sv-lncs ruhuna journal of science vol 9: 57-63, june 2018 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v9i1.33 university of ruhuna  faculty of science, university of ruhuna 57 sri lanka short paper optical and structural properties of cds thin films prepared using electro-deposition technique g.d.k. mahanama1*, d.a. madarasinghe1, w.g.d. dharmaratna1 and d. jayasundara2 1department of physics, university of ruhuna, matara, sri lanka 2 sri lanka institute of nanotechnology, pitipana, homagama, sri lanka correspondence: * mahanama@phy.ruh.ac.lk; orcid: 0000-0002-4761-4558 received: 30th october 2017, revised: 11th may 2018, accepted: 22nd june 2018 abstract. cadmium sulfide (cds) thin films were electrodeposited successfully on to indium tin oxide (ito) coated glass substrates from an aqueous solution of ph 1.4 containing 0.3m cdcl2 and 0.03m na2s2o3 or thiourea with the aim of using in cds/cdte solar cells. properties of cds thin films prepared at different deposition voltages, deposition time periods, deposition temperatures and annealing temperatures were investigated using the current-voltage (i-v) plots. it was found that good quality cds layers were formed under the deposition conditions of -1.13v for a period of 45 minutes in a solution of temperature at 46 oc. the performance of the cds layers was improved significantly after annealing the samples at 400 oc for a period of 20 min. the properties of cds thin films prepared by two and three electrode configurations and using two different electrolytes were compared using the current-voltage plots. it was found that there is a significant improvement of photocurrent of the samples prepared with two electrodes in thiourea as the s source in comparison with the samples prepared with na2s2o3. the analysis of xrd spectra showed the hexagonal crystal structure of cds films confirming the quality of the films prepared by this method. in addition, absorption spectra showed band gap value of 2.42 ev proving that the samples were of good quality. atomic force microscopy (afm) analysis showed that the roughness values of cds samples were in the range of 10-15 nm. film thicknesses of the samples were in the range of 175-225 nm according to the optical profilometric data. keywords. cadmium sulfide, electrodeposition, electrolyte, two-electrode, photocurrent. 1 introduction electrodeposition technique is a perspective competitor in thin film preparation because of several advantages such as the possibility for largeg.d.k. mahanama et al. optical and structural properties of cds thin films ruhuna journal of science 58 vol 9: 57-63, june 2018 scale production, minimum waste of components and easy monitoring of the deposition process (dharmadasa and haigh 2006). this technique is generally less expensive than the other methods. the composition of the electrolytes and deposition conditions play an important role in determining the quality of the deposited films. cds/cdte solar cells have emerged as a possible candidate for practical applications which can be produced at a low cost. here we report the results of the cds thin films prepared by varying the deposition voltage, period, temperature and annealing temperature. cds layers were studied using x-ray diffraction, scanning electron microscopy, atomic force microscopy, optical absorption spectroscopy and optical profilometry techniques. 2 materials & methods electrodeposition was performed in a conventional three-electrode cell. the three electrode cell contained a saturated potassium chloride calomel electrode (sce) as the reference, a platinum sheet as the counter electrode and an ito/glass substrate as the working electrode. a hokuto denko ha-301 potentiostat was used to control the electro-deposition process and to monitor the photocurrent and voltage profiles. all plates were cleaned using an ultrasonic bath. the ito coated glass substrates and counter electrode were polished and cleaned with acetone, methanol, dilute acetic acid and distilled water, respectively, before the deposition process. the bath consisted of cdcl2 (0.3m) and na2s2o3 (0.03m) is used. the ph of the bath was maintained at a constant value of 1.4 (dahiru 2011) throughout the study using hydrochloric acid. the solutions were prepared using analytical grade chemicals and de-ionized water. the area of the substrate to be deposited was kept fixed at 1.21 cm2 for all studies. the temperature of the electrolyte was maintained at a constant value during the deposition period. the photocurrent-voltage characteristics of cds deposited samples, as deposited as well as annealed at different temperatures (200–500 °c) were studied. an incandescent lamp of 100 w was used as the source of light to illuminate the cds layer for photocurrent-voltage measurements. the same procedure was repeated and the deposition process was carried out by changing the s source (thiourea) (aliyev and el-rouby 2013) instead of na2s2o3) and two electrode configuration instead of three electrodes. 3 results and discussion current-voltage characteristics of cds layers prepared at various deposition voltages [(-) 900 mv – (-)1300 mv] and as deposited and annealed samples g.d.k. mahanama et al. optical and structural properties of cds thin films ruhuna journal of science vol 9: 57-63, june 2018 59 were studied. the best deposition voltage was found to be -1100 mv from the above measurements. this was verified by a voltammogram (aliyev and elrouby 2013), which provided the exact deposition voltage (-1130 mv) for a stable cds thin layer. the deposited cds samples were annealed at different temperatures for 20 min and it was found that the best performance was shown by the layers annealed at 400 °c. the current-voltage curves of cds thin films deposited at different time periods (15 min to 60 min) were also studied. it was found that the layer thickness was low for a shorter deposition time period (15 min) due to insufficient time. as the deposition time was increased to 45 minutes the deposited films were smooth and adhered well on to the substrate. however, when the deposition time period was increased to 60 minutes and above the film dissolved into the bath due to the longer immersion time in the acidic medium (figen kadirgan et al. 1997). by varying the temperature of the electrolyte solution in the range 30 °c to 50 °c, it was found that good quality cds layers were produced at 46 °c. at lower temperatures (≤30 °c) the deposition process took a longer time. at higher temperatures (>50 °c) the deposition was rapid but the cds layer easily flaked off from the glass substrate showing a rapid decrease of current. fig. 1. absorption spectra of a good quality cds sample absorption spectra of prepared cds samples were measured in three different laboratories and all measurements showed that the band gap value was 2.42 ev confirming the quality of the samples. this is the exact value of the band gap expected from a good quality cds material according to the literature (sze 1981). absorption spectra of a good quality cds sample prepared is shown in figure 1. xrd measurements were taken in the range of 2θ = (20° 80°) using cu kα radiation (λ = 1.5418 å). the xrd spectrum obtained for a heat treated cds g.d.k. mahanama et al. optical and structural properties of cds thin films ruhuna journal of science 60 vol 9: 57-63, june 2018 layer deposited on glass/ito substrate using electrodeposition method is shown in figure 2. according to the standard x-ray diffraction patterns, the three broad peaks observed in the spectra at around 26.52°, 43.93°, 51.90° reveal a cubic lattice structure of cds (zinc blend). these peaks could be related to the (222), (220), and (322) planes of the cubic phase, respectively (reddy et al. 2003). the positions of several peaks were used to determine the cds phases and there are more hexagonal phases compared to the cubic phases. it is difficult to separate cubic peaks from hexagonal peaks because they are overlapping in the same regions. the results show that the films have highly oriented crystallites with the hexagonal structure (wurtzite type) with preferential orientation (aliyev and el-rouby 2013). fig. 2. xrd spectrum of a cds layer produced crystallite sizes related to xrd peaks shown in figure 2 were calculated using the scherer formula and given in the table 1. average grain size is ~ 69.1 nm which is correlated with sem image given below. table 1: crystallite sizes of xrd peaks shown in figure 2 2θ (degrees) fwhm hkl crystallite size (nm) 26.52 37.84 43.93 51.90 61.65 65.52 0.030 0.015 0.050 0.045 0.040 0.050 002 102 220 200 104 400 55.8 122.6 40.2 58.8 67.3 69.9 sem image of an annealed electrodeposited cds layer taken by the philips xl30 scanning electron microscope is shown in figure 3. the material g.d.k. mahanama et al. optical and structural properties of cds thin films ruhuna journal of science vol 9: 57-63, june 2018 61 clusters are in the nanoscale range with an average size of 175-200 nm. these grains seem to be packed together each other with small gaps in between. small gaps or pin holes seen in sem between the grains indicate the growth of cds as islands. fig. 3. sem image of heat treated ed-cds layer on glass/ito substrate atomic force microscopy (afm) measurements were performed on heat treated cds samples produced. the result shown in figure 4 indicates that the film roughness values are in the range of 10-15 nm. fig. 4. afm image of a cds sample g.d.k. mahanama et al. optical and structural properties of cds thin films ruhuna journal of science 62 vol 9: 57-63, june 2018 cds layers fabricated are densely packed in some regions and indicate perpendicular orientation to the glass/ito substrates and grow upwards after nucleation on the ito surface. the afm image reveals the existence of tightly packed nano-rods with length equal to 175-200 nm the cds layer. these observations are consistent with the sem results. the presence of nano-rods provides advantages because of band bending due to enhanced surfaces and hence creating an additional internal electric field perpendicular to their axis. it also minimizes recombination of electrons and holes in which the electrons flow along the axis of the nano-rod and holes could flow in the opposite direction along the vicinity of the surface layer of the nano-rod (dahiru 2011). thickness of the cds samples was measured using optical profilometer and the values are in the range of 175-225 nm. these values are very well matched with thickness values determined from afm measurements. 4 conclusion cds layers were successfully electrodeposited on ito glass substrate by using an aqueous solution containing 0.3m cdcl2 and 0.03m na2s2o3 or thiourea as the electrolyte. the analysis of photocurrent-voltage characteristics indicated that good quality cds layers were formed when the deposition was conducted for a period of 45 minutes under the deposition voltage of -1.13v in the solution of ph value 1.4 at the temperature of 46 oc. a more stable cds layer with better photo response was obtained by annealing the deposited layer at the temperature of 400 oc for 20 minutes. cds thin films prepared with thiourea as the s source have produced better currents than the films prepared with na2s2o3 as the s source. in addition, cds samples prepared with two-electrode configuration have shown an improvement of the photocurrent in compared with three-electrode configuration. this might be due to the poisoning of the electrolyte with the leakage of unwanted ions into the bath from the calomel electrode in the three electrode configuration. band gap of 2.42 ev calculated from absorption spectra of cds samples confirms the good quality of the sample prepared. this result was confirmed by measurements performed at three different laboratories. xrd spectra of electrodeposited cds samples have shown hexagonal phase crystal structure confirming the good quality of cds layers in compared with the cubic phase crystals reported with chemical bath deposition (cbd) method. atomic force microscopy data shows that the roughness values of the cds samples are in the range of 10-15 nm. film thickness was in the range of 175-225 nm. g.d.k. mahanama et al. optical and structural properties of cds thin films ruhuna journal of science vol 9: 57-63, june 2018 63 acknowledgments the authors thank turis project of university of ruhuna for the financial assistance and mr. w. m. n. m. b. wanninayake, department of physics, university of peradeniya for xrd measurements. prof. i.m. dharmadasa and mr. n. a. abdul-manaf at sheffield hallam university, uk are also acknowledged for the thickness measurements. three anonymous reviewers are acknowledged for their comments. references aliyev mas, el-rouby m. 2013. electrochemical studies on the cathode electrodeposition on n-type semiconductor cds thin films from thiosulfate acidic aqueous solution. international journal of thin film science and technology 2(3): 195-205. das sk, morris gc. 1993. preparation and characterization of electrodeposited ncds/pcdte thin film solar cells. solar energy materials and solar cells 28(4): 305-316. dahiru g. 2011. research and development of cdte based thin film pv solar cells. doctoral thesis, sheffield hallam university: p76 dharmadasa i, haigh j. 2006. strengths and advantages of electrodeposition as a semiconductor growth technique for applications in macro-electronic devices. journal of the electrochemical society (jes) 153(1): g47-g52. figen kadirgan d, mao a, balcioglu be, mccandless, song w, ohno tr, trefny ju. 1997. electrodeposited cds thin films and their application in cds/cdte solar cells, proceedings of 26th ieee photovoltaic specialists conference: 443-446 reddy ktr, shanthini gm, johnston d, miles rw. 2003. highly transparent and conducting cdo films grown by chemical spray pyrolysis, thin solid films 427 (1-2): 397-400. sze sm. 1981. physics of semiconductor devices, john wiley & sons, usa, 449 rjs-vol-1-sept-2006-5.dvi ruhuna journal of science vol. 1, september 2006, pp. 41–46 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. infestation of ceratothripoides claratris (shumsher) (thysanoptera: thripidae) on selected food crops in thailand w.t.s. dammini premachandra* institute of plant diseases and plant protection, university of hanover, herrenhäuser str. 2, 30419 hanover, germany. dammini@zoo.ruh.ac.lk christian borgemeister† institute of plant diseases and plant protection, university of hanover, herrenhäuser str. 2, 30419 hanover, germany. infestation of ceratothripoides claratris, the most prevalent thrip pest species on tomatoes in thailand was studied on eight different food crops, chili, cowpea, cucumber, eggplant, pumpkin, tomato, water melon and yard-long bean, in an open side-wall plastic house at the campus of asian institute of technology, thailand. percentage infested leaves and abundance of c. claratris on leaves of these crop species were determined at two-week intervals in three separate sampling occasions. the findings indicated that c. claratris has a potential to infest all the food crops tested. however, the abundance of thrips per unit leaf area and degree of infestation varied significantly. two varieties of tomatoes, i.e., king kong ii and luktho were found to be the most preferable food crop species for c. claratris. the degree of host preference of c. claratris was determined by the crop species, time and their combined effects. key words : abundance, ceratothripoides claratris, infestation, food crops. 1. introduction ceratothripoides claratris shumsher (thysanoptera:thripidae) is the predominant thrips species infesting tomatoes lycopersicon esculentum mill (solanaceae) in thailand. voracious feeding by larvae and adults, and/or oviposition by female c. claratris damage leaves, stems and fruits (murai et al. 2000; rodmui 2002; premachandra et al., 2004). in addition, it indirectly damages tomatoes by transmission of capsicum chlorosis virus (cacv) (genus tospovirus, family bunyaviridae) (premachandra et al., 2005a). hence, c. claratris causes significant yield losses on fieldand greenhouse-grown tomatoes in thailand, especially in greater bangkok area. previous studies indicated that c. claratris is particularly adapted to high temperatures frequently prevailing in tropics (premachandra et al. 2004). apart ∗ permanant address: department of zoology, university of ruhuna, matara, sri lanka. † permanant address: icepe african insect science for food and health, p.o. box 30772-00100, nairobi, kenya. 41 premachandra and borgemeister: infestation of ceratothripoides claratris ... 42 ruhuna journal of science 1, pp. 41–46, (2006) from thailand it has been reported in india (jangvitaya, 1993) and malaysia (s. okajima cited in murai et al., 2000). although c. claratris has been identified as a specialized feeder on tomatoes little is known about its host range. jangvitaya (1993) recorded luffa acutangula l. (cucurbitaceae) and clitoria ternatea l. (fabaceae) as host plants of c. claratris in thailand and this was the only published report on additional host plants of c. claratris. the objective of this study was to determine the level of infestation of c. claratris on selected food crops cultivated in thailand. this study was conducted as a part of a larger research project aiming to develop sustainable vegetable production under protected cultivation in the humid tropics. 2. materials and methods this trial was conducted in a 200m2 open side-wall plastic house (polythene plastic, 200 micron uv-stabilized polyfilm; ludvig svensson, kinna, sweden) with an opening of 50-200 cm above the ground level, located at the campus of the asian institute of technology (ait), in bangkok, thailand, during september october 2002. during the study period, the mean range of temperature and relative humidity were 26-28�and 60-80%, respectively. chili capsicum annuum l. (solanaceae), cowpea vigna sinensis (l) savexhass (fabaceae), cucumber cucumis sativus l. (cucurbitaceae), egg plant solanum xanthocarpum schard and wendl (solanaceae) varieties; green and purple, pumpkin curcurbita moschata (duch.) poir (cucurbitaceae), tomato, lycopersicon esculentum mill (solanaceae) varieties; king kong ii and “luktho”, water melon citrullus lanatus (cucurbitaceae) and yard-long bean vigna unguilulata (l.) walp (fabaceae) were included for the trial. these crop species were selected based on their importance for consumption in thailand, especially in greater bangkok area. twelve, three-week-old seedlings of each crop species were planted in plastic pots (30 by 25 cm) filled with a commercial growing substrate composed of clay, sand, and silt in proportions of 31, 30, and 39%, respectively, and 29% of organic matter. plants were arranged in four rows, i.e., 30 per row, in completely randomized design. the inter-pot and inter-row distances were 60 cm and 160 cm, respectively. irrigation and fertilization [fertilizers; [hakaphos (n-p-k), 2.5 kg/100 l; compo austria, and bai-plus (calcium), 1.8 kg/1001; bayer, bangkok, thailand]] were done seven to nine times per day (2.5 l/d) with a drip irrigation system controlled by solar light integral. plants were supported by ropes that were fixed to the structure of the plastic house. in this plastic house, plants were naturally colonized by c. claratris. previous studies confirmed that c. claratris is the predominant thrips species in the greenhouses at the ait campus (premachandra et al., 2004). thrips palmi karny (thysanoptera: thripidae) also occurs in fields and greenhouses in the greater bangkok area, though in very low numbers. three separate sampling occasions were performed in two-week time intervals commencing from 14 days post-planting of the crop species in the plastic house. at each sampling, three plants were selected at random from each food crop species and the total number of leaves and the number of thrips-infested leaves, i.e., thrips-feeding damage (e.g. leaves showing scars and necrosis), were recorded. in addition, at each sampling date, three fully premachandra and borgemeister: infestation of ceratothripoides claratris ... ruhuna journal of science 1, pp. 41–46, (2006) 43 expanded non-senescent thrips-infested leaves were picked at random from three randomly selected plants of each food crop species by pulling them into self-sealing plastic bags. thereafter, the sealed bags were transported to the laboratory and were washed three times for 10 s in a plastic box (159 cm) containing 250 ml of 70% ethanol. subsequently, the thrips-containing solution was poured into a conical fask (200 ml), shook thoroughly, and stored for 30 min for settling. thereafter, the supernatant was gently decanted to 50 ml, the remaining suspension was poured onto a counting plate, and the thrips, i.e., both adults and larvae, were counted under a stereomicroscope. total leaf area of sampled leaves was recorded using a leaf area meter (li-3100 area meter, li-cor. inc, nebraska, usa). percentage infested leaves per crop (i.e., percentage infestation) and abundance of thrips per unit leaf area (i.e., cm2) were calculated with respect to each food crop species. data on thrips abundance and percentage infested leaves were subjected log transformation and arcsine transformation, respectively, prior to the statistical analysis. two factorial anova was performed to compare the abundance of thrips and percentage infested leaves among the different plant species using proc glm (sas institute 1999). when the analysis of variance yielded significant f values, means were compared using the least significant difference (lsd) multiple range test. whenever significant interactions were observed between two factors, i.e., food crop species and sampling occasion, treatment means of one factor were compared at each level of the other factor. in contrast, no significant interactions were found means of the level of the one factor were compared irrespective of the levels of the other factor. all analyses were performed at 5% significance level using sas statistical package (sas institute, 1999). 3. results and discussion the mean abundance of c. claratris per cm2 leaf area varied significantly among the different food crop species (f = 21.70; df = 9, 60; p < 0.0001) and sampling occasions (f =18.16; df = 2, 60; p < 0.0001). in addition, significant interactions were detected between the food crop species and the sampling occasions (crop species*sampling occasion: f = 6.12; df = 18, 60; p < 0.0001). hence, thrips abundance on each food crop species was compared with each of the sampling occasion. the highest c. claratris abundance per cm2 was recorded on tomato-king kong ii followed by tomato-luktho and water melon on the second sampling occasion, but was not significantly different. the thrips abundance was significantly lower (p < 0.05) on chili, cowpea and yard-long bean on all the three sampling dates. in general, in the first sampling, highest number of thrips was recorded on cucumber and water melon while in the second and third sampling highest thrips abundance was detected on two tomato varieties. percentage infested leaves also differed significantly among the different crops species (f =23.49; df = 9, 60; p < 0.0001) over three sampling occasions (f = 29.81; df = 9, 60; p < 0.0001). moreover, significant interactions were detected between the crops species and sampling occasions (crop species*sampling occasion: f = 4.73; df = 18, 60; p < 0.0001). hundred percent of the leaf infestation was detected on premachandra and borgemeister: infestation of ceratothripoides claratris ... 44 ruhuna journal of science 1, pp. 41–46, (2006) figure 1 total number of c. claratris (mean (± se) per unit leaf area (cm2) on different food crops at three sampling occasions). cucumber, both varieties of tomato, water melon and yard-long bean. the lowest leaf infestation was found on chili (figure 2). only on cucumber maximum leaf infestation, i.e., 100% was detected at all the sampling occasions. this was the first study on infestation of c. claratris on different food crops. our results indicated that all the food crops tested in this trial could act as potential hosts for c. claratris. however, these food crops differed in their suitability as hosts for c. claratris. the degree of host preference of c. claratris was determined by the crop species, time and their combined effects. the higher abundance per unit leaf area coupled with 100% leaf infestation on two tomato varieties indicated that tomatoes were good host plants for c. claratris. conversely, chili was not considered as a suitable host plant for c. claratris. rosenthal and berenbaum (1991) indicated that differences in host plant preferences of a particular pest species depends on differences in food quality determined by the level of primary and secondary plant metabolites. it has also been shown that many plants contain secondary substances that deter feeding and oviposition in phytophagous insects (gupta and thorsteinson,1960). previous research work indicated that natural infestation by c. claratris commenced one week post-planting of tomatoes in the greenhouses at the ait campus. after invasion the prevailing microclimatic conditions (i.e., high temperatures and relative humidity) inside the plastic house provided favorable conditions for growth and reproduction of this thrips species (premachandra et al., 2004). the higher thrips abundance of c. claratris on tomatoes, eggplants, pumpkin and water melon on the second sampling occasion could most probably be associated premachandra and borgemeister: infestation of ceratothripoides claratris ... ruhuna journal of science 1, pp. 41–46, (2006) 45 figure 2 mean (± se) percentage leaf infestation of c. claratris on different food crops at three sampling occasions. with their reproduction on these food plants after invasion. it was also reported that the abundance of c. claratris on tomatoes started to decline five weeks postplanting tomatoes in the greenhouse at ait campus (premachandra et al., 2005b) due to deterioration of leaves caused by voracious feeding of the thrips adults and larvae, corroborating findings of the current study. though 100% infestation was found on cucumber, water melon and yard-long bean the thrips abundance was substantially low. this implies these food plants were preferable for feeding, but not for reproduction for c. claratris which act as the main factor for rapid increase in thrips population. however, in this study, the degree of damage on leaves was not estimated in this study. in conclusion, our study added ten additional host plants for the host range of c. claratris implying c. claratris has a broad host range. this information is important for manipulating of plant species in crop rotations. references gupta pd, thorsteinson aj. 1960. food plant relationships of the diamondback moth [plutella maculipennis (curt.)]. ii. sensory regulation of oviposition of the adult female. entomologia experimentalis et applicata. 3: 305-314. jangvitaya p. 1993. studies on the family thripidae (insecta: thysanoptera) from thailand. m.sc. thesis, tokyo university of agriculture, tokyo, 254 pp. premachandra and borgemeister: infestation of ceratothripoides claratris ... 46 ruhuna journal of science 1, pp. 41–46, (2006) murai t, kawai s, chongratanameteekul w, nakasuji f. 2000. damage to tomato by ceratothripoides claratris (shumsher) (thysanoptera: thripidae) in central thailand and a note on its parasitoid, goethena shakespearei girault (hymenoptera: eulophidae). applied entomology and zoology 35: 505-507. premachandra wtsd, borgemeister c, chabi-olaye a, poehling hm. 2004. influence of temperature on the development, reproduction and longevity of ceratothripoides claratris (thysanoptera: thripidae) on tomatoes. bulletin of entomological research 94: 377-384. premachandra wtsd, borgemeister c, maiss e, knierim d, poehling hm. 2005a. ceratothripoides claratris, a new vector of a capsicum chlorosis virus isolate infecting tomato in thailand. phytopathology 95: 659-663. premachandra wtsd, borgemeister c, mamoudou s, achilles t, poehling hm. 2005b. spatio-temporal distribution of ceratothripoides claratris (thysanoptera: thripidae) on tomatoes in thailand. environmental entomology, 34: 883-890. rodmui p. 2002. population dynamics and biological control of thrips, ceratothripoides claratris (shumsher) (thysanoptera: thripidae), on tomato under protected cultivation in thailand. m.sc. thesis, kasetsart university, bangkok, thailand, 50 pp. rosenthal ga, berenbaum mr. 1991. herbivores, their interaction with secondary plant metabolites, 2nd edn, vol 1. new york: academic press. sas institute. 1999. sas/stat user’s guide. sas institute, cary, nc. acknowledgments this study was funded by the german research council (deutsche forschungsgemeinschaft) within the framework of the dfg-for 431 project. rjs-vol-1-sept-2006-13.dvi ruhuna journal of science vol. 1, september 2006, pp. 125–132 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. clinics management system (cms) based on patient centric process ontology prasad jayaweera, jeewanie jayasinghe, lal wellakkage vajira malawaraarachchi, samantha liyanage department of computer science, university of ruhuna, matara, sri lanka, prasad@cc.ruh.ac.lk, jeewanie@cc.ruh.ac.lk, lalw@cc.ruh.ac.lk, vajiram@cc.ruh.ac.lk, rls@cc.ruh.ac.lk, piyal perera general hospital matara constantly created new specialties, different roles, public and private organizations in patients’ healthcare have resulted complexity issues. an ontological framework is very much promising to develop interoperable it solutions for various segments of today’s healthcare systems. in this paper we report an on-going project that develops patient centric process ontology. as a useful application of this framework, we brief clinics management system (cms) for general hospital, matara, sri lanka. key words : ontology, patient centric process, layered processes, clinics management 1. introduction a major problem faced by today’s healthcare is the increasing diversity and differentiation. new medical specialties are constantly created: there exist a large number of roles in patient care, and many different public and private organizations. all these roles in healthcare organizations are involved in the care of a single patient. this differentiation creates a need for the interoperability of organizations as well as their supporting systems and applications. a key instrument for achieving such interoperability is to develop a sharable healthcare ontology. however, today’s process oriented business ontologies are developed in customer centered manner in order to develop business applications to provide more and more value added services to consumers. this trend is also valid in healthcare sector although there are some restrictions and limitations that can be noticed in sri lankan scenarios. in this context, we would like to report on some interesting and noticeable developments are taking place at the general hospital, matara, sri lanka. among these, one of the initial it solution, called cms (clinic management system) has been introduced in this paper. cms will be one of the sub systems developed by the (ec4)1 research group at the department of computer science, university of ruhuna to be deployed at general hospital, matara. 1 (ec4) : electronic commerce and collaboration competence center, http://www.ruh.ac.lk/uni/science/computer/ec4 125 prasad m. jayaweera, et. al.: clinics management system... 126 ruhuna journal of science 1, pp. 125–132, (2006) the rest of the paper is organized as follows: section 2 reports related research in process oriented ontology in general and in the healthcare context. section 3 describes proposed patient centric process ontology developed in the selected domain. in sections 4, as a useful real world application, we summarize the proposed system giving its basic functionalities. the paper concludes with explaining possible benefits and future directions of the clinic management system. 2. related research one of the important trends in business managements is the focus on processes to create value added services for their ultimate customers. this is to get rid of unwanted, time consuming, unnecessarily repeated business activities of the processes and to monitor how value for customer is as efficiently as possible (vissers j. m. h, 1998, (johannesson and jayaweera, 2000). this is also very much valid for healthcare industry and there is a great need for process orientation and transparent communication between various healthcare roles and healthcare applications. in addition to process orientation, to achieve aforementioned inter and intra communication (or in other words interoperability) between healthcare units, utilization of sharable healthcare ontology is the promising approach. in general, “an ontology is an explicit specification of a conceptualization”, which is a widely accepted definition by gruber (gruber, 1993). this definition is an elaboration on “an ontology is the object, concepts, other entities that are assumed to exist in some area of interest and the relationships that hold among them” found in genesereth and nilsson’s work (genesereth and nilsson 1987). we adapt these definitions also in developing our healthcare ontology. as mentioned earlier, patient centric approaches are providing more competitive advantages also in healthcare industry. some of the interesting work in these converging areas are listed below. interesting orthodox process thinking can be found in samba (samba, 2006) (structured architecture for medical business activities). samba is a process modeling technique to define the concepts of the health care processes. the concepts to support continuity of healthcare processes have been dealt in contsys (contsys, 2006). for the development of proposed healthcare ontology, we have looked into the above mentioned and some other related approaches. this healthcare ontology is the basis of clinic management system developed for the general hospital, matara. in our work, bpmn (business process modeling notation) has been used as notational framework for process specifications. the main motivation of selecting bpmn is the popularity of service composition in particular web services for specification and enactment of business processes. in that context, bpmn is much more convenient in mapping to process execution languages such as wsbple (wsbple, 2006). however, complete service specification and healthcare process enactment and monitoring are out of the scope of this paper. prasad m. jayaweera, et. al.: clinics management system... ruhuna journal of science 1, pp. 125–132, (2006) 127 3. patient centric process ontological framework the development of patient centric process ontology for healthcare mainly focuses on the different roles around a patient to provide different healthcare services and different activities those roles are to play in order to provide services. however, construction of a single complete ontological framework to cover all aspects of healthcare is tedious and time-consuming task. in this work we have selected one of the initial phases to offer specialty-services at clinics in general hospitals. the paper reports on-going project at the department of computer science, university of ruhuna to develop clinics registration system for the general hospital, matara, based on the proposed ontological framework. 3.1. three layered processes the clinic management of general hospitals are complex involving many different healthcare roles and systems. these roles perform activities in providing wide spectrum of healthcare and related services towards a patient. as we have been proposing in our earlier work [jayaweera and johannesson, 2004], one of the approaches to tackle down this inherited complexity in healthcare is also to apply separation of concern where we take particular perspective at a time. this approach results three different layers that can be considered different activities happening in selected domains. the following are those we can consider in our study at the general hospital, matara: 1. referring processes 2. registering processes 3. clinical processes some of the high level activities that are possible in each of layers mentioned above are shown in figure 1. bpmn (bpmn) notation has been used here to illustrate process semantics. 3.1.1. referring process we have identified referring healthcare process as the first triggering process of particular patient’s clinic management scenario. in the referring party set, there can be various healthcare roles directing patients to relevant clinics. illustrated figure 1 in above is a couple of typical such activities where the referring parties could play for a given case. a patient could be directed to a specific clinic by a consultant, a district medical officer, another clinic, or some other authorized healthcare unit or personnel. 3.1.2. registering process having referred to a specific clinic by an authorized healthcare personnel or unit, it is the registering process which triggers next. this process basically collects required information about the patient and about the party who made the reference together with referral note. as the final step of the process it makes an appointment to meet a physician at the clinic. 3.1.3. clinical process final and the most important healthcare process is the clinical process where patient is subjected to clinical diagnosis, tests and various forms of healthcare treatment. however, for the sub-systems that we are discussing in the paper detailed coverage of these processes are beyond our scope. prasad m. jayaweera, et. al.: clinics management system... 128 ruhuna journal of science 1, pp. 125–132, (2006) figure 1 three main layers of clinic management processes. figure 2 metamodel of the clinic registration system (key healthcare roles). 3.2. clinics registration ontology as introduced in the section 2, there are many research going on in ontology development in general and also healthcare ontology in particular. however , rather than adapting very formal specification, we have developed a metamodel for our purpose as depicted below in figure 2. prasad m. jayaweera, et. al.: clinics management system... ruhuna journal of science 1, pp. 125–132, (2006) 129 figure 3 architecture of the proposed clinic management system. in the above metamodel, different roles in registration of clinics are highlighted. our understating here is identification of roles in healthcare is the key to model different activities those are authorized and in position to perform (see figure 2). even in the metamodel for clinic management system, we can notice that the patient being the centered concept and very much motivated in selecting patient centered approach. the two common abstraction levels defined in (fowler, 1997) and (martin and odell, 1994), are the operational level and the knowledge level. the operational level models concrete, tangible individuals in a domain. the knowledge level models information structures that characterise categories of individuals at the operational level. martin and odell, (martin and odell, 1994), employ the concept of power types to refer the correspondence between the objects of the knowledge and operational levels. a power type is a class whose instances are subtypes of another class. the power type of clinic, called clinic type, is also added for the purpose of differentiating between the modeling of concrete, tangible objects in a domain, and the abstract characteristic categories of these objects. 4. clinic management system in healthcare industry there is an increasing demand to find affordable, easy-todeploy technology to improve patient oriented services and communications effectiveness among all the parties involving in the hospital activities. the patient centric process ontology which was proposed in section 3.2, has been adapted in the development of our clinic management system. in figure 3, the architecture of the clinic management system is shown. the main idea here is to develop and maintain a central knowledgebase that extracts necessary information from the databases layer down. there may be several of databases to hold patient, clinic and healthcare roles’ information. there are five interfaces to the cms knowledgebase. three of them are to interact and to monitor the process layers mentioned in section 3.1. the remaining two prasad m. jayaweera, et. al.: clinics management system... 130 ruhuna journal of science 1, pp. 125–132, (2006) interfaces are to provide value added healthcare services to patients and also to administrative services to relevant authorities. 4.1. clinic management at the general hospital, matara the first step of the clinic process starts from directing patient to a clinic by different referring parties. a referring party can be a consultant, other hospital, a ward of a hospital, the opd (out door patient department) or any other medical authority. the next step of the process is registering a patient to a particular clinic by a registering officer. after the registration the patients are directed to the relevant clinic for the clinical activities. in each clinic the patient is examined by a consultant at the first time. the patient has to visit the clinic more than once in some cases. in the existing procedure, each and every time the patient has to register to a clinic in order to get an appointment to meet a consultant. the proposed system is designed to cater different types of users with different levels of accessibilities to the clinic information. the main purpose of having different user levels is to provide more secure and flexible environment to all the participants in the clinic system. the subsystems described here work with existing manual systems at the hospital. therefore to support the existing manual processes patient registry information are printed on a sticker and could be pasted on patient’s clinic record document (which is currently a printed book). based on the architecture of clinic management system the portal which is designed to register a new patient is described in the next section. 4.1.1. registering a new patient registering officer has the authority to register a patient. the proposed system is to issue a colored registration sticker for the patient at different stages when they are directed to a clinic. the registration sticker holds the registration number, clinic number, name of the patient, clinic date and the time and the room number. the color codes are as follows. the red color issued for those patients who are registered for a clinic at the first time. the blue color is for the patients second and consequent visits and the black color is for patients those who are to submit prescribed test reports (e.g. blood tests, xrays, etc). then the authorized persons can identify the patients by the color code and it will help arrange clinical activities in efficient manner. figure 4 is the web portal to register a patient at the first time to a clinic (issuing red color card). in this interface it could be noticed that there are three different areas of information shown. firstly, patient information; secondly, reference made; finally, the clinic details. when selecting the clinic, some essential data is filled automatically such as the registration number of the patient, the room number and the current date and time. in the appointment time slot, it contains additional information such as the number of maximum places for that slot within brackets and occupied places immediately left outside the brackets. this will be much helpful for patients when finding most convenient time slot to come to clinics. 5. concluding remarks and future directions in this paper we have reported an on going project at ec4 research group. the theoretical contributions of this work is the development of patient centric process prasad m. jayaweera, et. al.: clinics management system... ruhuna journal of science 1, pp. 125–132, (2006) 131 figure 4 new patient registration portal. ontological framework to serve as the basis for achieving interoperability in healthcare systems and its applications and also for developing value added healthcare services for consumers, in this context patients. as a real applicability of proposed patient centric ontological framework that has been proposed, we have developed an application to clinic management at general hospital, matara. although the system yet to be deployed, we are very much certain of possible benefits not only to patients but also healthcare personnel to carry out in their day-to-day activities. references bij j.d. v. d., dijkstra l., 1999, vries g. d., walburg j., improvement and renewal of healthcare processes: results of an empirical research project, health policy , 48, pp. 135-152. 26. bpel4ws, business process execution language for web services, oasis wsbpel technical committee, valid on 20060623, http://www.ebpml.org/bpel4ws.htm bpmn, business process modeling notation, valid on 20060623, http://www.bpmn.org. contsys, system of concepts to support continuity of care, cen/tc 251, valid on 20060624, http://www.centc251.org. fowler m., 1997, analysis patterns: reusable object models, addison-wesley genesereth m. r., & nilsson n. j., 1987, logical foundations of articial intelligence, kaufmann, los altos, ca, 3. prasad m. jayaweera, et. al.: clinics management system... 132 ruhuna journal of science 1, pp. 125–132, (2006) gruber t., 1993, a translation approach to portable ontology, knowledge acquisition. jayaweera p. and johannesson p.,2004, a patient centred process ontology for information visualisation in health care, emoi interop 2004 (enterprise modeling and ontologies for interoperability) at 16th international conference on advanced information systems engineering (caise ’04), riga-latvia. johannesson p., wangler b., jayaweera p., 2000, application and process integration concepts, issues, and research directions, brinkkemper s., lindencrona e., sølvberg a, editors, information systems engineering, state of the art and research themes, springer. martin, j., odell, j., 1994, object-oriented methods. a foundation, prentice hall. samba, structured architecture for medical business activities, valid on 20060623, http://www.centc251.org/tcmeet/doclist/doclist2006.htm scheer a.-w.,2000, aris business process frameworks, springer. vissers j. m. h.,1998, health care management modelling: a process perspective, health care management science, 1:77-85. acknowledgments authors would like to acknowledge all the support and guidance received without any constraint from prof. ranjith senaratne, vice chancellor of the university of ruhuna, matara and dr. k.i padmatilake, director of the general hospital matara. rjs-vol-1-sept-2006-4.dvi ruhuna journal of science vol. 1, september 2006, pp. 32–40 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. variation of magnetic energy in oriented ultra-thin ferromagnetic films p. samarasekara department of physics, university of ruhuna, matara, sri lanka, pubudus@phy.ruh.ac.lk the angle dependence of magnetic energy of ultra-thin films with n=1 to 5 layers has been investigated. the effect of stress and the demagnetization factor on the classical heisenberg hamiltonian has been taken into account in addition to other magnetic factors. films of sc(001) and bcc(001) ferromagnetic indicate preferred in plane and perpendicular orientations for n≤3 and n≥4, respectively. easy and hard directions of all sc(001), bcc(001) and fcc(001) films with five layers are θ = 7.2°and θ = 114°respectively. according to our simulation, ultra-thin films with one or two layers can be easily oriented in θ = 279°direction. key words : spin, magnetic moment, magnetic anisotropy, ferromagnetic materials, thin films, heisenberg hamiltonian 1. introduction. although the experimental evidence of oriented and non-oriented ferromagnetic and ferrimagnetic films can be found in many early reports, the theoretical aspects of oriented ferrimagnetic films were not revealed in a considerable manner yet. the sputter synthesis of aligned strontium hexaferrites, sro.6(fe2o3), films can be given as an example of the deposition of oriented ferromagnetic film (hegde et al. 1994a). the sputter synthesis of tbcu7 type sm(cofecuzr) films with controlled easy axis orientation (hegde et al. 1994b) and nitriding studies of aligned high anisotropy thmn12-type ndfe11co1−ymoyn film samples can be found in some experimental reports of oriented films (navarathna et al. 1994). easy axis oriented nickel ferrite (samarasekara et al. 1996) and lithium mixed ferrite (samarasekara and cadieu 2001) thin films have been fabricated on single crystal substrates by means of pulse laser deposition technique. the magnetic properties of ferromagnetic films have been investigated using bloch-spin wave theory (martin and robert 1951). monte carlo simulation of hysteresis loops of fef2(110)/fe bilayers has been performed to study the exchange interaction between antiferromagnetic and ferromagnetic sublattices (lederman et al. 2004). the energy of oriented ferromagnetic thick films has been calculated using the heisenberg hamiltonian in one of our early report. the energy of oriented ultra-thin ferromagnetic films will be described in this report. the effect of spin exchange interaction, magnetic dipole interaction, anisotropies up to eighth order, demagnetization factor and stress induced anisotropy on heisenberg hamiltonian was studied. 32 samarasekara: variation of magnetic energy... ruhuna journal of science 1, pp. 32–40, (2006) 33 2. model and discussion the classical heisenberg hamiltonian for ferromagnetic thin films can be given as following. h = − j 2 ∑ m,n ~sm.~sn + ω 2 ∑ m 6=n ( ~sm.~sn r3mn − 3(~sm.~rmn)(~rmn.~sn) r5mn ) − ∑ m d (2) λm (szm) 2 − ∑ m xd (4) λm (szm) 4 − ∑ m d (6) λm (szm) 6 − ∑ m d (8) λm (szm) 8 − ∑ m,n [ ~h − (nd ~sn/µ0)].~sm − ∑ m ks sin 2θm above equation for oriented ferromagnetic thin films can be deduced to following form (usadel and hucht 2002). e(θ) = − 1 2 n ∑ m,n=1 (j z|m−n| − ω 4 φ|m−n|) + 3ω 8 cos 2θ n ∑ m,n=1 φ|m−n| − cos2 θ n ∑ m=1 d(2)m − cos 4 θ n ∑ m=1 d(4)m − cos6 θ n ∑ m=1 d(6)m − cos 8 θ n ∑ m=1 d(8)m − n (hin sin θ + hout cos θ − nd µ0 + k5 sin 2θ) here j, d(2)m , d (4) m , d (6) m , d (8) m , n, hin, hout, nd, k5, z|m−n|, φ|m−n|, ω and θ are spin exchange interaction, second, fourth, sixth and eight order anisotropy coefficients, number of layers of the film, in plane applied magnetic field, out of plane applied magnetic field, demagnetization factor, stress induced anisotropy coefficient, number of nearest neighbors between layers, constants related to partial summation of dipole interaction, strength of long range dipole interaction and the angle between the spin and the film normal, respectively. for microscopic anisotropy, sixth and eighth order anisotropies are also induced. for an ultra thin film with few layers, it is reasonable to assume that d(2)m , d (4) m , d (6) m , d (8) m are constants. because zδ>1 = 0, ∑n m,n=1 z|m−n| = n z0 + 2(n − 1)z1. after assuming that φδ>1 = 0, ∑n m,n=1 φ|m−n| = n φ0 + 2(n − 1)φ1. then above equation can be reduced into following form, e(θ) = − j 2 [n z0 + 2(n − 1)z1] + ω 8 [n φ0 + 2(n − 1)φ1](1 + 3 cos 2θ) −n [d(2)m cos 2 θ + d(4)m cos 4 θ + d(6)m cos 6 θ + d(8)m cos 8 θ +hin sin θ + hout cos θ − nd µ0 + k5 sin 2θ] (1) samarasekara: variation of magnetic energy... 34 ruhuna journal of science 1, pp. 32–40, (2006) figure 1 the graph between number of layers and the angle (radians) for minimum energy. for the minimum energy, ∂e ∂θ = 0, and following equation of n can be obtained. n = 1.5ωφ1 sin 2θ k . here, k = 0.75ω(φ0 + 2φ1) sin 2θ − 2d (2) m sin θ cos θ − 4d(4)m sin θ cos 3 θ − 6d(6)m sin θ cos 5 θ − 8d(8)m sin θ cos 7 θ + hin cos θ − hout sin θ + 2k5 cos 2θ k5 ω = 10, d(2)m ω = 30, d(4)m ω = 20, d(6)m ω = 10, d(8)m ω = 5, hin ω = hout ω = 10, and nd µ0ω = 10 will be used for the simulation in this report. for sc(001) lattice, φ0 = 9.0336, φ1 = −0.3275 and n = − 0.4913 sin 2θ k . here k = −23.716 sin 2θ − 20(4 + 3 cos2 θ + 2cos4θ) cos3 θ sin θ + 10(cos θ − sin θ) + 20 cos 2θ. the graph between n and angle (θ) in radians is given in figure 1. the angles indicated in all the graphs are given in radians. according to this graph, films with n=1, 2 can be easily oriented in θ = 279°direction. similarly n corresponding to the minimum energy of fcc(001) and bcc(001) can be found using above equation. when the summations of d(2)m , d (4) m , d (6) m , d (8) m are considered, e(θ) = − j 2 [n z0 + 2(n − 1)z1] + ω 8 [n φ0 + 2(n − 1)φ1](1 + 3 cos 2θ) samarasekara: variation of magnetic energy... ruhuna journal of science 1, pp. 32–40, (2006) 35 − cos2 θ n ∑ m=1 d(2)m − cos 4 θ n ∑ m=1 d(4)m − cos 6 θ n ∑ m=1 d(6)m − cos 8 θ n ∑ m=1 d(8)m − n (hin sin θ + hout cos θ − nd µ0 + k5 sin 2θ) when θ = 0, e(0) = − j 2 [n z0 + 2(n − 1)z1] + ω 2 [n φ0 + 2(n − 1)φ1] − n ∑ m=1 d(2)m − n ∑ m=1 d(4)m − n ∑ m=1 d(6)m − n ∑ m=1 d(8)m − n (hout − nd µ0 ) when θ = 90°, e(90°) = −j 2 [n z0 + 2(n − 1)z1] − ω 4 [n φ0 + 2(n − 1)φ1] − n (hin − nd µ0 ) when e(0) > e(90), 3ω 4 [n φ0 + 2(n − 1)φ1] + n (hin − hout) > n ∑ m=1 d(2)m + n ∑ m=1 d(4)m + n ∑ m=1 d(6)m + n ∑ m=1 d(8)m when e(0) < e(90), 3ω 4 [n φ0 + 2(n − 1)φ1] + n (hin − hout) < n ∑ m=1 d(2)m + n ∑ m=1 d(4)m + n ∑ m=1 d(6)m + n ∑ m=1 d(8)m according to above equations, the in plane or perpendicular orientation is preferred depending on the marginal value of n ∑ m=1 d(2)m + n ∑ m=1 d(4)m + n ∑ m=1 d(6)m + n ∑ m=1 d(8)m given by following equation. n ∑ m=1 d(2)m + n ∑ m=1 d(4)m + n ∑ m=1 d(6)m + n ∑ m=1 d(8)m = 3ω 4 [n φ0 + 2(n − 1)φ1] + n (hin − hout) samarasekara: variation of magnetic energy... 36 ruhuna journal of science 1, pp. 32–40, (2006) for sc(001) with hin = hout n ∑ m=1 d(2)m + n ∑ m=1 d(4)m + n ∑ m=1 d(6)m + n ∑ m=1 d(8)m = 6.284n + 0.491 the anisotropy energy of a thin film is given by following equation. e(0) − e(90°) = 3ω 4 [n φ0 + 2(n − 1)φ1] − n ∑ m=1 d(2)m − n ∑ m=1 d(4)m − n ∑ m=1 d(6)m − n ∑ m=1 d(8)m + n (hin − hout) for hout = 0, e(0) − e(90) ω = 16.284n − 64.51 when n ≥ 4, e(0)>e(90°) and perpendicular orientation is preferred. when n ≤ 3, e(0)<e(90°) and in plane orientation is preferred. for bcc(001) φ0 = 5.8675, φ1 = 2.7126 e(0) − e(90) ω = 18.47n − 69.07 therefore, the marginal values of number of layers for bcc(001) is same as those for sc(001). using z0=4 and z1=1 for sc(001) ferromagnetic film, j ω = 10 and above other values in equation (1) e(θ) ω = − 18.95n + 10.082 + (3.142n + 0.246) cos 2θ − n [cos2 θ(30 + 20 cos2 θ + 10 cos4 θ + 5 cos6 θ) + 10 sin θ + 10 cos θ + 10 sin 2θ] 3-d graph of energy versus n and θ is shown in figure 2. for n = 5 e(θ) ω = −84.67n + 15.96 cos 2θ − 5[cos2 θ(30 + 20 cos2 θ + 10 cos4 θ + 5 cos6 θ) + 10 sin θ + 10 cos θ + 10 sin 2θ] graph between energy and θ is given in figure 3. for bcc(001) ferromagnetic thin films, z0=0, z1=4, φ0 = 5.8675, φ1 = 2.7126 e(θ) ω = − 28.6n + 39.322 + (4.235n − 2.034) cos 2θ − n [cos2 θ(30 + 20 cos2 θ + 10 cos4 θ + 5 cos6 θ) + 10 sin θ + 10 cos θ + 10 sin 2θ] samarasekara: variation of magnetic energy... ruhuna journal of science 1, pp. 32–40, (2006) 37 figure 2 3-d graph of energy versus n and θ (radians) for sc(001) ferromagnetic film figure 3 graph between energy and θ (radians) for n=5 of sc(001) ferromagnetic film 3-d graph of energy versus n and θ is shown in figure 4. for n=5, e(θ) ω = − 103.68 + 19.141 cos2θ − 5[cos2 θ(30 + 20 cos2 θ + 10 cos4 θ + 5 cos6 θ) + 10 sin θ + 10 cos θ + 10 sin 2θ] samarasekara: variation of magnetic energy... 38 ruhuna journal of science 1, pp. 32–40, (2006) figure 4 3-d graph of energy versus n and θ (radians) for bcc(001) ferromagnetic film figure 5 3-d graph of energy versus n and θ (radians) for fcc(001) ferromagnetic film. for fcc(001) ferromagnetic thin films, z0=4, z1=4, φ0 = 9.0336, φ1 = 1.4294 e(θ) ω = − 48.51n + 39.643 + (4.46n − 1.072) cos 2θ − n [cos2 θ(30 + 20 cos2 θ + 10 cos4 θ + 5 cos6 θ) + 10 sin θ + 10 cos θ + 10 sin 2θ] 3-d graph of energy versus n and θ is shown in figure 5. samarasekara: variation of magnetic energy... ruhuna journal of science 1, pp. 32–40, (2006) 39 figure 6 graph between energy and θ (radians) for n=5 of fcc(001) ferromagnetic film. for n=5, e(θ) ω = − 203 + 21.23 cos 2θ − 5[cos2 θ(30 + 20 cos2 θ + 10 cos4 θ + 5 cos6 θ) + 10 sin θ + 10 cos θ + 10 sin 2θ] graph between energy and θ is given in figure 6. according to these graphs, first minimum value of e(θ) ω for sc(001), bcc(001) and fcc(001) are -460, -470 and -570, respectively. therefore, fcc(001) can be oriented easily in the easy direction compared with other two. easy (θ = 7.2°) and hard (θ = 114°) directions for all three types with five layers remain same. but the maximum and minimum energies are different for three different types. also the maximum and minimum energy in each graph gradually increases with angle. 3. conclusion the magnetic energy of ultra-thin films with n=1 to 5 layers mainly depends on the azimuthual direction of magnetic moment orientation. the effect of stress induced anisotropy and the out of plane demagnetization factor on the heisenberg hamiltonian has been taken into account in addition to other magnetic factors. films of sc(001) and bcc(001) ferromagnetic with n ≤3 and n ≥4 layers can be preferably oriented for in plane and perpendicular directions, respectively. easy and hard directions of all sc(001), bcc(001) and fcc(001) with n=5 were found to be θ = 7.2°and θ = 114°, respectively. ultra-thin films with n=1 or 2 can be easily oriented in θ = 279°direction. samarasekara: variation of magnetic energy... 40 ruhuna journal of science 1, pp. 32–40, (2006) references hegde h, samarasekara p and cadieu f.j 1994a. nonepitaxial sputter synthesis of aligned strontium hexaferrites, sro.6(fe2o3), films, j. appl. phys., 75(10): 6640-6642 hegde h, samarasekara p, rani r, nanavarathna a, tracy k, and cadieu fj 1994b. sputter synthesis of tbcu7 type sm(cofecuzr) films with controlled easy axis orientation, j. appl. phys. 76(10): 6760-6762 lederman david, ricardo ramirez and miguel kiwi 2004. monte carlo simulations of exchange bias of ferromagnetic thin films on fef2(110), phys. rev. b70: 184422 1-7 martin j. klein and robert s. smith 1951. thin ferromagnetic films, phys. rev. 81: 378-380 navarathna a, samarasekara p, hegde h, rani r and cadieu fj 1994. nitriding studies of aligned high anisotropy thmn12-type ndfe11co1-ymoyn film samples, j.appl.phys. 76(10): 6068 samarasekara p, rani r, cadieu fj and shaheen sa 1996. variable texture niofe2o3 ferrite films prepared by pulsed laser deposition , j. appl. phys. 79(8): 5425-5427 samarasekara p and cadieu fj 2001. magnetic and structural properties of rf sputtered polycrystalline lithium mixed ferrimagnetic films , chinese j. phys. 39(6): 635-640 usadel kd and hucht a 2002. anisotropy of ultrathin ferromagnetic films and the spin reorientationh transition, phys. rev. b 66: 024419 1-6 ruhuna journal of science vol 8: 103-111, december 2017 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v8i2.29 university of ruhuna  faculty of science, university of ruhuna sri lanka 103 short paper comparative gc-ms analysis of all curcuma species grown in sri lanka by multivariate test h.m.i.c. herath, t.d.c.m.k. wijayasiriwardene*, g.a.s. premakumara industrial technology institute, bauddaloka mawatha colombo, sri lanka *correspondence: drchandima@iti.lk received: 26 th october 2017, revised: 28 th december 2017, accepted: 30 th december 2017 abstract curcuma is clinically valuable genus in traditional medicine. people use various plants under the same vernacular name may lead to adulteration or substitution. whole plants of curcuma species were collected in 2016 in the flowering season. voucher specimens of the plants were authenticated from the national herbarium, peradeniya. essential oils were extracted from clevenger’s apparatus and analyzed separately by gc-ms. the analyses were carried out with rtx wax capillary column. sampling and experiments were done according to who guidelines. one hundred sixty four phytochemicals were analyzed by simple correspondence and by cluster variable method. by cluster varibale as per phytochemicals present, mainly two groups were identified. c. albiflora and c. oligantha were identified as one group and the rest of the three plants were kept in the other group. a total of 64 constituents of essential oil obtained from whole plant of c. albiflora were identified by gc-ms, where α-pinene (10.87 %), caryophyllene oxide (8.85 %), alcanfor (5.12 %), aromadendrene oxide-(1) (4.81 %), n-hexadecanoic acid (4.74 %), α-famesene (3.93 %), camphene (3.52 %), and isoborneol (3.4 %) were detected as major compounds. the essential oil of c. aromatica possesses 7–methanoazulene (13.75 %) and curcumene (25.71%). caryophyllene (15.07%), phytol (13.38%), humulene (8.24%), elemene (6.11%), caryophyllene oxide (5.82%) were found in c. oligantha. this preliminary study has identified chemical markers present in all curcuma species grown in sri lanka. keywords: curcuma species, essential oils, gc-ms, multivariate. 1 introduction curcuma is an important genus in traditional medicine of sri lanka. commonly known as turmeric, five species are reported in sri lanka (e.g. c. albiflora thw., c. aromatica salisb., c. longa l., c. oligantha trimen., c. h.m.i.c. herath et al. gc-ms analysis of five curcuma species in sri lanka ruhuna journal of science vol 8: 103-111, december 2017 104 zedoaria roscoe). these plant materials are known to be used for centuries to manage various health conditions. people use turmeric in culinary as a flavor or color. in addition, curcuma species show antioxidant, anti-inflammatory, hypocholestraemic, choleratic, antimicrobial, antirheumatic, antifibrotic, antivenomous, antidiabetic, antihepatotoxic, anticancerous, larvicidal, pheromone, insecticidal, anti-plasmodium, hyperprotective, platelet aggregation inhibitory, antiarthritic, cox-1 inhibitory, antiviral and antiproliferative activity etc (afzal et al. 2013; tholkappiyavathi et al. 2013; abbasi and shah 2015; krup et al. 2013; sikha, et al. 2015;). however, many plants are reported under the same sinhalese vernacular name. for example, the people in mahiyanganaya area use both c. aromatica, and c. zedoaria as ‘harankaha’, while the people in erathna area (kegalle district) name c. aromatica as ‘beheth-harankaha’. in literature, zingiber zerumbet, c. albiflora, and c. zedoaria are reported as ‘harankaha’ (dassanayake 1983). regionally people have used various plants under the same vernacular name, even those plants have different chemical compositions and medicinal uses. furthermore, c. albiflora and c. oligantha are unexplored plants. this is the first report of the chemical constituents of these different species of curcuma, and the study was conducted to understand the similarities in terms of phytochemical constituents by analyzing data using multivariate test on five species available in sri lanka, c. albiflora, c. aromatica, c. longa, c. oligantha, and c. zedoaria. therefore, the present study will provide important information on similarities and dissimilarities of curcuma plants grown in sri lanka by chemical analysis. 2 material and methods 2.1 samples and chemical analysis whole plants of curcuma species were collected in the year 2016 from wet and dry zones of sri lanka in the flowering season; c. albiflora (kitulgala, kegalle and erathna, ratnapura district), c. aromatica (erathna, rathnapura), c. longa (kahathuduwa, colombo district), c. oligantha (hebarawa, badulla district), c. zedoaria (gonapola, colombo district). voucher specimens of the plants were authenticated and deposited in the national herbarium, peradeniya, sri lanka, and herbal technology section, industrial technology institute for future reference. phytochemicals in essential oil extracted from clevenger’s apparatus of five curcuma species were analyzed by gc-ms. gc-ms analysis was carried out on a thermoscientific trace 1300 detector and with rtx wax capillary column. mode of h.m.i.c. herath et al. gc-ms analysis of five curcuma species in sri lanka ruhuna journal of science vol 8: 103-111, december 2017 105 operating conditions was split (1:50), and the oven temperature program was 60 o c (after 10.00 min) to 240 o c at 5 o c/min with helium as carrier gas. identification of constituents was done by matching 1700 ev mass spectra, 250 o c quad temperature, 250 o c source temperature, 50 450 (amu) scan parameters, and matching with nist library. the fragmentation pattern and the retention time were compared with the standards to confirm the presence of a particular phytochemical. sampling and analysis were done according to who guidelines (who 2012). 2.2 statistical analysis statistical tests were performed using minitab 17. multivariate test was used to determine the complex relationship among variables (amel 2015). 164 phytochemicals were analyzed by simple correspondence analysis and by cluster variable method. but trace elements were discarded. statistical analysis examined the relationships between the 5 species and the associations between variables in two dimensions, and similar phytochemical contents were identified from their positions as described in greenacre (1983). 3 results and discussion phytochemicals detected in the extracts prepared from curcuma albiflora, c. aromatica, c. longa, c. oligantha, and c. zedoaria are summarized in table 1. table 1: chemical profiles of five curcuma species grown in sri lanka (af: curcuma albiflora, ar: c. aromatica, cl: c. longa, co: c. oligantha, cz: c. zedoaria). compound area% af ar cl co cz alpha ylangene 0.47 0 0 0 0 guaia-1(10),11-diene 1.07 0 0 0 0 myrtenal 0 0 0.1 0 0 o-cymene 0.2 0 0 0 0 p-cymene 0 0 13.3 0 0 thymol 0 0 0.2 0 0 (-)-alcanfor 5.12 0 0 0 0 (-)-myrtenol 0.81 0 0 0 0 (-)-spathulenol 0 0 0 1.97 0 h.m.i.c. herath et al. gc-ms analysis of five curcuma species in sri lanka ruhuna journal of science vol 8: 103-111, december 2017 106 table 1. continued. compound af ar cl co cz (1r)-(+)-nopinon 0.2 0 0 0 0 (e)-bocimene 0 0 0.3 0 0 (e)-nerolidol 0 0 0.5 0 0 (e)-p-famesene 0 0 0.2 0 0 (z)-p-ocimene 0 0 0.1 0 0 1bisabolone 0 1.5 0 0 0 1( 10),4-furanodien-6-one 0 0 0 0 0 2(10)-pinen-3-ol 0.75 0 0 0 0 2-heptanol 0 1 0 0 0 2-octanol 0 0 0 0 0 2-tridecanone 0.15 0 0 0 0 3-carene 0 0 0.9 0 0 6 camphenol 0.12 0 0 0 0 8-tunnerone 0 0 0 0 0 á-copaene 0.85 0 0 0 0 a-curcumene 0 0 0.2 0 0 a-humulene 0 0 0.2 0 0 á-linalool 0.69 0 0 0 0 alphabourbonene 0 0 0 0.06 0 alpha copaene 0 0 0 1.89 0 alpha terpineol 0 0 0.14 0 0 alpha-bisabolene epoxide 0 0 0 0 0 alpha-famesene 4.01 0 0 0 0 á-myrcene 0 0 0 0 0 andrographolide 0 0 0 0.07 0 a-phellandrene 0 0 18.2 0 0 a-pinene 14.5 0.3 2.6 0 0.4 ar-curcurnene 0 3.1 0 0 0 aromadendrene oxide-(1) 4.81 0 0 0.38 0 ar-turmerol 0 0.4 0.2 0 0 ar-turnerone 0 6.3 0.1 0 0 asarone 0 0 0 0.98 0 a-terpinene 0 0 0.4 0 0 a-terpineol 0.64 0.6 0.9 0 0 a-thujene 0 0 0.1 0 0 a-turmerone 0 6.7 0.3 0 0 azulene 0 0 0 0.37 0 bisabolene 0 0 0.2 0 0 b-elernene 0 1.4 0 0 0 b-sesquiphellandrene 0 1.7 0 0 0 carnphene 0 0.7 0 0 0 caryophyllene 2.57 0 0 15.07 0 caryophyllene oxide 9.35 0 0.4 5.82 0 h.m.i.c. herath et al. gc-ms analysis of five curcuma species in sri lanka ruhuna journal of science vol 8: 103-111, december 2017 107 table 1. continued. compound af ar cl co cz cawacrol 0 0 0.1 0 0 ç-elemene 0 0 0 6.11 0 cineole 0 0 14.6 0 0 cinnamal 0 0 0.4 0 0 cis-&elernenone 0 0.2 0 0 0 cis-sabinol 0 0 1 0 0 cis-verbenol 0.28 0 0 0 0 cubedol 0 0 0 0 0 curcuphenol 0 0.2 0 0 0 curdione 2.83 0 0.5 0 0 curzerene 0 0.2 0 0 0 curzerenone 0 11 0 0 0 de hydro-p-cymene 0 0 0.1 0 0 denderalasin 0.6 0 0 0 0 d-mannose 0 0 0 0.08 0 doconexent 0.5 0 0 0 0 elemene 0 0 0 1.86 0 elemicin 0 0 0.2 0 0 endo-borneol 1.44 0 0 0 0 epicurzerenone 0 0 0 0 0 eremophila-1(10),11-diene 0.91 0 0 0 0 eucalyptol 2.21 0 0 0 0 falcarinol 0.13 0 0 0.13 0 gamolenic acid 0 0 0 0.1 0 gerrnacrone 0 0.5 0 0 0 humulene 0 0 0 8.24 0 ionone 0 0 0 0.19 0 isoaromadendrene epoxide 0.69 0 0 0 0 ledene oxide-(ii) 2.11 0 0 0.52 0 limonene 0 1 3.3 0 0 linalool 0 0.2 1.2 0 0 myrcene 0 0 1.8 0 0 myrtanal 0.11 0 0 0 0 myrtenol 0 0 0 0 0 naphthalene 0 0 0 0.78 0 neocurdione 0.87 0 0 0 0 n-hexadecanoic acid 4.74 0 0 0 0 patchoulane 0.16 0 0 0 0 p-bisabolene 0 1.8 0 0 0 p-caryophyllene 0 0.3 0.5 0 0 p-cymen-8-01 0 0 2.4 0 0 phytol 0.99 0 0 13.38 0 pinocarvone 0.46 0 0 0 0 h.m.i.c. herath et al. gc-ms analysis of five curcuma species in sri lanka ruhuna journal of science vol 8: 103-111, december 2017 108 table 1. continued. compound af ar cl co cz ppinene 0 0 7.2 0 0 p-pinene 0 0.4 0 0 0 p-sesquiphellandrene 0 0 0.2 0 0 p-turrnerone( curlone) 0 0.9 0 0 0 myrcene 0 0.2 0 0 0 terpinen-4-01 0.25 0.5 0.8 1.82 0 terpinolene 0 0 11.6 0 0 trans-á-ionone 0 0 0 0.2 0 trans-carvone oxide 0.11 0 0 0 0 trans-geranylacetone 0.15 0 0 0 0 t-rnuurolol 0 0.1 0 0 0 y-terpinene 0 0 1 0 0 zingiberene 0 0.8 0.5 0 0 2-undecanone 0 0 0 0 0.02 verbenone 0 0 0 0 0.03 1-carvon 0 0 0 0 0.04 carvon 0 0 0 0 0.05 pichtosine 0 0 0 0 0.05 1 phelendrine 0 0 0 0 0.07 cyclene 0 0 0 0 0.08 beta myrcene 0 0 0 0 0.15 carbinol 0 0 0 0 0.15 2-nonanone 0 0 0 0 0.27 2-nonanol 0 0.5 0 0 0.29 phenyldihydropyran 0 0 0 0 0.31 linderazulene 0 0 0 0 0.32 terpineol 0 0 0 0 0.33 beta pinene 0 0 0 0 0.41 1-limonene 0 0 0 0 0.54 dl-limonene 0 0 0 0 0.61 sabinene 0 0 0.4 0 0.61 1 ,8-cineole 0 5.5 0 0 1.01 delta elemene 0 0 0 0 1.05 borneol 0 1.1 0.3 0 1.69 camphene 3.64 0 0 0 1.87 beta elemene 0 0 0 0 1.98 germacrene b 0 0.3 0 0 3.98 isoborneol 3.46 3.4 0 0 4.04 beta eudesmene 0 0 0 0 4.97 germacrone 0 0 0.2 0 5.17 benzofuran 0 0 0 0 8.84 aromadendrene 0 0 0 0.19 11.75 camphor 0 32.3 0 0 11.82 debromofiliforminol 0 0 0 0 31.46 h.m.i.c. herath et al. gc-ms analysis of five curcuma species in sri lanka ruhuna journal of science vol 8: 103-111, december 2017 109 among chemicals in c. albiflora, alpha-pinene (14.5%), caryophyllene oxide (9.35%), and alconfor (5.12%) have been found as major compounds. caryophyllene (15.07%), humulene (8.24%), and phytol (13.38) were found as major compounds in c. oligantha. camphor (32.3%), debromofiliforminol (31.46%), and alpha-phellendrene (18.2%) were found as major compounds in c. longa, c. zedoaria, and c. aromatica essential oil. by cluster varibale analysis of phytochemicals present in five speces, mainly two subgroups were clustered; c. albiflora and c. oligantha into one and c. aromatica and c. zedoaria into another. these two subgroups are separated from c. longa (figure 1). fig. 1: cluster variable analysis of dendrogram obtained from phytochemicals present in five curcuma species in sri lanka (af: c. albiflora, ar: c. aromatica, cl: c. longa, co: c. oligantha, cz: c. zedoaria). this is in agreement with the seperation of curcuma species according to morphological characters explained by dassanayake (1983). symmetric row and column plot and column plot from simple correspondence analysis showed comparable results (figure 2). variables s im il a r it y czarclcoaf 48.47 65.65 82.82 100.00 dendrogram with single linkage and correlation coefficient distance h.m.i.c. herath et al. gc-ms analysis of five curcuma species in sri lanka ruhuna journal of science vol 8: 103-111, december 2017 110 fig. 2: column and symmetric plot obtained from phytochemicals present in five curcuma species in sri lanka (af: c. albiflora, ar: c. aromatica, cl: c. longa, co: c. oligantha, cz: c. zedoaria). the present study has identified chemical markers present in all curcuma species grown in sri lanka. however, collection of more samples of same species from different locations is needed to be analyzed for the establishment of chemo-taxonomical identity and distribution pattern by multivariate test. 4 conclusion cluster variable method and simple correspondence analysis enabled the clustering of curcuma species grown in sri lanka into two major groups based on their phytochemical composition. the chemo-taxonomy of curcuma plants grown in sri lanka could be established with the extension of number samples. references abbasi k, shah aa. 2015. biological evaluation of turmeric (curcuma longa). international journal of current microbiology and applied science 4(11): 236-249. h.m.i.c. herath et al. gc-ms analysis of five curcuma species in sri lanka ruhuna journal of science vol 8: 103-111, december 2017 111 afzal a, oriqat g, khan ma, jose j, afzal m. 2013. chemistry and biochemistry of terpenoids from curcuma and related species. journal of biologically active products from nature 3(1): 1-55. amel b. 2015. microscopic analysis of curcuma longa l. using multivariate test. international journal of pharmacognosy 2(4): 173-177. dassanayake md. 1983. flora of ceylon, vol. iv. amerind publishing co. pvt. ltd, new delhi 495 – 496, 501-504 pp. greenacre mj. 1983. correspondence analysis in practice. academic press, harcourt, brace & company. krup v, prakash hl, harini a. 2013. pharmacological activities of turmeric (curcuma longa linn): a review. journal of traditional medicine & clinical naturopathy 2(133): 21671206. sikha a, harini a, prakash h. 2015. pharmacological activities of wild turmeric (curcuma aromatica salisb): a review. journal of pharmacognosy and phytochemistry 3(5): 01-04. tholkappiyavathi k, selvan k, neyanila m, yoganandam sk, gopal gp. 2013. a concise review on curcuma zedoaria. international journal of phytotherapy 3(1): 1-4. who. 2012. quality control methods for medicinal plant materials. world health organization, genève, 33-40 pp. rjs-vol-1-sept-2006-11.dvi ruhuna journal of science vol. 1, september 2006, pp. 104–112 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. evaluation of antibacterial activity of different mangrove plant extracts p.d. abeysinghe, r.p. wanigatunge department of botany, university of ruhuna, matara, sri lanka , pushpa@bot.ruh.ac.lk, r.n. pathirana department of chemistry university of ruhuna, matara, sri lanka , ranjithnp@chem.ruh.ac.lk antibacterial activity of mature leaves, tender leaves and bark extracts of avicennia marina, avicennia officinalis and bruguiera sexangula were evaluated using soxhelt extraction method. petroleum ether, chloroform, ethyl acetate and ethanol were used as solvents in order to get the plant extracts. the antibacterial activity was screened by using agar diffusion technique against pathogenic bacteria species of staphylococcus sp. (from urine), proteus sp. (from a wound), escherichia coli (from infected blood), shigella sp. (from a wound) and pseudomonas sp. (from a wound). the length of inhibition zone was measured in millimeters from the edge of the well to the edge of the inhibition zone. twelve different plant extracts in a. marina, a. officinalis and b. sexangula exhibited different degree of growth inhibition against tested bacterial strains. mature leaf extracts of a. marina and tender leaf extracts of a. officinalis in ethyl acetate exhibited promising antibacterial activity than other plant extracts. all plant extracts in ethyl acetate showed strong inhibition compared to other extracts on all tested bacterial strains. among all bacterial strains, pseudomonas sp. and e. coli showed considerable growth inhibition against almost all plant extracts. mature leaf extract of only a. marina was used for further investigations since a large amount of tender leaves of a. officinalis was not readily available. charcoal treated mature leaf extracts of a. marina showed more inhibition for all tested bacterial strains than untreated plant extracts. combinations of mature leaf extracts of a. marina in different solvents failed to exhibit synergistic activity against tested bacterial strains. most combinations showed antagonistic effect against proteus sp., e. coli and shigella sp. antibacterial activity of plant extracts of a. marina gradually declined during 4 months after extraction, for all tested bacterial strains. components of mature leaf extracts of a. marina in chloroform, ethyl acetate and ethanol were separated by analytical thin layer chromatography (atlc) when the eluents were hexane and diethyl ether. isolated components in preparative thin layer chromatography (ptlc) exhibited moderate antibacterial activity against pseudomonas sp., shigella sp. and e. coli. separated components did not show any antibacterial activity against proteus sp. and staphylococcus sp. phytochemical screening revealed that mature leaf of a. marina contained alkaloids, steroids, triterpenoids and flavonoids. key words : a. marina, antibacterial activity, plant extracts, growth inhibition 1. introduction microorganisms have potential to cause human diseases. most of the time viruses, bacteria and fungi act as major pathogenic organisms. the discovery of antibiotics in the early twentieth centaury provided an increasingly important tool to combat bacterial diseases. as antibiotics are increasingly used and misused, the bacterial 104 abeysinghe, wanigatunge and pathirana: evaluation of antibacterial... ruhuna journal of science 1, pp. 104–112, (2006) 105 strains become resistant to antibiotics rapidly. therefore, screening of antibacterial activity of medicinal plants is very important since vast number of medicinal plants have been used for centuries as remedies for human diseases. among them extracts from different parts of mangroves and mangrove associates are widely used throughout the world. for instance, stem of avicennia marina is used for ulcers and bark of bruguiera sexangula is used for antitumors (bandaranayake 1998). mangrove and mangrove associates contain biologically active antiviral, antibacterial and antifungal compounds (bandaranayake 1998). they provide a rich source of steroids, triterpenes, saponins, flavonoids, alkaloids and tannins (bandaranayake 1995). therefore, it is worth to screen mangrove plants for the presence of new antibacterial compounds to combat the normal pathogenic bacterial strains and hospital acquired antibiotic resistant bacterial strains. the main objectives of this study were to screen antibacterial activity of some selected mangrove plant species against pathogenic and antibiotic resistant bacterial strains and also to isolate and characterization of chemical components which are responsible for antibacterial activity. 2. materials and methods 2.1. plant materials and preparation of plant extracts mature leaves, tender leaves and bark of avicennia marina, avicennia officinalis and bruguiera sexangula were used as plant materials. sequential soxhlet extraction method was carried out to obtain plant extracts. fifty grams of mature leaves of each selected mangrove species were ground and extracted separately with petroleum ether (300 ml) followed by chloroform (300 ml), ethyl acetate (300 ml) and finally with ethanol (300 ml). the extraction time period was three hours for each solvent. the same procedure was followed for tender leaves and bark of tested plants. extracts were concentrated to 3 ml portions using a rotary evaporator at 50◦c and stored at 4◦c . four mature leaf extracts of a. marina obtained from soxhlet extraction method were treated with appropriate amount of activated charcoal and incubated at 40◦c for 10 minutes in a water bath. charcoal treated extracts were centrifuged at 10,000 rpm for 2 minutes and the organic layer was obtained. mature leaf extracts of a. marina were mixed with the following combinations of solvents: petroleum ether and chloroform, petroleum ether and ethyl acetate, petroleum ether and ethanol were used to test synergistic activity of plant extracts. plant extracts for phytochemical screening were prepared using 100g of ground fresh mature leaves of a. marina. they were extracted using 300ml of 95% (v/v) ethanol in soxhlet apparatus. the mixture was refluxed on a steam bath about one hour, then cooled to room temperature and plant extract was washed with 50 ml of fresh 95 % (v/v) ethanol. the final volume of the extract was measured and portions of this extracts were used in phytochemical screening for alkaloids, flavonoids, saponins, terpenoids, steroids and cardiac glycosides. abeysinghe, wanigatunge and pathirana: evaluation of antibacterial... 106 ruhuna journal of science 1, pp. 104–112, (2006) table 1 tested bacterial strains for antibacterial activity. bacterial strain source escherichia coli (from blood) general hospital, (matara) proteus sp.* (from a wound) general hospital, (matara) pseudomonas sp. (from a wound) general hospital, (matara) shigella sp. (from a wound) faculty of medicine, (university of ruhuna) staphylococcus sp.* (from urine) general hospital, (matara) *antibiotic resistant bacterial strains. 2.2. tested bacterial strains staphylococcus sp. is resistant to many common antibiotics in use such as ceftazidime, cephalexin, cotrimoxazole, cloxacillin, gentamycin, kenamycin and ticarcillin/clavulanic acid although proteus sp. is resistant only to gentamycin and kenamycin as indicated in the microbiology report of the general hospital, matara 2003. 2.3. determination of antibacterial activity antibacterial activity of tested bacterial strains was assayed using all plant extracts by agar diffusion technique (de castillo et al. 1998). antibacterial activity was observed as inhibition zone on petri plates containing nutrient agar. size of the inhibition zone was measured in millimeters using a metric ruler from the edge of the well to the edge of the inhibition zone. petroleum ether, chloroform, ethyl acetate and ethanol were used as controls instead of plant extracts. stability of antibacterial activity of plant extracts was tested during a period of 4 months and results were obtained once a month using agar diffusion technique. antibacterial activity against tested bacterial strains (table 1) was tested for components separated by ptlc. 2.4. separation of active components mature leaf extract obtained from soxhlet extraction was separated using thin layer chromatography (tlc). pure solvents of hexane, diethyl ether, petroleum ether, ethyl acetate and water, a mixture of hexane : diethyl ether; petroleum ether: ethyl acetate; ethyl acetate: water; hexane: water were used as mobile phase with different ratios as 9 : 1 , 8 : 2 , 7 : 3 , 6 : 4 , 5 : 5 , 4 : 6 , 3 : 7 , 2 : 8 and 1 : 9. plates were observed under uv light and developed using i2 vapour. 2.5. phytochemical screening phytochemical screening was carried out for crude extract of mature leaves of a. marina for alkaloids, flavonoids, saponins, terpenoids, steroids and cardiac glycosides. mayer’s reagent and wagner’s reagent were used to detect alkaloids. dragendorff’s reagent was used to estimate the probable number of alkaloids present. mayer’s reagent was also used to determine quaternary alkaloids. saponins were screened using froth test. liebermann burchardt test and salkoweski test were carried out to determine steroids and triterpenoids. kellerkilliani test was performed for screening cardiac glycosides. flavonoids were tested using octanol (harbone 1984). abeysinghe, wanigatunge and pathirana: evaluation of antibacterial... ruhuna journal of science 1, pp. 104–112, (2006) 107 figure 1 inhibition of e. coli by plant extracts of mature leaf of a. marina, a. officinalis and b. sexangula extracted in petroleum ether (pe), chloroform (c), ethyl acetate (ea), and ethanol (et) using soxhlet extraction method. 3. results twelve different extracts from a. marina, a. officinalis and b. sexangula exhibited different degree of growth inhibition against tested bacterial strains (table 2). according to table 2, mature leaves of a. marina and tender leaves of a. officinalis exhibited considerable antibacterial activity against tested bacterial strains. further tests were carried out using only mature leaves of a. marina because tender leaves of a. officinalis are not easily obtainable. plant extracts of mature leaf of a. marina showed more inhibition than mature leaf extracts of a. officinalis and b. sexangula against e. coli, staphylococcus sp., and shigella sp. (fig. 1 and table 2). tender leaf extracts of a. officinalis exhibited more pronounced inhibition against staphylococcus sp. and pseudomonas sp. (fig. 2 and table 2). bark extracts of a. marina showed strong inhibition against shigella sp. and e. coli (fig. 3 and table 2) compared to the extracts of b. sexangula and a. officinalis. mature leaf extract of a. officinalis was unable to inhibit the growth of shigella sp. and staphylococcus sp. while tender leaf extract did not exhibit antibacterial activity against proteus sp. moreover, none of the tender leaf extracts of b. sexangula were able to inhibit proteus sp., pseudomonas sp. and shigella sp. also mature abeysinghe, wanigatunge and pathirana: evaluation of antibacterial... 108 ruhuna journal of science 1, pp. 104–112, (2006) figure 2 inhibition of staphylococcus sp. and pseudomonas sp. by plant extracts of tender leaf of a. officinalis extracted in petroleum ether (pe), chloroform (c), ethyl acetate (ea) and ethanol (et) using soxhlet extraction method. figure 3 inhibition of e. coli and shigella sp. by plant extracts of bark of a .marina extracted in petroleum ether (pe), chloroform (c), ethyl acetate (ea) and ethanol (et) using soxhlet extraction method. leaf extract of b. sexangula did not exhibit antibacterial activity against pseudomonas sp. and bark extracts were unable to inhibit staphylococcus sp (table 2). no inhibition of the growth of proteus sp., staphylococcus sp. and shigella sp. was observed in extracts in petroleum ether except bark extract of a. marina. ethyl acetate extracts of bark of a. officinalis, mature leaf and tender leaf extract of a. marina and chloroform extract of mature leaf of a. marina were able to inhibit the growth of all tested bacterial strains. ethanolic extract of bark of a. officinalis greatly inhibited, the growth of pseudomonas sp, e. coli and shigella sp. controls did not exhibit inhibitory effect against any of the tested bacterial strains. charcoal treated mature leaf extracts of a. marina were able to inhibit the growth of all tested bacterial strains more than untreated extracts (table 3). none of the tested possible combinations of mature leaf extracts were able to exhibit synergistic activity abeysinghe, wanigatunge and pathirana: evaluation of antibacterial... ruhuna journal of science 1, pp. 104–112, (2006) 109 figure 4 a diagrammatic representation of ptlc plates showing separated bands in plant extracts of mature leaves in chloroform, ethanol and ethyl acetate when the eluents were hexane: diethyl ether 4: 6. table 2 a comparison of the degree of growth inhibition of bacterial species in millimeters (mm) from the edge of the well to the edge of the inhibition zone by charcoal treated (t) and untreated (ut) plant extracts of mature leaves of a. marina in petroleum ether (pe), chloroform (c), ethyl acetate (ea) and ethanol (et). plant material proteus sp. staphylococcs sp. pseudomonas sp. e. coli shigella sp. ut t ut t ut t ut t ut t m a tu re le a v e s pe 2 3 1 2 c 1 3 1 1 1 3 2 2 2 3 ea 4 5 4 7 4 6 6 8 7 9 et 2 4 5 7 5 8 ( mark indicates no inhibition) against tested bacterial strains. most combinations exhibited antagonistic effect against proteus sp., e. coli and shigella sp. (table 4). the degree of antibacterial activity of mature leaf extract of a. marina decreased with time for all tested bacterial strains. when tlc was carried out, chloroform and ethanol extracts of mature leaves separated into two bands and ethyl acetate extract of mature leaves separated into three bands (fig. 4). separated components of tlc could inhibit the growth of tested bacterial species except proteus sp. and staphylococcus sp. (table 5). according to phytochemical screening, mature leaves of a. marina contained alkaloids, steroids and flavonoids. 4. discussion a. marina, a. officinalis and b. sexangula were used as test plants due to the presence of much evidence that prove their therapeutic value against microbial infections (bandaranayake 1998). also preliminary studies have been demonstrated that the mangrove plant extracts have antibacterial activity against pathogenic bacterial strains; staphylococcus sp. and e. coli and pseudomonas sp. (abeysinghe et al. 2002) and antibiotic resistant bacterial strains; staphylococcus sp and proteus sp. (abeysinghe et al. 2003). the results of the present study clearly showed that mangrove plant extracts showed antibacterial activity against tested pathogenic bacterial strains including antibiotic resistant strains. the effectiveness of the active abeysinghe, wanigatunge and pathirana: evaluation of antibacterial... 110 ruhuna journal of science 1, pp. 104–112, (2006) table 3 degree of growth inhibition of bacterial species measured in millimeters (mm) from the edge of the well to the edge of the inhibition zone by possible combinations of mature leaf extracts of a. marina in petroleum ether (pe), chloroform, ethyl acetate (ea) and ethanol (et) bacterial strains combination of plant extracts p e c e a e t p e + c p e + e a p e + e t c + e a c + e t e a + e t p e + c + e a p e + c + e t e t + c + e a e a + p e + e t p e + c + e t + e a proteus sp. 1 3 2 1 2 1 2 1 3 2 2 4 pseudomonas sp. 1 1 7 4 staphylococcus sp. 1 5 1 shigella sp. 2 8 4 1 2 4 9 4 6 4 2 4 5 7 e. coli 1 1 8 6 2 5 7 6 3 1 4 3 4 ( indicates no inhibition). table 4 level of inhibition caused by the components of mature leaf extracts of a. marina in chloroform (mc), in ethanol ((me) and in ethyl acetate (mea) separated by ptlc measured in millimeters (mm) from the edge of well to the edge of inhibition zone against proteus sp. (pr), staphylococcus sp. (st), pseudomonas sp. (ps), e. coli and shigella sp. (sh) by tlc method,b1band one, b2band two, b3-band three ) bacterial strain isolated bands mcb 1 mcb 2 meb 1 meb 2 meab 1 meab 2 meab 3 pr ps 2 1 1 st e.coli 2 1 1 1 2 sh 1 1 2 ( -mark indicates no inhibition). compounds present in plant extracts cause the production of growth inhibition zones that appear as clear areas surrounding the wells. antibacterial activity may be due to active components which are present in plant extracts. however, some plant extracts were unable to exhibit antibacterial activity against tested bacterial strains. these bacterial strains may have some kind of resistance mechanisms e.g. enzymatic inactivation, target sites modification and decrease intracellular drug accumulation (schwarz and noble 1999) or the concentration of the compound used may not be sufficient. no inhibition was observed with controls, which proves that solvents could not act as antibacterial agents. further investigations were done using extracts of mature leaves of a. marina only, since they showed considerable antibacterial activity. in almost all tests, crude ethyl acetate extracts showed better inhibition against all tested bacterial strains, indicating that active ingredients in plant materials could be extracted into ethyl acetate. however, highest antibacterial activity was observed against e. coli. abeysinghe, wanigatunge and pathirana: evaluation of antibacterial... ruhuna journal of science 1, pp. 104–112, (2006) 111 charcoal treated plant extracts showed more inhibition against all tested bacterial strains than untreated plant extracts. it may be assumed that plant pigments may increase the survival ability of bacteria. the degree of antibacterial activity of plant extracts decreased during the period of four months after extraction. it may be due to degradation or volatility of antibacterial compounds or they may have converted into non-antibacterial compounds. a combination of plant extracts in different solvents unable to exhibit synergistic activity against tested bacterial strains. mixing of plant extracts in different solvents seems to have diluted down the toxic effect against tested bacterial stains or might have increased the survival ability of pathogenic bacteria. tlc was carried out using different solvent systems. but only hexane and diethyl ether solvent system gave best separation. separated components showed positive results against pseudomonas sp., e. coli and shigella sp. whereas antibacterial activity was not observed against staphylococcus sp and proteus sp. according to preliminary studies, it has been reported that active components from mature leaves of a. marina have been separated using tlc (withanawasm 2002). phytochemical screening of mature leaf extracts of a. marina revealed the presence of biologically active substances such as alkaloids, steroids, triterpenoids and flavonoids. the results obtained from preliminary phytochemical screening are comparable with the results reported earlier (bandaranayake, 1995). preliminary phytochemical screening of luminetzera racemosa reveled the presence of secondary metabolites such as alkaloids, flavonoids and steroids (abeysinghe et al., 2003). these secondary metabolites may exert antibacterial activity against tested bacterial strains. in addition to above tested phytochemical groups tannins, anthocyanins, polyphenols, coumarins and essential oils should be subjected to phytochemical screening. it is promising that the tested mangrove plant species could be used to synthesis novel antibiotics for bacterial infections, especially for antibiotic resistant bacterial infections. further research is necessary for successful separation, purification and characterization of biologically active compounds using chromatographic methods and spectroscopic techniques. further studies are being carried out in order to separate the individual components that are present in plant extracts of a. marina using column chromatography. references abeysinghe pd, withanawasam m, pathirana rn, abeysinghe s. 2002. preliminary in vitro screening of antibacterial compounds of some mangrove plant extracts for clinical isolates from different sources. proceeding of the first science symposium, university of ruhuna, 22-25 pp. abeysinghe pd, wijesekara d, pathirana rn. 2003. inhibition of growth of antibiotic resistant staphylococcus sp. and proteus sp. by mangrove plant extracts. proceeding of the first science symposium, university of ruhuna, 1-9 pp. bandaranayake wm. 1995. survey of mangrove plants from northern australia for phytochemical constituents and uv-absorbing compounds. current topics in phytochemistry (life science advances) 14:69-78 abeysinghe, wanigatunge and pathirana: evaluation of antibacterial... 112 ruhuna journal of science 1, pp. 104–112, (2006) bandaranayake wm. 1998. traditional and medicinal uses of mangroves. mangroves and salt marshes. 2:133-148. de castillo mc, de nader om, de r. holgado ap. 1998. in vitro comparison of disk diffusion and agar dilution antibiotic susceptibility test methods for neisseria gonorrhoeae. 93(4):517. harbone jb. 1984. phytochemical methods. 2nd edition. chapman and hall, london. schwarz s, noble wc. 1999. aspects of bacterial resistance to antimicrobials used in veterinary dermatological practice. 163-176 pp. withanawasam dm. 2002. preliminary in vitro screening of antibacterial and antifungal compounds of mangrove plant extracts for pathogens from different sources. acknowledgments this work is partially funded by the twas (third world academy of sciences) award to the first author. microsoft word rjs-vol-ii-saman-abeysinghe-1.2.doc © 2007 faculty of science university of ruhuna ruhuna journal of science vol. 2, september 2007, pp. 82-88 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x 82 abstract. root rot, caused by fusarium solani f. sp. phaseoli, is one of the main root diseases impacting production of common bean in sri lanka. rhizobacteria were screened in dual petri plate assays to select antagonistic strains against f. solani f. sp. phaseoli. b. subtilis ca32 effectively antagonized the pathogen. t. harzianum ru01 also showed the antagonistic activity. the efficacy of the b. subtilis ca32 and the t. harzianum ru01 were tested in greenhouse pot experiments against f. solani f. sp. phaseoli. seed bacterization with b. subtilis ca32 and t. harzianum ru01 significantly protected bean seedlings from f. solani f. sp. phaseoli compared to the untreated control plants. plant protection was more pronounced in t. harzianum ru01 treated plants than bacterized plants. enhanced root growth was observed only t. harzianum ru01 treated plants, suggesting that the biotic modifications of the mycorrhizosphere as a result of colonization with t. harzianum ru01. keywords: rhizobacteria, biological control, bean, antagonism 1 introduction root rot of common bean (phaseolus vulgaris l.) is a soil-borne disease that is incited by several fungal pathogens including fusarium spp., pythium spp. and, rhizoctonia solani. it occurs in all bean-growing areas of the world. root rot caused by f. solani f. sp. phaseoli is a major concern in many bean growing areas in sri lanka leading to enormous crop losses. the pathogen is known to be very persistent in soil and capable of surviving in infested fields for very long period and is difficult to control. fusarium root rot is characterized by reddish-brown lesions along the tap roots and lower hypocotyls. diseased areas of the plant enlarge with age and gradually turn brown. longitudinal cracks may develop in older lesions and the cortical tissues be discolored and decayed. root rots are particularly severe under water-stress condition (burke and hall, 1991). disease management options include crop rotation, improving soil fertility levels, use of resistant cultivars, use of fungicides and biological control. the impetus for developing biological control agents has been the public perception of pesticide toxicity in the environment. residue-free produce has become a valuable commodity. biological control is biological control of fusarium solani f. sp. phaseoli the causal agent of root rot of bean using bacillus subtilis ca32 and trichoderma harzianum ru01 saman abeysinghe department of botany, faculty of science, university of ruhuna, matara, sri lanka correspondence: saman@bot.ruh.ac.lk abeysinghe: biological control of fusarium solani. . . 8 3 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 82-88 ( 2 0 0 7 ) compatible with pesticide-free agriculture and the environment. roots of plants support the growth of a complex of microorganisms that can have a profound effect on the growth and survival of the plant. among these organisms, arbuscular mycorrhizal fungi (amf) are known to improve the nutritional status of their host, provide alterations in the host’s physiology and exudation from roots (filton et al., 2006). there is accumulating evidence that amf can reduce disease incidence and propagule number of several soil-borne pathogens including fusarium (filton et al., 2006). apart from the amf, other microbial biological control agents such as rhizospheric bacteria (handelsman & stabb, 1996) and certain saprophytic fungi (steyaert et al., 2003) have been identified as promising candidates to combat with various deleterious soil-borne pathogens on different crops. certain biological control agents protect plants under field condition and commercial formulations of these organisms are now available (spadaro & gullino, 2005). however, control of f. solani f. sp. phaseoli by using bacterial biological control agents and trichoderma spp. do not exist. at present, there is very limited knowledge and experience regarding the biological control of soil-borne diseases in sri lanka. therefore, the main objective of the current study was to screen rhizospheric bacteria and test trichoderma isolate ru01 for in vitro antagonism against f. solani f. sp. phaseoli and evaluate their efficacy under greenhouse conditions in order to select potential biological control agents which could be used in the field. 2 materials & methods biological control agents used in this study bacterial cultures bacteria used in this study were isolated from capsicum annuum l. rhizosphere as described in abeysinghe, (2007). bacterial isolates were maintained in 80% glycerol (v/v) at -80 0c. in order to culture these bacteria, a loopful of inoculum was streaked on caa plates (bactocassamino acid 5g (difco, detroit, mi), k2hpo4 0.18 g, mgso4.7h2o 0.25 g, 18 g of bactoagar, water 1 l). after incubation for 24 h at room temperature, single colonies were streaked on fresh caa plates. trichoderma harzianum ru01 trichoderma spp. was isolated from soil samples obtained from a commercial chili field at angunukolapalasse by using dilution petri plate technique on trichoderma selective medium (askew and laing, 1993) and maintained on pda. identification was performed by using colony characteristics described in gams & bissett (1998). t. harzianum ru01 was selected for further studies because of its antagonistic activity against different fungal pathogens. screening of antagonistic bacteria against f. solani in petri plate assay all bacterial isolates were initially screened for the ability to inhibit fungal growth on caa plates. single colonies were selected and patched along the perimeters of plates and incubated overnight at 28 oc. the following day each plate was inoculated at the centre with 84 abeysinghe: biological control of fusarium solani. . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 82-88 ( 2 0 0 7 ) a 5 mm diameter plug of the growing fungus. the plates were incubated at 28oc for 2 -3 days and observed for inhibition of fungal growth. the bacteria positive for antagonism were selected. zones of inhibition indicated antifungal activity and the strains were ranked according to the inhibition zone. an inhibition zone was defined as the distance between the leading edge of the fungal growth and the closest edge of the bacteria. inhibition was expressed relative to a control strain spotted on the same plate. interaction between t. harzianum and f. solani in petri plate assay five millimeter-diameter discs of t. harzianum isolate were removed from the edge of colonies of 4-day old pda cultures and placed on one side of a petri dishes containing pda medium. similar dishes of f. solani isolate grown in the same manner were placed on the opposite side of petri plates and made three replicates. cultures were observed daily and recorded for antagonism or parasitism of trichoderma isolate against f. solani. isolation of pathogen phaseolus vulgaris plants showing root rot were collected from commercial bean fields at balangoda, in the sabaragamuwa province of sri lanka. infected parts of the plants were excised with a sterile scalpel and, surface sterilized with 3% (w/w) naocl for 2 min. sterilized pieces were washed twice with sterile water for 60s and, cut into small pieces (1 cm length) and transferred on to antibiotic amended pda plates. plates were incubated at room temperature for 48 h, and white mycelium growth from the infected stem pieces were transferred to new pda plates. after incubation for 5 days, a single spore was isolated and cultured on new pda plates. the pathogen was identified as fusarium solani f. sp. phaseoli, based on the characteristics described by booth (1977). koch’s postulates were demonstrated for the pathogen and confirmed as the causal agent of root rot of p. vulgaris. bacterial innoculum preparation for seed bacterization among the bacterial isolates the most antagonistic bacterial isolate, according to the in vitro petri plate assay, bacillus subtilis ca32 was selected for greenhouse pot experiments. fresh cells were obtained from stock cultures stored at – 80 oc and grown in caa broth overnight at room temperature in a shaker. from this 100 ml of caa broth in a 250 ml flask was inoculated and incubated for 48 h at room temperature in a rotary shaker (100 round/min). the bacterial culture was centrifuged (6000 g for 10 min at 4 oc) and the supernatant was discarded. the cell pellet was resuspended in sterile 0.85% nacl and centrifuged again under the same conditions. the supernatant was discarded and washed bacterial cells were resuspended in sterile distilled water (sdw). the concentration of cells in the suspension was spectrophotometricaly adjusted to 108 cfu/ml and used for seed bacterization. inoculum production of trichoderma harzianum for seed inoculation, trichoderma was grown on pda in petri plates for seven days at room temperature under dark, to allow profuse sporulation. sterile distilled water was added to each plate and a conidial suspension was obtained by scraping the colony surface with a abeysinghe: biological control of fusarium solani. . . 8 5 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 82-88 ( 2 0 0 7 ) sterile spatula and filter through cheese cloth. conidial suspension was adjusted to 106 spores/ml and mixed (0.01%) tween-20 as a wetting agent. f. solani inoculum and experimental set-up the pathogen inoculum was produced on potato dextrose agar (pda). the plates were inoculated with an agar plug (5 mm in diameter) containing actively growing f. solani mycelium, and incubated under fluorescence for 10 days at room temperature. spores were washed from the plates with sterile distilled water and the concentration was adjusted to 106 conidia/ml with a haemocytometer. ten days after planting when the primary leaves were fully expanded, the best seedlings were selected by thinning to three plants per pot. five milliliters of spore suspension was applied by pipette just below the collar region around the hypocotyls of each plant. disease severity was assessed 21 days after inoculation. seedlings were removed from the pots and excess soil clumps were removed by gently shaking and by dipping the roots in water. roots were dried with a paper towel and rated immediately for symptoms of root rot. the severity of root rot was visually scored by assessing necrotic lesions on the roots and hypocotyls using a rating scale of 0-5 described according to filion et al., (2003). treatments consisting of 3 seedlings/pot and 5 replicates were included per treatment. the trial was conducted twice and the experiment was arranged in a completely randomized block design. 3 results pathogen identification and characteristics colonies grown on pda or caa became brown 7 day after incubation, and produced macro and micro conidia. the conidia were approximately 0.4 mm diameter. the pathogenicity tests for the f. solani isolate was examined with different local bean cultivars. based on the koch’s postulates fusarium isolate was pathogenic to all tested local bean cultivars to different levels (data not shown). identification and characterization of trichoderma spp. trichoderma spp. grew rapidly at room temperature on pda. cultures at first were white and cottony then turned to bright green, finally they became dark green. chlamydospores were intercalary and/or terminal, globose and smooth walled. single phialides arose laterally on the conidiophores in clusters. therefore trichoderma isolate was identified as t. harzianum and named as ru01. in vitro antagonism tests t. harzianum ru01 against f. solani antagonistic properties of t. harzianum ru01 was tested using dual petri plate method. trichoderma inhibited the mycelial growth of f. solani but could not overgrow the pathogen until 3 to 4 days. however, several days later f. solani over grew the trichoderma mycelia. furthermore, conidia production decreased compared to the control plates (data not 86 abeysinghe: biological control of fusarium solani. . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 82-88 ( 2 0 0 7 ) shown). t. harzianum hyphae coiled around hyphae of f. solani causing vacuolization and disintegration of f. solani hyphae indicating strong antagonistic activity of the t. harzianum isolate ru01. screening of rhizobacteria against f. solani from the preliminary round of screening for antagonism in petri plate assay, only four bacterial isolates were selected for further study. among these isolates, b. subtilis ca32 was selected for greenhouse pot experiments because ca32 was the most antagonistic isolate in terms of inhibition zones in the plate assay (table i). table 1: mycelial inhibition of f. solani f. sp. phaseoli in caa by bacterial isolates isolate mycelial inhibition mean (%)x bacillus subtilis ca32 55.50 a pseudomonas fluorescens ca05 32.00 b pseudomonas putida ca28 30.32 b bacillus subtilis ca16 28.70 b control 0 c x the percent mycelial inhibition data were analyzed using analysis of variance (anova) and a mean separation was done by fisher’s least significant difference at p < 0.05. three replication plates were included per bacterial isolates and the data were pooled from two separate experiments. mycelial inhibition means with the same letter are not significantly different from each other (95% confident limit). plant protection ability of biological control agents under greenhouse conditions all bean seedlings grown in soil inoculated with a conidial suspension of f. solani f. sp. phaseoli showed red lesions on hypocotyls and tap roots characteristically distinctive of fusarium root rot. in contrast, all plants from seeds bacterized with ca32 or treated with t. harzianum ru01 conidia showed significantly less number of lesions and low disease severity. the noninfested controls showed no symptoms. however, t. harzianum treated plants were well protected from the pathogen infection than ca32 treated plants (table 2). table 2. effect of application of b. subtilis ca32 and t. harzianum to the seeds of p. vulgaris on infection of f. solani f. sp. phaseoli in pot experiments disease severity x treatment trial i y trial ii trial iii healthy control 0 a 0 a 0 a disease control 4.2 b 4.5 b 3.8 b b. subtilis ca32 2.4 c 2.2 c 2.6 c t. harzianum 1.6 c 1.8 c 2.0 c x disease severity was assessed using a rating scale of 0-5, where 0 = no disease symptoms, 1 = slightly brown <50% surface discoloration of the hypocotyl, 2 = >50% surface discoloration, 3 = abeysinghe: biological control of fusarium solani. . . 8 7 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 82-88 ( 2 0 0 7 ) discolored hypocotyls and extensive root pruning, 4 = darkly discolored hypocotyls and root completely collapsed and severe root pruning, and 5 = dead or dying plant. y values followed by a different letter within columns are significantly different according to the lsd test at p<0.05 using duncan’s multiple range test. the length and fresh weight of root was significantly increased (p < 0.05) in plants colonized by t. harizanum than ca32 treated plants as compared with nontreated control seedlings (table 3). table 3. effect of biological control agents on p. vulgaris root development growth parameters treatment root length (cm) root fresh weight (g) healthy control 5.2±1.25 ax 9.25±1.25 a disease control 3.2±1.87 c 5.28±0.98 c b. subtilis ca32 5.6±1.92 a 10.25±1.75 a t. harzianum 9.8±2.12 b 15.35±2.02 b x values are the means of three replicates ± standard error of the mean. values followed by a different letter within columns are significantly different using duncan’s multiple range test (p<0.05). 4 discussion although the interactions between many bacteria and fungi have been studied those involving fusarium solani f. sp. phaseoli have received less attention. this study presents the data of screening of bacteria isolated from c. annuum rhizosphere and trichoderma harzianum against fusarium solani f. sp. phaseoli, one of the major causal agents of root rot of bean in sri lanka. among 53 bacterial isolates only 4 isolates were able to antagonize f. solani in dual petri plate assay. b. subtilis ca32 was the strongest antagonistic isolate (table 1). more over, trichoderma harzianum ru01 also inhibits the growth of f. solani and reduces the production of conidia indicating antagonistic properties of t. harzianum ru01 against the pathogen. however, spore production inhibition did not quantify. the in vitro culture of f. solani and t. harzianum ru01 in culture media led to a variety of interactions. f. solani growth was generally inhibited; the host cell contents disorganized and the hyphae were intensively parasitized by t. harzianum ru01. similar reactions have been reported on other fungal pathogens (hanson & howell, 2004) but no reports on f. solani. the results reported here suggest that the presence of either b. subtilis ca32 or t. harzianum ru01 in the rhizosphere significantly reduces the root rot caused by f. solani in bean plants. the reduction might be related to the decline of the population density of f. solani in soil and also due to alterations caused by t. harzianum ru01 in the f. solani hyphae as observed in vitro. however, the protection exerted by the t. harzianum ru01 against f. solani was pronounced than b. subtilis ca32. this difference may be due to the more than one mode of mechanisms exerted by the t. harzianum ru01 which may have an 88 abeysinghe: biological control of fusarium solani. . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 82-88 ( 2 0 0 7 ) additive effect in plant protection. most likely the enhanced growth of root system by t. harzianum as evidenced by increased biomass may be positively acted in this respect. more over t. harzianum is a well-known producer of cell wall-degrading enzymes and the antibiotics thus could act synergistically with other mechanisms (vinale et al., 2006). in conclusion, the present study clearly demonstrated that antagonistic bacterial strain b. subtilis ca32 and t. harzianum ru01 can be used as biological control agents in order to protect bean plants from f. solani f. sp. phaseoli under greenhouse conditions. the combine use of these biocontrol agents and the evaluation of the biological control efficacy under field conditions are underway. references abeysinghe, s. the effect of mode of application of bacillus subtilis ca32r on control of sclerotium rolfsii on capsicum annuum. archives of pytopathology & plant protection (in press). askew, d. g., laing, m. d. (1993) an adapted selective medium for the quantitative isolation of trichoderma species. plant pathology 42:686-690. booth c. 1977. fusarium: laboratory guide to the identification of the major species. kew, uk: commonwealth mycological institute. burke, d. w., and hall, r. compendium of bean diseases. st. paul, minnesota, usa, aps press, 1991, 9-10pp. filion, m., st-arnaud, m., jabaji-hare, s. h. (2003) quantification of fusarium solani f. sp. phaseoli in mycorrhizal bean plants and surrounding mycorrhizosphere soil using realtime polymerase chain reaction and direct isolations on selective media. phytopathology 93:229-235. gams, w., bisset, j. (1998) morphology and identification of trichoderma. in: kubicek, c. p., herman, c. e. (eds.), trichoderma and gliocladium: basic biology, taxonomy and genetics. taylor and francis, london, uk, pp. 3-34. handelsman, j. & stabb, e. c. (1996) biocontrol of soilborne plant pathogens. the plant cell 8:1855-1869. hanson, l. e. & howell, c. r. (2004) elicitors of plant defense responses from biocontrol strains of trichoderma virens. phytopathology 94:171-176. spadaro, d., & gullino, m. l. (2005) improving the efficacy of biocontrol agents against soilborne pathogens. crop protection 24:601-613. steyaert, j. m., ridgway, h. j., elad, y., stewart, a. (2003) genetic basis of mycoparasitism: a mechanism of biological control by species of trichoderma. new zealand journal of crop and horticultural science 31:281-291. vinale, f., marra, r., scala, f., ghisalberti, e. l., lorito, m., sivasithamparam, k. (2006) major secondary metabolites produced by two commercial trichoderma strains active against different phytopathogens. letters in applied microbiology 43:143-148. sv-lncs ruhuna journal of science vol 9(2): 150-159, december 2018 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v9i2.43 university of ruhuna  faculty of science, university of ruhuna 150 short paper detecting and correcting real-word errors in tamil sentences r. sakuntharaj1* and s. mahesan2 1centre for information and communication technology, eastern university, sri lanka 2department of computer science, faculty of science, university of jaffna, sri lanka *correspondence: *sakuntharaj@esn.ac.lk; orcid: 0000-0002-5689-6470 received: 15th march 2018, revised: 9th november 2018, accepted: 19th november 2018 abstract. spell checkers concern two types of errors namely non-word errors and real-word errors. non-word errors fall into two subcategories: first one is that the word itself is invalid; the other is that the word is valid but not present in a valid lexicon. real-word error means that the word is valid but inappropriate in the context of the sentence. an approach to correcting real-word errors in tamil language is proposed in this paper. a bigram probabilistic model is constructed to determine appropriateness of the valid word in the context of the sentence using a 3gb volume of corpora of tamil text. in case of lacking appropriateness, the word is marked as a real-word error and minimum edit distance technique is used to find lexically similar words, and the appropriateness of such words is measured by a word-level n-gram language probabilistic model. a hash table with word-length as the key is used to speed up the search for words to check for the lexical similarity. words of length differing less than two with the length of the ‘inappropriate’ word are considered to search in the hash table. test results show that the suggestions generated by the system are with 98% accuracy as approved by a scholar in tamil language. keywords. tamil, real-word error, bigram, minimum edit distance, error correction. 1 introduction tamil language possesses a rich history of more than 2000 years and spoken primarily in sri lanka, india, malaysia and singapore. tamil, a dravidian language, an official language of sri lanka, is an agglutinative language having rich morphological structure. tamil words are formed by lexical roots r. sakuntharaj and s. mahesan real-word errors in tamil sentences ruhuna journal of science vol 9(2): 150-159, december 2018 151 followed by one or more suffixes (navalar 1998, sangar 2006, nuhman 2013). spell checker is a tool that finds and corrects misspelt words in a text document. spelling error detection and correction techniques are widely used by text editing systems, search engines, machine translation systems, speech recognition systems, text to speech and speech to text conversion systems, speech recognition systems and optical character recognition systems. misspelt words can be classified into two categories of errors namely non-word errors and real-word errors (kukich 1992, samanta and chaudhuri 2013). non-word errors fall into two sub-categories: first one is that the word itself is invalid (e.g. மளை) and the other is that the word is valid (e.g. அைகு) but not present in a valid lexicon. real-word error means that the word is valid but inappropriate in the context of the sentence. for example, the sentence மளை நாட்டில் மளை பெய்யும் becomes non-sensical when the two words மளை and மளை are swapped. in this paper, we focus on detecting and correcting real-word errors in tamil sentences. we propose a method that uses bigram probabilistic model to detect the real-word errors and minimum edit distance technique to generate the suggestions for them while making use of a hash table to speed up the lookup. 2 methodology the input word is first checked with lexicon using a tree-based lookup algorithm to see whether the word is valid or invalid. letter-level and word-level bigrams & trigrams techniques are used to generate suggestions for the invalid words with two different types of hash tables to speed up the search. after correcting all the invalid words in the sentences, real-word errors are detected and corrected. as highlighted in figure 1, real-word errors are detected by considering the appropriateness of the words in the context of the sentence. the appropriateness of the words is determined by a bigram probabilistic model constructed using a pooled corpus of tamil text (as described in section 2.1). in case of lacking appropriateness, the word is marked as a real-word error, and then minimum edit distance (med) technique is used to find lexically similar words. these words are taken as candidates for suggestions to the word of concern that is marked as a realword error. the appropriateness of these candidates is measured by a wordlevel n-gram language probabilistic model (described in section 2.2). finally, the suggestions are refined by considering the probabilities of bigrams formed by each of these candidates with the word that follows the word of concern in the given sentence. it is interesting to note that the real-word errors are found r. sakuntharaj and s. mahesan real-word errors in tamil sentences ruhuna journal of science 152 vol 9(2): 150-159, december 2018 to be lexically similar to the word that was intended in the context. it is rather rare or unnatural in tamil to see real-word errors that are not lexically similar. fig. 1. flowchart of the real-word error detection and correction steps 2.1 bigram probabilistic model to determine the appropriateness of words the bigram probabilistic model constructed to determine the appropriateness of a word in the context of the sentence is described below. let the sequence of words (𝑤1, 𝑤2, … … , 𝑤𝑛 ) denote the sentence in the corpus. several tamil corpora collected from various sources including tamil news websites, tamil articles, tamil story books etc. were incrementally pooled into a single corpus. while they were collected, validity of words was checked by our system and texts were added to the corpus after correcting the invalid words. also new valid words were added to the lexicon. let (($𝑠𝑡𝑎𝑟𝑡, 𝑤1 ), (𝑤1, 𝑤2 ), (𝑤2, 𝑤3 ), (𝑤3, 𝑤4 ), … … … , (𝑤𝑛−1, 𝑤𝑛 ), (𝑤𝑛 , $𝑒𝑛𝑑 )) denote the word-level bigrams of the corpus. let 𝑝𝑚 be the bigram probability matrix that is constructed for the corpus using equation 1. 𝑝𝑚𝑖𝑗 = 𝑝𝑟 (𝑤𝑗 |𝑤𝑖 ) = 𝑐𝑜𝑢𝑛𝑡 𝑜𝑓 𝑏𝑖𝑔𝑟𝑎𝑚(𝑤𝑖 , 𝑤𝑗 ) + 1 𝑐𝑜𝑢𝑛𝑡 𝑜𝑓 𝑤𝑖 + 𝑛 𝑓𝑜𝑟 𝑖, 𝑗 = 1,2, . . 𝑛 (1) where n = |𝑉| is the size of the vocabulary set v. r. sakuntharaj and s. mahesan real-word errors in tamil sentences ruhuna journal of science vol 9(2): 150-159, december 2018 153 the size of the vocabulary set is 1,778,676. to calculate the probability 𝑝𝑚𝑖𝑗 , the laplace smoothing technique (jurafsky and martin 2018) is used to get rid of zero probabilities. as the result of this smoothing, the probability 𝐶 𝑁𝑤 of a 𝑃𝑟 (𝑤𝑖 | 𝑤𝑖−1) becomes 𝐶 + 1 𝑁𝑤+ 𝑛 . that is, the zero probability now becomes 1 𝑁𝑤+ 𝑛 after smoothing. to determine the appropriateness of word 𝑤𝑖, the probabilities 𝑃𝑟 (𝑤𝑖 | 𝑤𝑖−1) and 𝑃𝑟 (𝑤𝑖+1 | 𝑤𝑖 ) are considered. when either one of them is found to be too small (that is less than a prescribed threshold), 𝑤𝑖 is considered inappropriate in the context of the sentence. in the experiments it found that words with bigram probabilities less than 10-6 are not suitable in the context of the sentences, as determined by the scholar in tamil language, and thus the value 10-6 is chosen to be a good threshold for word to be considered as appropriate or inappropriate. let us consider an example to see how to detect a real-word error in a sentence: ‘அவன் ெைங்களைச் சாப்ெிட்டான்’. the bigram probabilistic model built for the corpus gives the bigram probabilities for this sentence as follows: pr (அவன் | $start) = 4.3623*10 -4 pr (ெைங்களைச் | அவன்) = 5.3920*10 -7 pr (சாப்ெிட்டான் | ெைங்களைச்) = 5.6221*10 -7 pr ($end| சாப்ெிட்டான்) = 8.7481*10 -5 of the above values, pr (ெைங்களைச் | அவன்) and pr (சாப்ெிட்டான் | ெைங்களைச்) are found to be less than the threshold value of 10-6, indicating the word ‘ெைங்களைச்’ is inappropriate in this sentence making a real-word error. how to find words lexically similar to the word under consideration, and how to choose the most appropriate one out the lexically similar ones are described below. 2.2 finding lexically similar appropriate words as the first step to generate suggestion for alternative words for inappropriate word 𝑤𝑖 of length mi, lexically similar words are found using minimum edit distance (med) technique. med is the minimum number of editing operation required to transform one string into another (damerau 1964, wagner and fisher 1974). a lexicon word that gives minimum edit distance with erroneous word less than four is considered as a lexically similar suggestion for the erroneous word. for the erroneous word of length mi, words of lengths from mi-1 to mi+1 are checked for lexical similarity. this range was determined by analysing the words with wider range of length. r. sakuntharaj and s. mahesan real-word errors in tamil sentences ruhuna journal of science 154 vol 9(2): 150-159, december 2018 after generating the lexically similar words for real-word errors, the appropriateness of the generated words is measured by a word-level n-gram language probabilistic model. the model helps choosing suggestions for erroneous words from the constructed corpus. the model is constructed for the sentences in the corpus as follows: let the sequence of words (𝑤1, 𝑤2, 𝑤3, … . . , 𝑤𝑛 ) denote the sentence in the corpus, (($𝑠𝑡𝑎𝑟𝑡), (𝑤1), (𝑤2), (𝑤3) … … , (𝑤𝑛 ), ($𝑒𝑛𝑑 )) denote its unigram sequence, (($𝑠𝑡𝑎𝑟𝑡, 𝑤1 ), (𝑤1, 𝑤2 ), (𝑤2, 𝑤3 ), … … , (𝑤𝑛−1, 𝑤𝑛 ), (𝑤𝑛 , $𝑒𝑛𝑑 )) denote its bigram sequence and (($𝑠𝑡𝑎𝑟𝑡, 𝑤1, 𝑤2), (𝑤1, 𝑤2, 𝑤3), (𝑤2, 𝑤3, 𝑤4 ), … … , (𝑤𝑛−2, 𝑤𝑛−1, 𝑤𝑛 ), (𝑤𝑛−1, 𝑤𝑛 , $𝑒𝑛𝑑 )) denote its trigram sequence. let 𝑝𝑖 denote a measure for a word 𝑤𝑖 calculated by using equation 2 as instructed in (jurafsky and martin 2018). 𝑝𝑖 = 𝜆1𝑝𝑖 (1) + 𝜆2𝑝𝑖 (2) + 𝜆3𝑝𝑖 (3) (2) where 𝜆𝑗s are positive scalars such that 𝜆1 + 𝜆2 + 𝜆3 = 1, 𝑝𝑖 (1) denotes the probability 𝑝𝑟 (𝑤𝑖 ) in the corpus, 𝑝𝑖 (2) denotes the probability 𝑝𝑟 (𝑤𝑖 | 𝑤𝑖−1) in the corpus, and 𝑝𝑖 (3) denotes the probability 𝑝𝑟 (𝑤𝑖 | 𝑤𝑖−2, 𝑤𝑖−1) in the corpus. the values of 𝜆𝑗s are tuned by using a set of test sentences, and the respective values 0.05, 0.7 and 0.25 are found to be working well with the constructed corpus. the probabilities 𝑝𝑖 (1) , 𝑝𝑖 (2) and 𝑝𝑖 (3) are calculated using equations 3, 4 & 5 respectively listed below: 𝑝𝑖 (1) = 𝑝𝑟 (𝑤𝑖 ) = 𝑐𝑜𝑢𝑛𝑡 𝑜𝑓 𝑤𝑖 𝑇𝑜𝑡𝑎𝑙 𝑛𝑜. 𝑜𝑓 𝑤𝑜𝑟𝑑𝑠 𝑖𝑛 𝑡ℎ𝑒 𝑐𝑜𝑟𝑝𝑢𝑠 (3) 𝑝𝑖 (2) = 𝑝𝑟 (𝑤𝑖 | 𝑤𝑖−1) = 𝑐𝑜𝑢𝑛𝑡 𝑜𝑓 𝑏𝑖𝑔𝑟𝑎𝑚𝑠 (𝑤𝑖−1, 𝑤𝑖 ) 𝑐𝑜𝑢𝑛𝑡 𝑜𝑓 𝑤𝑖−1 (4) 𝑝𝑖 (3) = 𝑝𝑟 (𝑤𝑖 | 𝑤𝑖−2, 𝑤𝑖−1) = 𝑐𝑜𝑢𝑛𝑡 𝑜𝑓 𝑡𝑟𝑖𝑔𝑟𝑎𝑚𝑠 (𝑤𝑖−2, 𝑤𝑖−1, 𝑤𝑖 ) 𝑐𝑜𝑢𝑛𝑡 𝑜𝑓 (𝑤𝑖−2, 𝑤𝑖−1) (5) after calculating the measures for all the suggestion words for an erroneous word using equation 2, the suggestion words are ranked based on their measures and top ranked words are picked as suggestions for the erroneous word under consideration. up to five words that give 𝑝 measure greater than the threshold value 10-6 are considered for suggestions. the value 10-6 is found to be a good threshold as being determined by analysing the words in r. sakuntharaj and s. mahesan real-word errors in tamil sentences ruhuna journal of science vol 9(2): 150-159, december 2018 155 the constructed corpus. moreover, it is found that the number of appropriate words is found to be less than five in many cases. let {�̂�𝑖 𝑗 }j=1…k, be a set of top k candidates for suggestions for inappropriate word 𝑤𝑖 . let 𝑤𝑖+1 be the word that comes next to 𝑤𝑖 in the sentence under consideration and the candidates �̂�𝑖 𝑗 that give the probability �̂�𝑗 = 𝑝𝑟 (𝑤𝑖+1|�̂�𝑖 𝑗 ) greater than 10-6 are taken as refined suggestions. let’s go back to our example sentence: ‘அவன் ெைங்களைச் சாப்ெிட்டான்’. once the word ‘ெைங்களைச்’ is found to be inappropriate as already described in section 2.1, the real-word error correction module generates lexically similar words for the word ‘ெைங்களைச்’ using med: the module finds the words {‘ெைங்களை’, ‘ெைங்களைச்’, ‘ெைங்கைிளைச்’, ‘ொைங்களைக்’, ‘ளவரங்களை’, ‘ககாைங்களைக்’, ‘சடைங்களை’} from the lexicon as lexically similar to ‘ெைங்களைச்’. these words are considered as candidates for suggestion. the word-level n-gram language probabilistic model ranks these candidates based on their measures of 𝑝𝑖 s that are calculated using equation 2 as described above. for the example considered above, 𝑝 measures for the suggestion candidates are: 𝑝1 = 𝑝 (ெைங்களைச்) = 2.7173*10 -4 𝑝2 = 𝑝 (ெைங்கைிளைச்) =1.7502*10 -5 𝑝3 = 𝑝 (ெைங்களை) = 8.0030*10 -6 𝑝4 = 𝑝 (ொைங்களைக்) = 8.3220*10 -6 𝑝5 = 𝑝 (ளவரங்களை) = 9.4058*10 -6 𝑝6 = 𝑝 (ககாைங்களைப்) = 2.8425*10 -8 𝑝7 = 𝑝 (சடைங்களை) = 7.3924*10 -9 after determining the measures of these lexically similar words, up to five words with measures more than a predetermined threshold 10-6 would be selected. for the above example, the words ‘ெைங்களைச்’, ‘ெைங்கைிளைச்’, ‘ெைங்களை, ‘ொைங்களைக்’ and ‘ளவரங்களை’ would be selected as their measures are higher than the threshold of 10-6. for these selected five words �̂�𝑗 = 𝑃𝑟 (𝑤𝑖+1|�̂�𝑖 𝑗 )j=1…5 are obtained from the bigram probability matrix as follows: �̂�1 = 𝑃𝑟(சாப்ெிட்டான் | ெைங்களைச்) = 2.3482 * 10 -3 �̂�2 = 𝑃𝑟(சாப்ெிட்டான் | ெைங்கைிளைச்) = 2.2393 * 10 -3 �̂�3 = 𝑃𝑟(சாப்ெிட்டான் | ெைங்களை) = 5.6187 * 10 -7 r. sakuntharaj and s. mahesan real-word errors in tamil sentences ruhuna journal of science 156 vol 9(2): 150-159, december 2018 �̂�4 = 𝑃𝑟(சாப்ெிட்டான் | ொைங்களைக்) = 5.5821 * 10 -9 �̂�5 = 𝑃𝑟(சாப்ெிட்டான் | ளவரங்களை) = 5.6219 * 10 -9 of these values, 𝑃𝑟 (சாப்ெிட்டான் | ெைங்களைச்) and 𝑃𝑟 (சாப்ெிட்டான் | ெைங்கைிளைச்) are greater than the threshold value of 10-6 and thus the words {‘ெைங்களைச்’, ‘ெைங்கைிளைச்’} are given as refined suggestions for the real-word error ‘ெைங்களைச்’. time taken for generating lexically similar words for 𝑤𝑖 is dependent on the lexicon size and lookup method. fortunately, using a hash table is a good approach to speed up the lookup as an alternative to mere linear search. in order to select the words of given length, the hash table with word-length as the key is built for the lexicon words and stored in a python pickle for easy and quick restoration, which significantly saves time. 3 results and discussion the programming language python 3.5 has been used to write the programs and run on ubuntu 16.04 on a dell machine with intel core i7-6500u processor of 2.5ghz clock speed and with 8gb ram. the proposed real-word error detection and correction system corrects following types of real-word errors that occur due to the wrong choice of letters from the confusion sets and due to ‘sandhi’ mistakes. sandhi inflects two words coming in between them. the following groups of letters form confusion sets: {ை, ை, ை}: in tamil, they sound somewhat similar to ‘la’, {ந, ண, ை}: these letters sound somewhat similar to ‘na’ and {ர, ற}: these letters sound somewhat similar to ‘ra’. each of these letters in each of these sets has distinct pronunciation, and has to be used appropriately with care. the error due to wrong choice of letters from any confusion set more likely to occur when a transliteration method is used to type tamil text or in speech to text conversion. for example, to input மளை one may type ‘malai’, ‘malai’, ‘malzi’ (as all sounds similar) instead of right choice for letter ‘ளை’ in the transliteration system. the sandhi mistake occurs due to the wrong suffix letter or missing of it. the right suffix depends on the word that follows it. it should be noted that the suggestions to all these errors are found to be lexically similar to the erroneous words. during the error detection and correction process, the inappropriate words in the sentences and suggestion words for them are recorded. the following five examples show the output of the real-word error detection and correction. example 1: (correcting contextual spelling mistakes that occur due to the r. sakuntharaj and s. mahesan real-word errors in tamil sentences ruhuna journal of science vol 9(2): 150-159, december 2018 157 letters in the confusion set {ை, ை, ை}) input sentence: அவன் ெைங்களைச் சாப்ெிட்டான். output: real-word errors: ெைங்களைச் suggestions: ெைங்களைச் / ெைங்கைிளைச் example 2: (correcting contextual spelling mistakes that occur due to the letters in the confusion set {ர, ற}) input sentence: அவள் களர ெடிந்த சட்ளட அணிந்திருந்தாள். output: real-word errors: களர or ெடிந்த suggestions: களற ெடிந்த or களர மடிந்த example 3: (correcting contextual spelling mistakes that occur due to the letters in the confusion set {ந, ண, ை}) input sentence: நாண் கநற்று ககாயிலுக்குப் கொகைன். output: real-word errors: நாண் suggestions: நான் example 4: (correcting sandhi (junctional inflection) mistakes between two words) input sentence: நான் அவளைச் கண்கடன். output: real-word errors: அவளைச் suggestions: அவளைக் / அவளைக் / அவளரக் example 5: (correcting any real-word error in a sentence) input sentence: அந்த விதியின் நடுகவ குைிபயான்று உள்ைது. output: real-word errors: விதியின் suggestions: வீதியின் / விடுதியின் / நதியின் in examples 1, 2 & 3, the real-word error detection and correction system detects and corrects the contextual spelling mistakes. in example 4, there is a ‘sandhi’ mistake between the words ‘அவளைச்’ and ‘கண்கடன்’. however, the proposed system identifies the word ‘அவளைச்’ as a real-word error and suggests the word ‘அவளைக்’ as the most appropriate alternative to ‘அவளைச்’ that results in correcting the sandhi mistake. in example 5, the real-word error detection and correction system identifies the word r. sakuntharaj and s. mahesan real-word errors in tamil sentences ruhuna journal of science 158 vol 9(2): 150-159, december 2018 ‘விதியின்’ as a real-word error and provides the words {‘வீதியின்’, ‘விடுதியின்’, ‘நதியின்’} as suggestions. this experiment is tested with two sets of sentences:  set 1 consists of one thousand sentences of random choice (none being taken from the corpus already built). these 1000 input test sentences have 131 real-word errors as confirmed by a scholar in tamil language, and the bigram probabilistic model detects all of them.  set 2 consists of 100 sentences deliberately made different from those in set 1 so that each has a real-word error, the bigram probabilistic model detects all of them. the following table shows the details of (a) the number of real-word errors existed in the chosen sets of sentences, (b) the number of real-word errors detected by the system, (c) the number of real-word errors with at least one suggestion with the bigram probabilistic measure greater than or equal to 10-6 as described in section 2.2, (d) the number of generated suggestions with the bigram probabilistic measure greater than or equal to 10-6 as described in section 2.2, and (e) the number of generated suggestions approved by tamil scholar. test set (a) (b) (c) (d) (e) set 1 131 131 99* 130 127 (97.7%) set 2 100 100 90* 156 154 (98.7%) * the reason for not getting any suggestion for 32 + 10 words is that the bigram involving the word in the input sentence that follows the suggestion candidate does not exist in the model constructed for the corpus. the test results show that 98% of the suggestions generated by the system are found to be suitable as approved by a scholar in tamil language. 4 conclusion in this work, a method has been proposed to detect and correct real-word errors in tamil sentences. in this regard, a bigram probabilistic model is constructed to detect real-word errors. minimum edit distance technique is used to generate suggestions. test results show that the bigram probabilistic r. sakuntharaj and s. mahesan real-word errors in tamil sentences ruhuna journal of science vol 9(2): 150-159, december 2018 159 model detects all the real-word errors in input sentences and suggestions generated by the system are with 98% accuracy as approved by a scholar in tamil language. acknowledgements two anonymous reviewers are acknowledged for their comments on the initial manuscript. references damerau f. 1964. a technique for computer detection and correction of spelling errors. communications of the association for computing machinery 7(3): 171–176, doi:10.1145/363958.363994 jurafsky d, martin jh. 2018. language modeling with n-grams. [online] available at: https://web.stanford.edu/~jurafsky/slp3/3.pdf kukich k. 1992. techniques for automatically correcting words in text. association for computing machinery computing survey 24(4): 377– 439, doi: 10.1145/146370.146380 navalar a. 1998. tamil grammar questions and answers. no. 366, kankesanthurai road, jaffna: vannai santhayarmadam, jaffna. nuhman ma. 2013. basic tamil grammar. university of peradeniya: readers association, kalmunai. samanta p, chaudhuri b. 2013. a simple real-word error detection and correction using local word bigram and trigram. proceeding of the 25th conference on computational linguistics and speech processing. kaohsiung, taiwan: 211–220. sangar v. 2006. tamil grammar. puduchcheri, india: nanmozi printers. wagner ra, fischer mj. 1974. the string to string correction problem. journal of the association for computing machinery 21(1): 168–173, doi:10.1145/321796.321811 https://web.stanford.edu/~jurafsky/slp3/3.pdf rjs-vol-1-sept-2006-2.dvi ruhuna journal of science vol. 1, september 2006, pp. 16–23 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. some hydrographic aspects of koggala lagoon with preliminary results on distribution of the marine bivalve saccostrea forskalli: pre-tsunami status k. b. suneetha gunawickrama and e. p. s. chandana department of zoology, faculty of science, university of ruhuna, matara, sri lanka, suneetha@zoo.ruh.ac.lk selected hydrographic parameters of koggala lagoon over a period of four months in 20022003 are reported together with the preliminary results on the distribution of a marine bivalve within lagoon. such data prior to the december 2004 tsunami may provide background information for comparison of post-tsunami conditions. moderate yet significant variation in most physico-chemical parameters was found among selected sites. mean salinity decreased from 34.0 ppt at the mouth to 11.4 ppt at 2150 m into the lagoon indicating pronounced seawater influx and mixing with freshwater. monthly inter-site variation in salinity was statistically significant (p<0.05). the marine bivalve saccostrea forskalli (chemnitz) was the dominant benthic mollusk with numbers (per m2) ranging from 111.5±9.7 at lagoon mouth to 23.5±2.1 at 2150m inner lagoon. density distribution of s. forskalli was significantly correlated (p<0.05) with the overall variation in salinity (r=0.889). salinity-dependent colonization success of s. forskalli demonstrates the dominance of marine conditions in koggala lagoon. key words : koggala lagoon, hydrography, saccostrea, seawater influx, lagoon dynamics. 1. introduction. coastal wetlands including lagoons are increasingly subject to human interventions that are driven by economic and developmental needs. such activities may cause natural hydrographic conditions of lagoons to change over time, consequently making a great impact on the wellbeing of the life forms as well. in addition to anthropogenic activities, infrequent natural events like the december 2004 indian ocean tsunami may have great impacts on environmental conditions in coastal wetlands. in conservation point of view, it is important to monitor the trends in changes within these ecosystems, while studies looking into such aspects have to rely on previously reported data to have comparative analyses. koggala lagoon, a basin estuary located in the southern coast of sri lanka (06°00′ n; 80°20′ e) (anon. 1988) is a nationally important wetland site according to the currently adopted valuation criteria (cea 1995, 1999). it covers a relatively large surface area (approx. 727 ha) measuring 4.8 km in length and 2 km across, and is relatively deeper (1.04.0 m) compared to other coastal water bodies along the 16 gunawickrama and chandana: some hydrographic aspects of... ruhuna journal of science 1, pp. 16–23, (2006) 17 southern coast (silva 1996). its innermost region receives a large influx of freshwater from few streams, and the seaward end has a narrow outlet named ‘pol oya’ (cea 1995) (figure 1). various land use practices exist around this wetland, which mainly includes small-scale fishing industry within lagoon and paddy cultivation close to the landward end of the lagoon (amarasinghe 1998). the koggala export processing zone (kepz), largely focused on textile manufacturing is located within the watershed area of the lagoon. presently, the natural outlet of koggala lagoon has been considerably modified with the construction of an artificial barrier (breakwater) at the mouth. as a result, the outlet has been diverted westward creating an open passage to the sea (approx. 30-40 m wide). it is hypothesized that the lagoon has been subjected to prolonged seawater intrusion affecting its hydrographic conditions. one indication for such a change may be the occurrence of marine bivalve saccostrea forskalli (chemnitz) in the koggala lagoon. scientific information is lacking on their inward distribution. the tsunami event in december 2004 caused a great environmental damage to the nearby area of the koggala lagoon. although the lagoon faced a full-scale impact of the surging wave, no studies have been carried out on the effects of the tsunami on the lagoon so far. scientific information gathered before the tsunami event is immensely important for such comparisons. the present paper reports selected hydrographic characteristics of koggala lagoon for a period of four months during 2002-2003 with preliminary results on the density distribution of the marine bivalve saccostrea forskalli. 2. materials and methods 2.1. sampling nine locations along the approximate lagoon axis were selected for sampling (figure 1). the distance (m) from the mouth to the consecutive sampling locations (1-9) were 0 (at the outlet), 50, 150, 400, 650, 900, 1150, 1650 and 2150. at each sampling location (1-9), four sites that were approximately 1-2 m apart from each other were selected for replicate sampling (figure 1). five consecutive sampling was carried out during the four-month period (12th of november, 7th and 28th of december in 2002, 21st of january and 23rd of february in 2003). water samples were collected from each site within the first meter of the water column. 2.2. physico-chemical parameters in-situ measurements of depth (m), secchi depth (transparency, cm), ph (by ph meter), water temperature (�) and conductivity (conductivity meter, in ms/cm) were taken at each sampling site. water samples were collected for salinity (by agno3 titrative method in parts per thousand; ppt), dissolved oxygen (winkler’s method, mg/l) and dissolved phosphate ion concentration (spectrophotometric method; µg/l) to be analyzed in the laboratory. the mean values (n=4) of each parameter at given sampling times were tested for among-location variation and short-term temporal variation using one-way anova (p<0.05 was considered significant). all statistical and data analyses were done using statisticarv 7.0 (statsoft, usa). gunawickrama and chandana: some hydrographic aspects of... 18 ruhuna journal of science 1, pp. 16–23, (2006) figure 1 the map of koggala lagoon showing the selected sampling sites (insert at top left corner). 0 5 0 1 5 0 4 0 0 6 5 0 9 0 0 1 1 5 0 1 6 5 0 2 1 5 00 5 0 1 5 0 4 0 0 6 5 0 9 0 0 1 1 5 0 1 6 5 0 2 1 5 0 0 5 0 1 5 0 4 0 0 6 5 0 9 0 0 1 1 5 0 1 6 5 0 2 1 5 00 5 0 1 5 0 4 0 0 6 5 0 9 0 0 1 1 5 0 1 6 5 0 2 1 5 00 5 0 1 5 0 4 0 0 6 5 0 9 0 0 1 1 5 0 1 6 5 0 2 1 5 0 45 40 35 30 25 20 15 10 45 40 35 30 25 20 15 10 c o n d u c ti v it y ( m s ) a n d s a n it y ( p p t) distance from lagoon mouth (m) conductivity (ms) salinity (ppt) november 12, 2002 december 7, 2002 december 28, 2002 january 21, 2003 february 23, 2003 figure 2 variation in mean salinity and conductivity at the selected sites of koggala lagoon over a period of four months during 2002-2003 (±sd, n=4). note. (x-axis in the plots is shifted for the two parameters for clarity). gunawickrama and chandana: some hydrographic aspects of... ruhuna journal of science 1, pp. 16–23, (2006) 19 o y st e r d e n si ty ( n u m b e r/ m )2 o y st e r d e n si ty ( n u m b e r/ m )2 distance from lagoon mouth (m) salinity (ppt) 140 120 100 80 60 40 20 0 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 r = 0.789 ; r = 0.889 , p<0.001 ; y = -18.27 + 3.36 * x 2 0 50 150 400 650 900 1150 1650 2150 110 90 70 50 30 10 figure 3 variation in mean oyster (saccostrea forskalli) density in koggala lagoon (top) and the plot of regression analysis between oyster density and salinity. note. the error bars represent standard deviation (top) while dashed lines indicate 95% confidence limits (bottom). 2.3. density distribution of saccostrea forskalli density of the bivalve (number per m2) was estimated by placing four random replicate quadrates (1 m2) at each sampling location, and by underwater observation and counting the numbers in-situ (it was easily done at sampling locations where bottom was visible), on one sampling date only (28th december 2002). statistical correlation of bivalve density with the among-site variation in salinity and conductivity during december 2002 was tested by pearson’s correlation analysis and least square regression analysis. 3. results 3.1. lagoon characteristics the mean depth of the sampling area ranged from 0.69 m near lagoon mouth to 2.9 m at 1.15 km distant from mouth over the four months. secchi disc transparency of the water column was high over the months ranging between 119 cm (site 4) and 136 cm (site 9). bottom of the lagoon was visible up to a distance of 150-200 m from the mouth. frequent rapid inflow of sea water into the lagoon was observed whenever moderate to strong wave action was present while the outflow was less frequent, occurring under very calm sea conditions and low tide. among the physico-chemical parameters studied, most displayed sharp gradients from the lagoon mouth towards the inner lagoon. particularly, conductivity and salinity of lagoon water (table 1) at each sampling day markedly decreased towards gunawickrama and chandana: some hydrographic aspects of... 20 ruhuna journal of science 1, pp. 16–23, (2006) the inner lagoon (figure 2). mean conductivity was highest at the mouth (site 1 & 2) under high influence of seawater ranging from 25.3 ms cm−1 to 39.4 ms cm−1 during the four months. conductivity at each sampling site was significantly different (anova, p<0.05) between the five sampling times (sites 8 and 9 were not included in this test) while along-axis variation was highly significant (anova, p<0.001) at all sampling times. salinity was highest at the mouth and ranged between 31.3 and 34.0 ppt, while the lowest salinity of 11.4±0.8 ppt was recorded at the innermost sampling site (2.15 km from mouth) during third week of december 2002 (figure 2). salinity ranged from 15.5 to 18.5 ppt during the four-month period. significant temporal variation in salinity was found at all sampling locations except at locations 7 and 8. among site variation in salinity was highly significant in all sampling times indicating a strong salinity gradient (reverse) from mouth towards inner lagoon (anova, p<0.05). mean water temperature ranged from 28.6 �(at the mouth) to 30.1 �(innermost site). the lagoon water was largely alkaline with ph varying between 8.7 and 9.6. although some significant differences in ph were found among sites (results not given) there was no discernible gradient (table 2). amount of dissolved oxygen was relatively high (10.0-11.6 mg/l), at some point up to the level of saturation, and no marked among-site variation was found. the concentration of total dissolved phosphate ions was highly variable and significantly different among most of the sites, but no discernible trend was found. during november and december, 2002, significantly high phosphate ion concentration was found close to the mouth (sites 1,2 & 3) compared to inner lagoon (anova, p<0.05). 3.2. distribution of the marine bivalve the marine oyster saccostrea forskalli (chemnitz 1758) (mollusca: bivalvia) was the dominant mollusk in the littoral areas and benthic regions with some hard substrates. most submerged rocky substrates (boulders) at the mouth region were covered with the oyster and its distribution has extended up to more than half way into the lagoon (figure 3). the mean densities (number/ m2) ranged from 111.5 ± 9.7 at the outfall to 23.5 ± 2.1 at 2.15 km, close to the midpoint of the lagoon axis (table 1). the among-site variation in the densities was highly significant and correlated with the variation in salinity and conductivity during december, 2002 (figure 3). 78.9% of the among-site variation in oyster density can be explained by the variation in salinity alone (r2 = 0.789; pearson’s r = 0.889; p<0.001). oyster density was significantly correlated with conductivity as well (pearson’s r = 0.65, p<0.001). 4. discussion the observed gradients in salinity and conductivity from the lagoon mouth towards inner lagoon could be attributed to the strong influence of sea water and freshwater fluxes. high rainfall was reported during november in the south region of the country owing to the second inter-monsoonal season (october-november) and the reduced salinity found during december could be attributed to the enhanced freshwater gunawickrama and chandana: some hydrographic aspects of... ruhuna journal of science 1, pp. 16–23, (2006) 21 table 1 descriptive statistics of salinity and conductivity together with the oyster density in koggala lagoon (displayed as mean±sd; the values in each column are respectively from five consecutive sampling times (1: nov 12, 2002; 2: dec 7, 2002; 3: dec 28,2002; 4: jan 21, 2003; 5: feb 23, 2003)). site 1 2 3 4 5 6 7 8 9 distance from mouth (m) (0) (50) (150) (400) (650) (900) (1150) (1650) (2150) conductivity1 28.3±0.1 25.3±0.2 25.9±0.3 25.4±0.3 26.3±0.1 23.6±0.1 23.5±0.8 n.m n.m (ms) 2 29.2±0.7 29.9±4.5 27.9±2.5 27.8±1.3 24.3±0.6 22.2±3.0 15.7±1.3 n.m n.m 3 39.0±2.2 39.5±0.6 39.1±0.1 38.5±0.4 37.4±0.2 38.1±0.1 37.3±0.1 31.7±1.2 28.4±0.3 4 27.8±3.6 25.3±0.9 25.2±0.7 22.5±0.3 21.2±2.7 18.8±0.8 18.6±0.3 18.3±0.7 n.m 5 33.2±0.8 30.5±1.7 29.2±0.5 28.3±0.9 23.3±0.9 25.0±2.7 22.6±0.3 20.3±0.5 n.m salinity 1 33.5±1.6 28.9±0.2 22.6±1.1 22.2±0.8 18.8±0.9 19.5±0.3 18.1±0.6 n.m n.m (ppt) 2 34.0±1.2 30.3±4.5 26.9±0.9 22.4±0.2 19.8±1.1 17.3±0.6 15.5±2.5 n.m n.m 3 33.8±1.8 31.5±0.6 27.8±1.4 26.5±1.1 24.0±0.7 17.7±1.2 16.5±1.7 15.7±2.6 11.4±0.8 4 31.3±0.7 28.3±0.9 24.0±1.7 23.4±0.6 18.4±0.8 17.1±0.8 17.1±0.7 17.2±0.6 n.m 5 33.5±0.6 29.2±1.7 27.5±0.5 24.4±0.9 22.1±0.7 20.5±0.5 18.5±0.4 18.4±0.7 n.m oystera 3 111.5±9.7 97.3±3.8 69.8±15.246.8±2.2 50.3±1.7 42.5±2.6 45.0±3.2 36.0±7.1 23.5±2.1 (asaccostrea forskalli density in number per m2, n. m.=not measured ). table 2 selected physical and chemical parameters of koggala lagoon displayed as the mean values (within parentheses are the number of sampling times). among site variation was significant (anova, p<0.05) for all parameters. site 1 2 3 4 5 6 7 8 9 distance from 0 50 150 400 650 900 1150 1650 2100 mouth(m) ph 9.5 (3) 9.6 (2) 9.6 (3) 9.1 (3) 8.7 (3) 8.9 (3) 9.0 (3) 9.0 (1) 8.7 (1) temperature (�) 28.6 (5) 28.9 (5) 29.0 (5) 29.1 (5) 29.2 (5) 29.2 (5) 29.4 (5) 29.3 (3) 30.1 (1) depth (m) 0.86 (5) 0.69 (5) 1.67 (5) 2.05 (5) 2.16 (5) 2.46 (5) 2.92 (5) 3.40 (3) 2.60 (1) secchi depth b.v. b.v. 119 (5) 121 (5) 130 (5) 128 (5) 128 (5) 133 (3) 136 (1) (cm) dissolved 10.3 (5) 10.9 (5) 10.0 (5) 10.7 (5) 11.6 (5) 11.5 (5) 11.1(5) 10.7 (3) 10.5 (1) o2 (mg/l) phosphate 242.7 (3) 238.4 (3) 225.5 (3) 169.9 (3) 168.2 (3) 137.9 (3) 143.5 (3) 152.9 (1) 208.7 (1) (µg/l) b.v.= bottom visible influx from the streams connected to the head region. characteristics of the koggala lagoon have been reviewed previously by presenting earlier reported values by silva (1996). the salinity values reported in the present study in november are apparently higher than those reported during 1981-1982, (min-max: 0.01-18.0 ppt) (silva 1996). the influence of seawater influx enabled by the breakwater seems to have dominated gunawickrama and chandana: some hydrographic aspects of... 22 ruhuna journal of science 1, pp. 16–23, (2006) the lagoon water as far as 1-2 km from the lagoon mouth. evaporation is identified as a governing factor of the characteristics of lagoon waters especially under closed conditions (hadlec et al. 1988). therefore, evaporation may affect lagoon hydrography causing higher saline conditions especially during droughts. under present situation, salinity is likely to be largely regulated by prolonged seawater intrusion but not by evaporation losses. salinity value as high as 19.6 ppt was reported at the middle of the lagoon during march 1994 (silva 1996), in agreement with the rise in saline conditions reported during that period (amarasinghe 1998). the values reported in the present study further support the characterization of koggala lagoon as a dynamic system having large annual variations in hydrographic parameters (cea 1995). strong inward flow of seawater forced by the near-shore wave action at the koggala lagoon mouth reflected the potential influence of seawater on the lagoon hydrography. under natural conditions, the lagoon hydrography is strongly governed by the tidal influence and freshwater influx from the head region (silva 1996). although the tidal influence on lagoon dynamics still exists, the effect of wave action apparently dominates over it. the predominance of more saline conditions in the lagoon seems to have enabled invasion of some marine species into the lagoon. the marine bivalve saccostrea forskalli has never been reported in previous studies within the lagoon (see cea 1995). it is a marine species highly abundant in eulittoral to sublittoral zones of many rocky shores around sri lanka (kirthisinghe 1978). the occurrence of this species strongly indicates the environment is presently conducive to their existence within lagoon, and perhaps the lagoon provides better conditions deprived of the near-shore intertidal hardships usually faced by them on rocky shores. the habitat expansion of the oyster inward to the lagoon exemplifies the overwhelming influence of seawater probably lasting over a long period, and hence this marine bivalve could be considered as an indicator organism displaying the dominance of marine conditions within the lagoon. further investigations are needed in this regard. koggala lagoon is known to harbor a diverse fauna as reported previously (cea 1995, 1999), and thus has a high ecological significance. several endemic species have been reported and especially a large species diversity of fishes is known to exist (cea 1995, 1999). therefore, from conservation point of view, the lagoon bears a high value. as most of the lagoons in southern coastal belt remain closed for large part of the year (silva 1996), they may serve as separate geographical entities that offer conditions for the inhabiting populations to become locally adapted and becoming isolated from other such populations. this process is of high evolutionary significance. however, in recent times, the existing biodiversity in the koggala lagoon is reportedly declining owing to various human activities (cea 1995). undesirable changes to the lagoon hydrography inevitably cause habitat deterioration for the inhabitants that no longer can tolerate the changed environmental conditions. recent indian ocean tsunami (december 2004) caused great damage to the surrounding area of the lagoon while the lagoon itself has been subjected to the direct impact of the big wave. it seems a timely requirement to investigate the changes gunawickrama and chandana: some hydrographic aspects of... ruhuna journal of science 1, pp. 16–23, (2006) 23 that may have occurred due to this natural event, yet such studies have to rely on data gathered prior to the event. in this sense, such data provide important background information for comparative studies. the present results may provide some information from the pre-tsunami era of the lagoon system. in summary, hydrological conditions studied during the four-month period provide evidence for the dynamic nature of the koggala lagoon. strong sea water influence apparently caused by the prolonged opening of the lagoon system and strong onrushing wave action seem to have felt by the lagoon as far as about 2 km from the lagoon outlet. the seawater influence is further indicated by the colonization success of the marine oyster s. forskalli within the lagoon, which has not been reported from the koggala lagoon prior to 1995. references amarasinghe, o. (1998). profitability of current land use practices in five saltwater exclusion and drainage (wsed) schemes. report of the swed project under southern province rural development project, sri lanka. anonymous (1988). national atlas of sri lanka, survey department, colombo, sri lanka. cea (1995). wetland site report and conservation management plan, koggala lagoon. wetland conservation projectcentral environmental authority of sri lanka/ arcadis/ euroconsult. 1-82 p. cea (1999). wetland atlas of sri lanka. wetland conservation projectcentral environmental authority of sri lanka/ arcadis/ eurocunsult. 1-75 p. hadlec, r. h., williams, r. b. and scheffe, r. d. (1988). wetland evapotranspiration in temperate and arid climates. in: hook, d. d., mckee jr, w. h., smith, h. k., gregory, j., burrell jr, v. g., devoe, m. r., sojka, r. e., gilbert, s., banks, r., stolzy, l. h., brooks, c., matthews, t. d. and shear, t. h. (eds.), the ecology and management of wetlands, vol. 1, pp. 146-160, timber press, portland, oregon. kirthisinghe, p. (1978). sea shells of sri lanka: including forms scattered throughout the indian and pacific oceans. rutland, vt: c. e. tuttle. silva, e. i. l. (1996). water quality of sri lanka, a review on twelve water bodies. institute of fundamental studies, peradeniya. ruhuna journal of science vol 11 (1): 13-22, june 2020 eissn: 2536-8400 © faculty of science doi: http://doi.org/10.4038/rjs.v11i1.83 university of ruhuna © faculty of science, university of ruhuna 13 sri lanka estimation of selected biological parameters of the purple mangrove crab goniopsis pelii (herklots 1851) from a tropical mangrove swamp in nigeria r.o. moruf department of fisheries and aquaculture, faculty of agriculture, bayero university, kano, p.m.b. 53011, gwarzo road, kano, kano state, nigeria correspondence: tunjimoruf@gmail.com; https://orcid.org/0000-0002-0459-0621 received: 7th august 2019, revised: 24th may 2020, accepted: 8th june 2020 abstract. the purple mangrove crab, goniopsis pelii is one of the more elusive crab species inhabiting most of the microhabitats in the mangrove ecosystem of southern nigeria. this study was carried out to investigate the size composition, growth pattern, diets, and reproductive biology of g. pelii from abule-eledu mangrove swamp in lagos lagoon, nigeria. a total of 116 specimens were studied. data revealed that the carapace length and total weight ranged from 2.5 to 4.9 cm (3.72 ± 0.39) and from 11 to 37 g (23.90 ± 4.21) respectively. the crab exhibited negative allometric growth pattern (b<3), and the carapace lengthweight relationship was positively correlated (r = 0.577-0.900). the condition factor (k) ranged from 3.5-8.0 (male), 3.2-7.5 (female) and 3.4-8.0 (sexes combined). the highest k-values were recorded for the smaller size group (2.5 2.9 cm). in terms of sex, the male has the higher k-value (8.0). identified food particles included plant materials, algae, detritus, diatoms, and protozoa. the sex ratio was 1:0.57 (male: female), and was significantly different from the theoretical ratio of 1:1. fecundity ranged between 1.9 and 2.8 million eggs with egg diameter ranging from 0.26 to 0.38 mm. crabs with partial gonad development were predominant in the mangrove, while adults with fully ripe gonads were encountered occasionally throughout the study period. in this study, the reproductive indices and body conditions are indicators of a viable population. keywords: allometric growth, crab feeding, reproductive biology, lagos lagoon. 1 introduction mangroves are salt-tolerant, characteristically complex plant communities occurring in sheltered coastline areas in the tropical and subtropical intertidal regions such as bays, estuaries, lagoons, and creeks (venkataraman 2007). one of the more elusive crab species occurring in brackish water environments in nigeria is the purple mangrove crab (goniopsis pelii herklots, 1851), inhabiting almost every microhabitat in the mangrove ecosystems. the purple mangrove crab populations are dependent on mangroves for shelter and food, hence any impact that significantly alters the https://creativecommons.org/licenses/by-nc/4.0/ https://orcid.org/0000-0002-0459-0621 https://orcid.org/0000-0002-0459-0621 r.o. moruf biology of the purple mangrove crab ruhuna journal of science vol 11 (1): 13-22, june 2020 14 mangrove ecosystem may affect them. although the mangrove crab does not really constitute a food item for the coastal communities, they play a major ecological role in the mangrove ecosystem through processing fallen leaves for feeding (usese et al. 2018). in nigeria, mangroves are threatened by anthropogenic activities such as pollution and dredging as well as natural phenomenon like sea level rise (lawal-are et al. 2019) the purple mangrove crab is a brightly colored crab, having a dark carapace with white spots, and hairy red legs with white or yellow spots. female crab has a wide abdomen for egg storage, while males have thin abdomen covering the gonopods. most mangrove crabs are amphibious in habit and can be found around the intertidal areas with moist/ wet muddier regions of the mangrove (moruf and lawal-are 2018). crabs are mainly opportunistic feeders, while some are scavengers with plant-derived detritus comprising the main food consumed. in common with other benthic detritus feeding invertebrates, crabs are not indifferent to cannibalism when kept in captivity and even in the wild (hill and o’keefe 1992). generally, crabs retain the ability to eat different things but tend to be best adapted for particular diets. according to feller and sitnik (1996), mangrove crabs have been observed feeding on mangrove propagules, insects, and organic material with special affinity for a foraging existence. they will eat anything that gets washed upon shores or in mangrove such as algae, carrion, seaweeds, debris and anything small enough to be grasped with the chelae, because their chelae and gastric mill are non-specialized (moruf and ojetayo 2017). however, mangrove crabs might have been overlooked as they are certainly able to shred leaf litter very effectively (dobson et al. 2002). this combined with the general abundance and high biomass makes them potentially important in the dynamics of energy resources in mangrove ecosystems (hill and o’keeefe 1992). purple mangrove crab provides food for other species and reduces the insect population within the mangrove ecosystem. by eating the leaf litter and organic material around the mangrove roots, mangrove crabs contribute to nutrient cycling (moruf and ojetayo 2017). as little is known about the mangrove crabs of lagoon creeks in lagos, this paper bears significance to provide baseline data on the biology of the purple mangrove crab in the abule-eledu mangrove swamp of lagos lagoon, with particular emphasis on size composition, growth pattern, condition factor, feeding habits and reproductive biology. 2 materials and methods 2.1 study site the study was carried out in the mangrove swamp bordering the abule-eledu creek (figure 1), which forms part of the many sluggish tidal creeks that drain into the lagos lagoon, nigeria (gps coordinates: 60 31.015'n, 3023.948'e, elevation: 14ft (4.27m) above sea level. r.o. moruf biology of the purple mangrove crab ruhuna journal of science vol 11 (1): 13-22, june 2020 15 fig. 1. map of study location in mangrove swamps of abule-eledu creek, nigeria 2.2 sample collection random sampling was carried out between 17.00 and 19.00 hours on monthly bases from march to july 2017. a total of 116 specimens of g. pelii (figire 2) were caught from the mangrove swamps of abule-agege creek using baited traps (emmanuel 2009) and hand-picked using protective rubber gloves. specimens were identified using published taxonomic keys (schneider 1990) and immediately transported to the laboratory for further analysis. fig. 2. purple mangrove crab (goniopsis pelii) (collected from mangrove swamps of abule-eledu creek, nigeria) 2.3 morphometric measurements the length of the crab was measured to the nearest 0.1cm using a vernier caliper and plotted against their respective frequencies. the total weight was measured in grams using sartorius top loading balance (model: dt1001a). the length-weight relationship of the crabs is represented using the equation (pauly 1983) as follows: 𝑊 = 𝑎𝐿𝑏 r.o. moruf biology of the purple mangrove crab ruhuna journal of science vol 11 (1): 13-22, june 2020 16 where, w= body weight (g), l= carapace length (cm), a= regression constant (intercept), and b= regression coefficient (slope). the relationship was logarithm transformed into a straight-line (parsons 1988) by 𝐿𝑜𝑔 𝑊 = 𝐿𝑜𝑔 𝑎 + 𝑏 𝐿𝑜𝑔 𝐿 condition factor (k) of each crab was determined using the formula (bannister 1976) by, 𝐾 = 100 × ( 𝑊 𝐿𝑏 ) 2.4 stomach content analysis the cardiac stomach of each specimen was dissected, and the contents were extracted into a petri dish. the extracted contents were mixed with a little water and examined under a binocular microscope for the food types using the numerical and occurrence methods (cortés 1998). 2.5 reproductive biology the sex was distinguished using the species conspicuous external morphological features, i.e. t-shaped abdomen of the male and the triangular or rounded aprons of the female (schneider 1990). sex ratio was tested for any deviation from the expected 1:1 ratio using chi-square analysis. in fecund females, egg mass attached to the underside of the abdomen was removed for fecundity estimation. fecundity was calculated as the number of eggs carried externally by the female. it was estimated by the gravimetric method (kumar et al. 2000) from the ovaries at stage iii. the diameters of eggs per berried females (sub sample) were measured using an ocular micrometer. gonadal development was determined using the criteria of lawal-are (2010) given below: stage i: no gonad development stage ii: partial gonad development stage iii: gonads extending in carapace, which are orange in colour stage iv: ripe running bright red gonads 2.6 data analysis data were analyzed using microsoft excel 2010 and spss software. test for goodness of fit was determined statistically using chi-square (χ2) test on sex ratios of the species. 3 results and discussion 3.1 size composition amongst mangrove macrobenthos, crabs are one of the most significant groups in terms of species, numbers and total biomass. monthly abundance of g. pelii revealed r.o. moruf biology of the purple mangrove crab ruhuna journal of science vol 11 (1): 13-22, june 2020 17 that the highest occurrence was at the onset of wet season with 23.3 % in may, while the lowest number was recorded in july 2017 (figure 3). fig. 3. monthly collection of g. pelii (n= 116) from the mangrove swamp of abule-eledu creek, lagos lagoon the crabs with carapace length between 3.50 cm and 3.90 cm was the most frequently occurring size group, with the length frequency polygon showing a unimodal size distribution during the sampling period (figure 4). this report conformed to the work of onadeko et al. (2015) on brachyuran crabs of the same study area, which showed the highest abundance in may of the year. fig. 4. carapace length frequency distribution of g. pelii (n= 116) from the mangrove swamp of abule-eledu creek (marchjuly 2017) 3.2 growth pattern the growth pattern of the crabs was based on the length/weight relationship. the carapace length was between 2.5 and 4.9 cm and total weight between 11 and 37 g. the carapace length-total weight of the crabs was transformed into a logarithm form as shown below: male g. pelii: log tw = 0.5299 + 1.4271 log cl female g. pelii: log tw = 0.649 + 1.1988 log cl combined sexes: log tw = 0.148 + 2.1217 log cl r.o. moruf biology of the purple mangrove crab ruhuna journal of science vol 11 (1): 13-22, june 2020 18 fig. 5. relationship of log carapace length/log total weight of male g. pelii from mangrove swamp of abule-eledu creek (march july 2017) fig. 6. relationship of log carapace length/log total weight of female g. pelii from mangrove swamp of abule-eledu creek (march july 2017) according to figures 5, 6, and 7, the values of ‘b’ were 1.4271, 1.1988 and 2.1217 for the males, females and combined sexes respectively. these values showed that g. pelii exhibited a negative allometric growth pattern (b < 3). the correlation coefficient (r) was 0.8533 for the males, while 0.9001 for females and 0.5773 for combined sexes of g. pelii, showing correlation between the carapace length and weight. fig. 7. relationship of log carapace length/log total weight of g. pelii from the mangrove swamp of abule-eledu creek (march july 2017) r.o. moruf biology of the purple mangrove crab ruhuna journal of science vol 11 (1): 13-22, june 2020 19 the relationships between the length and weight suggest that the crab weight could be estimated from its length regardless of the developmental stages. another co-occurring mangrove crab, the hairy mangrove crab, sersema huzardii shows positive relationships between length and weight (lawal-are and nwankwo 2011). similarly, moruf and lawal-are (2017a) reported a positive correlation for portunid crabs of lagos coast, suggesting that the higher the carapace length the more weight an individual possesses. 3.3 condition factor the variations in condition factor (k) by size and sex of g. pelii from the abule-agege mangrove swamp are presented in table 1. the k-values ranged from 3.58.0 (male), 3.2 -7.5 (female) and 3.4 8.0 (combined sexes). the highest k-values were recorded for the small size (2.5-2.9 cm) group, while in terms of sex; the male had the higher kvalue (8.0). table 1: condition factor by sex and size of goniopsis pelii. n = number of crabs, cl = carapace length, w = total weight, k = condition factor in studies of population dynamics, high ‘k’ values of a crab show favorable environmental conditions such as habitat and prey availability. it was observed from this study that the small sized group had the highest condition factors, indicating successive growth as a result of molting activity. the older the crab, the more difficult it is for the crab to molt. on the average, the condition factors of the male g. pelii were higher than that of its female. 3.4 stomach contents of the 116 examined specimens of g. pelii, 14 (12%) of these had empty stomachs. the stomach contents consisted mainly of plant materials, filamentous algae, detritus, diatoms and protozoa. plant materials showed the highest numerical count (50.1%) and formed the most important food item, occurring in 97.1% of the crabs examined. diatoms and detritus occurred in 92.2% and 85.3% of the stomachs respectively (figure 8). g. pelii can be considered as herbivore-scavenger species as its diet contained a wide range of plant materials, filamentous algae, detritus, diatoms and protozoa. except for protozoa, the broad spectrum of the food items ingested by g. carapace length (cm) male female combined sex n cl (cm) w (g) k n cl (cm) w (g) k n cl (cm) w (g) k 2.52.9 18 2.7 15.7 8 11 2.7 14.8 7.5 2 2.7 15.7 8 3.0 -3.4 16 3.1 18.2 6.1 8 2.9 16.2 6.6 31 3.1 17.5 5.9 3.53.9 24 3.7 18.8 3.7 12 3.5 18.2 4.3 52 3.7 18.6 3.7 4.0 -4.4 11 4.2 33 4.5 5 4.2 28.2 3.8 27 4.2 30.7 4.1 4.5 -4.9 5 4.7 36.6 3.5 6 4.7 33.4 3.2 4 4.7 35 3.4 74 42 116 r.o. moruf biology of the purple mangrove crab ruhuna journal of science vol 11 (1): 13-22, june 2020 20 pelii is of plant-origin while detritus is regarded as debris of any kind. this report is similar to what was recorded by eteobong et al. (2016) on diet composition of g. pelii from south west nigeria. the feeding habits of mangrove crabs play a vital role in the preservation of wetland environments by sifting through the sands, and they aerate the substrate and prevent anaerobic conditions. fig. 8. stomach content of g. pelii from the mangrove swamp of abule-eledu creek 3.5 reproductive biology a total of 74 males and 42 females of g. pelii were examined giving a sex ratio of 1: 0.57 (male: female) which was significantly different (p<0.05) from the expected and theoretical ratio of 1:1. this is in contrary to the report of eteobong et al. (2017) who reported more female than male g. pelii in lagos lagoon. environment factors can influence the sex ratio of aquatic organism either directly or indirectly (moruf et al. 2018) fecundity was determined using 18 mature females (3.64.9 cm carapace length and 18.6-35.0 g weight) obtained only during the months of march and april. the fecundity ranged from 1.9 to 2.8 million eggs with egg diameter ranging between 0.26 and 0.38 mm. crablets with partial gonad development (stage ii) were predominant in the mangrove while ripe males and females (stage iv) occurred occasionally throughout the study period (table 2). table 2: percentage occurrence of different maturity stages of gonads of g. pelii from the mangrove swamp of abule-eledu creek (march – july 2017). gonad stages male female number % occurrence number % occurrence stage i 6 8.1 4 9.5 stage ii 41 55.4 20 47.6 stage iii 25 33.8 16 38.1 stage iv 2 2.7 2 4.8 total 74 42 r.o. moruf biology of the purple mangrove crab ruhuna journal of science vol 11 (1): 13-22, june 2020 21 fecundity in g. pelii was low when compared with what was reported in brachyuran species like callinectes sapidus (mean of 3.2 million eggs) (guillory et al. 1996) and portunus validus (mean of 3 million eggs) (moruf and lawal-are 2017b). according to shields et al. (1990), variations in fecundity in crab might be caused by several ecological factors including habitat and biological constraints. the egg diameter of g. pelii ranged from 0.26 to 0.38mm with a mean of 0.30mm. this result compared well with a study on a different species of estuarine crab, by lawal-are (2010) who reported an egg diameter range of 0.25 to 0.35mm with a mean of 0.25mm for callinectes amnicola in the lagos lagoon. 4 conclusions goniopsis pelii exhibits sexual dimorphism with males attaining larger sizes than females. as revealed by allometric growth, the crab weight can be estimated from its length. reproductive biology and conditions of the species are indicators of viable population. the present study has provided baseline data which will serve as a useful source of information in the management and conservation of mangrove ecosystem resources. further studies could assess reproductive activity using macroscopic and microscopic observations of gonad characteristics, trends of gonad indices, size at first sexual maturity, differences in the monthly sex ratio, estimation of gonado-somatic index etc. acknowledgments two anonymous reviewers are acknowledged for valuable comments on the initial draft of the manuscript. references bannister jv. 1976. the length-weight relationship, condition factor, gut contents in the dolphin fish, coryphaena hippurus l. in the mediterranean. journal of fish biology 9: 335-338. cortés e. 1998. methods of studying fish feeding: reply. canadian journal of fisheries and aquatic sciences 55(12): 27-38. dobson m, magana a, mathooko jm., ndegwa fk. 2002. detritivores in kenyan highland streams: more evidence for the paucity of shredders in the tropics? freshwater biology 47: 909-919. emmanuel be. 2009. assessment of crab traps selectivity and efficiency in a tropical riparian swamp. african journal of biotechnology 8 (18): 4680-4684 eteobong n, fola-mathews oo, aghogho kd. 2017. sex ratio of purple lagoon crabs goniopsis pelii (herklots, 1851) from the lagos lagoon – nigeria. journal of biology, agriculture and healthcare 7(9): 6-11. eteobong n, lawal-are ao, bernard e. 2016. diet composition of purple lagoon crab goniopsis pelii (herklots, 1851) from south west nigeria. journal of biology, agriculture and healthcare 6(21): 104112. feller ic, sitnik m. 1996. mangrove ecology. workshop manual.565pp r.o. moruf biology of the purple mangrove crab ruhuna journal of science vol 11 (1): 13-22, june 2020 22 guillory v, prejean e, bourgeois m, burdon j, merrell j. 1996. a biological and fisheries profile of the blue crab, callinectes sapidus. l.a. department of wildlife and fisheries management plan series, 8(1): 210 pp. hill mp, o’keefe jh. 1992. some aspects of the ecology of the freshwater crab (potamonautes perlatus milne edwards) in the upper reaches of the buffalo river, eastern cape province, south africa. south african journal of aquatic sciences 18: 42-50. kumar ms, ferguson g, xiao y, hooper g., venema s. 2000. studies on reproductive biology and distribution of the blue swimmer crab (portunus pelagicus) in south australia waters. sardi research report series no. 47. lawal-are ao, nwankwo h. 2011. biology of the hairy mangrove crab, sersema huzardii (decapoda: grapsidae) from a tropical estuarine lagoon. journal of american science 7 (7): 402-408. lawal-are ao, moruf ro, oluseye-are so, isola to. 2019. antioxidant defense system alternations in four crab species as a bio-indicator of environmental contamination. bulletin uasvm veterinary medicine 76(1): 73-80. lawal-are ao. 2010. reproductive biology of the blue crab, callinectes amnicola in the lagos lagoon, nigeria. turkish journal of fish and aquatic science 10: 1-7. moruf ro, bolaji od, lawal-are ao. 2018. biometrics, gut contents and sexual dimorphism of the west african mud creeper, tympanotonus fuscatus var radula from the mangrove swamps of a coastal estuary in nigeria. egyptian journal of aquatic biology and fisheries 22(1): 8796. moruf ro, lawal-are ao. 2017a. size composition, growth pattern and condition factor of two portunid crabs, callinectes amnicola (de rochebrune) and portunus validus (herklots) from lagos coast, nigeria. nigerian journal of fisheries and aquaculture 5(1): 15 – 21. moruf ro, lawal-are ao. 2017b. comparability in dietary elements, sex ratio and fecundity of portunid crabs, callinectes amnicola (de rochebrune) and portunus validus (herklots) off lagos coast. journal of aquatic sciences 32 (1a): 25 – 31. moruf ro, lawal-are ao. 2018. haemato-biochemical variations in estuarine crabs from a lagoon ecosystem. journal of applied sciences and environmental management 22(12): 1899–19033. moruf ro, ojetayo ta. 2017. biology of the west african fiddler crab, uca tangeri (eydoux, 1835) (decapoda: ocypodidae) from a mangrove wetland in lagos, nigeria. international journal of aquatic biology 5 (4): 263-267. onadeko ab, lawal-are ao, igborgbor os. 2015. habitat diversity and species richness of brachyuran crabs off university of lagos lagoon coast, akoka, nigeria. the bioscientist 3(1): 014028. parson r. 1988. statistical analysisa decision making approach. second edition. harper and row publishers, new york.791pp. pauly d. 1983. some simple methods for the assessment of tropical stocks. fao fish. tech. pap., 234: 52. schneider w. 1990. field guide to commercial marine resources of the gulf of guinea. fao species identification sheets for fishery purposes. 186pp shields jd, okazaki rk, kurts am. 1990. fecundity and the reproductive potential of the yellow rock crab, cancer anthony. fishery bulletin 89: 299-305. usese ai, lawal-are ao, moruf ro, chukwu lo. 2018. biomarker responses to environmental stressors in the hairy mangrove crab, sesarma huzardii (grapsidae) from a tropical lagoon mudflat in nigeria. alexandria journal of veterinary sciences 57 (1): 4-10. venkataraman k. 2007. marine ecosystems of india. indian journal of environmental education 7: 7-26. sv-lncs ruhuna journal of science vol 9(2): 140-149, december 2018 eissn: 2536-8400  faculty of science http://doi.org/10.4038/rjs.v9i2.42 university of ruhuna  faculty of science, university of ruhuna 140 short paper study on effectiveness and user satisfaction of antiinflammatory ointment of curcuma albiflora thw. in bentota divisional secretariat, southern sri lanka h.m.i.c. herath1, t.d.c.m.k. wijayasiriwardene 2*, g.a.s. premakumara2 1faculty of graduate studies, university of colombo, colombo, sri lanka 2industrial technology institute, bauddaloka mawatha, colombo, sri lanka *correspondence: drchandima@iti.lk; orcid: 0000-0002-8308-6748 received: 6th march 2018, revised: 31st august 2018, accepted: 27th november 2018 abstract. under the same sinhalese vernacular name, three plants (curcuma albiflora thw., curcuma zedoaria rosc., and zingiber zerumbet smith) are being reported. amongst them, c. albiflora is an endangered and less explored plant. the current study was conducted to study the anti-inflammatory activity of the ointment produced using c. albiflora. clinical trial was conducted using 175 participants in community based centers (n=2) in grama niladari divisions of bentota divisional secretariat, southern province, sri lanka. participants treated with the ointment (98 patients) or the placebos (77 patients) were asked to complete the structured questionnaire. treatment was applied for two weeks and the results were monitored by two ayurvedic doctors who were involved in the study. comparisons by chi square test showed that tm users differed from non-users in terms of previous side effects and current inflammatory conditions (p<0.05). by binary logistic regression analysis, treatment group satisfaction was found to be about 31 times the chance of the placebo group. the models indicated that the patients above 60 years of age, female, previous traditional medicine users, duration of current anti-inflammatory condition (≥ 1 year) were more likely to effectively respond to the treatment. ointment of c. albiflora thw. showed moderate antiinflammatory activity. keywords. binary logistic regression, curcuma albiflora, survey. 1 introduction arthritis is a major common health problem in the world, while osteoarthritis (3.8%) and rheumatoid arthritis (0.24%) of people are reported globally (richette and bardin 2010). according to survey results of the department of h.m.i.c. herath et al. effectiveness and user satisfaction of curcuma ointment ruhuna journal of science 141 vol 9(2): 140-149, december 2018 census and statistics on sri lankan people, nearly 11.4% were reported to have arthritis. moreover, the prevalence rate of arthritis reported is 1.25% and 2.8% among the males and females respectively (dcs 2014). to date, chemoprevention is a major strategy in controlling inflammatory diseases (gustin et al. 2002). non-steroidal anti-inflammatory drugs (nsaids) were originally used to alleviate pain and for the treatment of arthritis. nsaids lead to a number of side effects in the patients such as pain, gastro-intestinal tract issues, and leading to problems such as high blood pressure (o’leary 2014). therefore, day by day herbal medicine becomes popular throughout the world. but, some of the constraints exist in herbal medicine are adulteration of authentic plant materials with other similar looking plant materials (e.g. curcuma species), legal barriers on cultivation (e.g. kansa canabis sativa) and restrictions to harvest from forests (e.g. venivel coscinium fenestratum, kotala himbutusalasia reticulate, bimkohomba munronia pinnata, sudu handunsantalum album and rath handun pterocarpus santalinus) (kankanamalage et al. 2014). ‘harankaha’ is a medicinal plant used in traditional medicine in sri lanka. but, under the same common trivial name harankaha, three species are reported; curcuma albiflora thw., curucma zedoaria rosc., and zingiber zerumbet smith (dassanayaka 1983). among these three species, c. albiflora is an endangered and poorly explored endemic plant (moe 2012). however, c. zedoaria is one of the most extensively studied species of curcuma due to their high commercial value. proteins of c. zedoaria have shown a significant anti-inflammatory activity (i.e. 77% inhibition after 5h) in the carrageenan induced rat paw oedema model system (angel et al. 2013). according to ullah et al. (2014), ethanol extract of c. zedoaria (500 mg/kg) has shown maximum activity (94.67% inhibition) after 3h on carrageenan induced paw oedema test. moreover, its ethanol extract (500 mg/ml) has significantly inhibited protein denaturation by 77.15%. makabe et al. (2006) reported that furanodiene and furanodienone isolated from c. zedoaria using ethyl acetate (in hexane) showed 75% and 53% inhibitory effect on tpa (12o-tetradecanoylphorbol13-acetate) induced edema of mouse ear test respectively. further, kaushik and jalalpure (2011a) has studied histamine induced rat paw edema test on various extracts of c. zedoaria (root) and have shown a significant inhibition. petroleum ether and chloroform extracts have shown 46.3% and 51.85% inhibition at 200 mg/kg dose after 6h respectively. further, kaushik and jalalpure (2011a) found that petroleum ether and chloroform extracts of c. zedoaria have shown 55.81% and 51.16% inhibition at the dose of 200 mg/kg respectively (on carrageenan induced paw oedema model). but standard group (indomethacin) has shown 53.49% inhibition after 6h. the petroleum ether extract of c. zedoaria reduced edema 56.70% in the carrageenan induced paw oedema test on oral administration 200 mg/kg when compared to control group (kaushik and jalalpure (2011b). however, c. albiflora has not studied to that extent, yet investigation of antih.m.i.c. herath et al. effectiveness and user satisfaction of curcuma ointment ruhuna journal of science 142 vol 9(2): 140-149, december 2018 inflammatory action of c. albiflora is important for traditional medicine in sri lanka. c. albiflora hydro-ethanolic extract significantly impaired the paw oedema, at 1h (by 61%) by 200 mg/kg on carrageenan induced paw oedema model (herath et al. 2017). moreover, the drug group (400 mg/kg) significantly (p<0.05) reduced the paw oedema from the day 5 to 7, when compared with the control (water) on formaldehyde induced paw oedema model. although in vivo and in vitro assays are used to predict human response to drugs (or other chemicals), it is an argumentative issue (shanks et al. 2009). hence, the current study was conducted to study the anti-inflammatory activity of ointment which was produced from c. albiflora using the modified method of liu et al. (2012) and anand and bansal (2016). the ointment was tested through a clinical trial using 175 people, who were selected from community based centers (n=2) in grama niladari divisions of bentota divisional secretariat, southern province, sri lanka. 2 materials and methods 2.1 ethical clearance ethical clearance was obtained from the department of ayurveda in sri lanka to conduct the clinical trial. 2.2 collection of plants and preparation of the extract plants were collected from 2016 to early 2017 in sabaragamuwa province (erathna: n 6o 50' 07'', e 80o 24' 41'', kitulgala: n 6° 59' 41'', e 80° 24' 20'' and bopathella: n 6o 48' 07'', e 80o 22' 12''). voucher specimens (no. 02, 03, 04) of the plants were authenticated from national herbarium, peradeniya, sri lanka. 2.3 preparation of c. albiflora (ca) ointment decoction, sesame oil (2.5 l) as solvent, soft paraffin (5 l), bee-wax were used to prepare ca ointment (n = 100). 2.4 participants and settings a descriptive survey design was used to collect data through a questionnaire about ca ointment. data were collected from the people gathered at the community based centers (n= 2) in dedduwa and mahawila grama niladhari h.m.i.c. herath et al. effectiveness and user satisfaction of curcuma ointment ruhuna journal of science 143 vol 9(2): 140-149, december 2018 dividions in benthota, galle district, southern province on 22/12/2017 and 14/01/2018. 2.5 questionnaire in the present study, a modified version of the questionnaire developed by swisher was used (swisher et al. 2002). a draft questionnaire was prepared, and it was reviewed by ayurvedic doctors (n=2). then a pilot test was performed in two community groups as mentioned in 2.4. the questionnaire used to collect data was attached and it consisted of three parts; the first part on the demographic information of participants (age, gender, educational level, marital status and religions), the second part on participants' clinical condition, use of traditional medicine (tm) (activity of daily life, methods of obtaining information about tm), and the third part on the satisfaction or dissatisfaction on curcuma albiflora (ca) ointment after applying it for two weeks. they were advised to use ointment without using any antiinflammatory drug during these two weeks. tm users were defined as patients who had used tm at least once before the use of ca ointment. 2.6 statistical analysis of the survey data differences in demographic and clinical characteristics between tm users and non-users were assessed using test. factors associated with tm use were identified via binary logistic regression analysis. the analysis provided an odds ratio and a 95% ci for each variable while simultaneously controlling for the effects of other variables. data were analyzed with minitab 17 software. logistic regression is applicable when the response is binary and uses a transformation or link function to convert the binomial data into continuous variable using formula logit(p) = ln( ). the regression output provides the predicted logit value for each observation, which transform to a binary percentage (peng 2002). the event probability (e) is the estimated event probability value for that observation. e = where logit(p) = constant+α1x1+α2x2…+αnxn 3 results and discussion the survey conducted in the community based centers (dedduwa and mahawila) in benthota division, galle district, southern province noted mean h.m.i.c. herath et al. effectiveness and user satisfaction of curcuma ointment ruhuna journal of science 144 vol 9(2): 140-149, december 2018 ages as 62 and 58 for two populations (sd 13). required total sample size was obtained as 166. total of 175 participants attended the survey from the community based centers (n=2) in dedduwa and mahawila grama niladhari, in benthota division (plate 1). plate 1. survey on two community based centres (a: dedduwa grama niladhari, benthota division, galle district, southern province b: mahawila-west grama niladhari, benthota division, galle district, southern province, sri lanka). table 1: demographic and clinical characteristics of the participants. variable category tm users % tm non users % total % chi square value p value age >60 32.6 40 72.6 0.72 ref: 3.84 0.39 <= 60 14.3 13.1 27.4 education illiterate 4 4.6 8.6 2.75 ref: 7.82 0.43 primary 21.1 25.7 46.9 ol 17.1 14.3 31.4 > ol 4.6 8.6 13.1 marital status married 27.4 33.7 61.1 0.44 ref: 5.99 0.80 never 4 4 8 widowed 15.4 15.4 30.9 sex male 10.9 15.4 26.3 0.77 ref: 3.84 0.37 female 36 37.7 73.7 previous side effects free 41.7 53.1 94.8 10.76 ref: 3.84 < 0.001* simple 5.1 0 5.1 food type vegetarian 14.3 17.1 31.4 0.06 ref: 3.84 0.80 mix 32.6 36 68.6 current inflammatory condition never 0 4 4 6.42 ref: 5.99 0.01* slight 46.3 49.1 95.4 duration of current inflammatory condition <1 year 21.7 29.1 50.9 1.75 ref: 5.99 0.41 1-5 years 19.4 20 39.4 >5 years 5.7 4 9.7 a b h.m.i.c. herath et al. effectiveness and user satisfaction of curcuma ointment ruhuna journal of science 145 vol 9(2): 140-149, december 2018 out of these 175 people, 98 (56%) were eligible for receiving ca ointment. after screening past diagnosis, current health status of patients were assessed by ayurvedic doctors (n=2). participants (who have chronic diseases such as diabetes, kidney diseases etc.) and healthy participants were recommended to obtain placebo. patients with arthritis and joint disorders have received the ca ointment. of the 175 surveyed people consisting of 46 (26.3%) males and 129 (73.7%) females completed the survey with a response rate of 100%. their mean age was 64.6 ± 11.8 years; most participants were married (n = 107; 61.1%) and were widowed (n=54; 30.9%). demographic and clinical characteristics of participants were summarized in table 1. nine patients (5.1%) reported side-effects of the tm drugs such as traditional mixtures called ‘arishta’ and ‘kalka’. these side-effects included gastric upset, stomachache and increase of phlegm, but nobody experienced serious sideeffects. 46.9% (n = 82) of the total responders were tm users (patients who had used tm at least once before the use of ca ointment). comparisons by chi square test showed that tm users differed from non-users in terms of previous side effects and current inflammatory condition (p<0.05). tm users did not differ from non-users in terms of age, education, marital status, sex, food type, and duration of current inflammatory condition (p>0.05). effect of the ca ointment and placebo were reported in table 2. table 2: effect of treatment from surveyed result of curcuma albiflora (ca) ointment and placebo. exposure outcome odd ratio improved no improvement treatment group (ca ointment) 74 24 31 placebo group 07 70 odd ratio equals to 31 indicated the satisfaction of the treatment group who used ca ointment for two weeks is about 31 times that of the placebo group. according to response of the ca ointment users, results were summarized in table 3. ca ointment was received by 56% (n = 98) of the surveyed group. comparisons by chi square test showed that ca satisfied responders did not differ from dissatisfied in terms of age, education, food-type, sex, and the duration of current inflammatory condition (p>0.05). some of the factors were studied in detail and reported in table 4. h.m.i.c. herath et al. effectiveness and user satisfaction of curcuma ointment ruhuna journal of science 146 vol 9(2): 140-149, december 2018 table 3: bivariate analysis results by survey on satisfaction/dissatisfaction of curcuma albiflora ointment. variable category satisfied % dissatisfied % total % chisquare p-value age < 60y 25.5 5.0 30.6 1.97 0.16 >= 60y 64.3 5.0 69.4 education < ol 41.8 7.1 49.0 1.96 0.16 >= ol 37.8 13.3 51.0 sex male 16.3 8.2 24.5 1.34 0.24 female 59.2 16.3 75.5 food type vegetarian 23.4 6.1 29.6 2.73 0.09 mix 0.64 6.1 70.4 duration of current inflammatory condition <1 year or >5 0.32 6.1 37.8 0.18 0.67 1-5 years 0.54 8.2 62.2 table 4: factors associated with curcuma albiflora ointment users in the binary logistic regression. variable category response test or treatment placebo age >60 years improved 50 6 24.5 no improvement 18 53 <60 years improved 24 1 68 no improvement 6 17 sex male improved 16 3 12.6 no improvement 8 19 female improved 58 4 46.2 no improvement 16 51 previous tm use tm users improved 35 1 102.1 no improvement 12 35 tm non users improved 39 6 19.0 no improvement 12 35 duration of current inflammatory condition <1 year improved 19 5 12.4 no improvement 8 26 1-5 or >5 years improved 54 2 69.9 no improvement 17 44 h.m.i.c. herath et al. effectiveness and user satisfaction of curcuma ointment ruhuna journal of science 147 vol 9(2): 140-149, december 2018 the binary logistic models indicated that above 60 years of age, female, previous traditional medicine users, duration of current anti-inflammatory condition (≥ 1 y) were more likely to effectively response for ca ointment. according to binary logistic analysis number of satisfied ca ointment users were reported as 74, and dissatisfied as 24. predictor variable values were summarized in table 5. table 5. predictor variable values of curcuma albiflora ointment users by binary logistic regression analysis (tm= traditional medicine). predictor variable β s.e. β wald z p exp (β) or 95% ci for exp (β) lower upper age -0.02 0.02 -0.81 0.420 0.98 0.93 1.03 sex 0.44 0.58 0.76 0.445 1.56 0.50 4.92 marital status 0.09 0.32 0.30 0.762 1.10 0.58 2.10 edu1tional level -0.57 0.31 -1.84 0.06 0.56 0.31 1.04 food -0.59 0.60 -0.98 0.32 0.55 0.17 1.81 previous tm user -0.39 0.53 -0.74 0.46 0.67 0.23 1.93 diagnosis/months 0.00 0.00 0.75 0.45 1.00 0.99 1.02 side effects 0.20 1.25 0.16 0.87 1.23 0.11 14.31 arthritic condition 0.22 1.37 0.17 0.86 1.26 0.08 18.64 many researches have revealed effectiveness of curcumin incorporated skin formulations. gonçalves et al. (2014) has assessed topical formulations of curcuminoid extracted from curcuma longa by quantifying skin retention (using pig ear membrane) and skin permeation (modified franz diffusion cell system). they have prepared various gel formulations by water and hydroxyethylcellulose (at 70°c); the most stable formulation was βcyclodextrin containing hydro-ethanolic and dry extract of turmeric. although skin penetration lowered by curcumin, skin retention promoted. therefore, curcumin is used in anti-aging formulations. gonçalves et al. (2014) has reported that, a tonic prepared from the rhizome of c. longa has shown significant reduction in ulcers, cough, and the common cold as well as ointment form increased wound healing. furthermore, hamzah (2011) has found that turmeric gel reduced a 30% inhibition in oedema using the carrageenan-induced rat paw edema model. huang et al. (1997) found that curcumin inhibited induced skin tumor formation by topical application. lin and lin (2008) showed that curcumin prevents skin tumors, spots or wrinkles effectively. kuptniratsaikul et al. (2009) performed a single-blind randomized controlled trial to evaluate the efficacy of c. domestica extracts (2,000 mg/day) compared with ibuprofen (800 mg/day) of in 107 knee oa patients h.m.i.c. herath et al. effectiveness and user satisfaction of curcuma ointment ruhuna journal of science 148 vol 9(2): 140-149, december 2018 for 6 weeks. further, to determine the efficacy and safety of c. domestica extracts (1,500 mg/day) in pain reduction and functional improvement compared with ibuprofen (1,200 mg/day); about 96% were satisfied with the treatment while fewer gastrointestinal issues reported in the c. domestica extracts group. patient’s assessment of satisfaction on c. domestica extract (n=171) at week 4, satisfied and unsatisfied patients were about 97% and 1% respectively (p > 0.05). fig. 1. diagnostic plot of delta chi-square versus probability log-likelihood and g value were reported as -50.436 and 8.232 respectively. goodness-of-fit tests under the method pearson was shown higher chisquare value (109.854) indicates enough number of data for the analysis. this is supported with the figure 1. hydro-ethanolic extract of c. albiflora significantly impaired the paw oedema, at 1h (by 61%) by 200 mg/kg on carrageenan induced paw oedema model as well as from the day 5 to 7 on formaldehyde induced paw oedema model (herath et al., 2017). from the current study it was revealed that the treatment group satisfaction was about 31 times the chance of the placebo group. therefore, ca ointment showed moderate anti-inflammatory activity. 4 conclusion from the present findings it can be concluded that curcuma albiflora ointment possessed a moderate anti-inflammatory activity. acknowledgment national science foundation provided financial support (nsf/sch/2018/03). commissioner of ayurveda, southern province, sri lanka is acknowledged for probability d e lt a c h is q u a r e 1.00.80.60.40.20.0 25 20 15 10 5 0 delta chi-square versus probability h.m.i.c. herath et al. effectiveness and user satisfaction of curcuma ointment ruhuna journal of science 149 vol 9(2): 140-149, december 2018 providing ethical clearance. comments from two anonymous reviewers were acknowkedged. references anand a, bansal g. 2016. predicting customer’s satisfaction (dissatisfaction) using logistic regression. international journal of mathematical, engineering and management sciences 1(2): 77–88. angel gr, vimala b, nambisan b. 2013. antioxidant and anti-inflammatory activities of proteins isolated from eight curcuma species. phytopharmacology 4(1): 96-105. dassanayaka md. 1983. flora of ceylon, vol iv. amerind publishing co. pvt. ltd, new delhi 495 – 496, 501-504. dcs. 2014. national survey on self-reported health in sri lanka 2014. department of census and statistics. sri lanka: ministry of national policies and economic affairs. gustin dm, brenner de. 2002. chemoprevention of colon cancer: current status and future prospects. cancer and metastasis reviews 21:323-348. hamzah mm. 2011. evaluation of topical preparations containing curcuma, acacia and lupinus extracts as an anti-inflammatory drugs. international journal of applied research in natural products 4(2): 19-23. herath i, premakumara s, wijayasiriwardene c. 2017. anti-inflammatory activity of curcuma albiflora thw. grown in sri lanka. journal of ayurveda medical sciences 2(4): 90-93. huang mt, yen p, xie j, han j, frenkel j. 1997. inhibitory effects of topical application of low doses of curcumin on 12-o-tetradecanoylphorbol-13-acetateinduced tumor promotion and oxidized dna bases in mouse epidermis. carcinogenesis 18(1): 83-88. kankanamalage tn, dharmadasa rm, abeysinghe dc, wijesekara rg. 2014. a survey on medicinal materials used in traditional systems of medicine in sri lanka. journal of ethnopharmacology 155 (1):679–91. kaushik ml, jalalpure ss. 2011. anti-inflammatory efficacy of curcuma zedoaria root extracts. asian journal of pharmaceutical and clinical research 4(3):1-5. kaushik ml, jalalpure ss. 2011. effect of curcuma zedoaria rosc root extracts on behavioral and radiology changes in arthritic rats. journal of advance pharmaceutical technology & reaserch 2(3): 170-176. kuptniratsaikul v, thanakhumtorn s, chinswangwatanakul p, wattanamongkonsil l, visanu t. 2009. efficacy and safety of curcuma domestica extracts in patients with knee osteoarthritis. journal of alternative complementary medicine 15(8):891-897. lin cl, lin jk. 2008. curcumin: a potential cancer chemopreventive agent through suppressing nf-ıb signaling. journal of cancer molecules 4(1): 11-16. makabe h, maru n, kuwabara a, kamo t, hirota m. 2006. anti-inflammatory sesquiterpenes from curcuma zedoaria. natural product research 20(7): 680-685. moe. 2012. the national red list 2012 of sri lanka, conservation status of the fauna and flora. sri lanka: ministry of environment, 339p. o’leary mr. 2014. dr. thomas addison 1795-1860. usa: iuniverse. peng cyj. 2002. an introduction to logistic regression analysis and reporting. the journal of educational research, 3-14. richette, p, bardin t. 2010. gout. lancet 375(9711):318. shanks n, greek r, greek j. 2009. are animals models predictive for humans?. philosophy, ethics and humanities in medicine 4: 2. swisher, em, cohn de, goff ba. 2002. use of complementary and alternative medicine among women with gynecologic cancers. gynecologic oncology 84(3): 363–367. ullah hma, zaman s, juhara f, et al. 2014. evaluation of antinociceptive, in-vivo & in-vitro anti-inflammatory activity of ethanolic extract of curcuma zedoaria rhizome. bmc complementary alternative medicine 14(1): 1-12. ruhuna journal of science vol 11 (1): 59-70, june 2020 eissn: 2536-8400 © faculty of science doi: http://doi.org/10.4038/rjs.v11i1.87 university of ruhuna © faculty of science, university of ruhuna 59 sri lanka zagreb topological indices for hexadentate 3hydroxypyridinones-terminated dendrimers used in iron binding and anti-microbial activities thayamathy pio jude1, elango panchadcharam1*, koneswaran masilamani2 1department of mathematics, faculty of science, eastern university, sri lanka, sri lanka. 2department of chemistry, faculty of science, eastern university, sri lanka, sri lanka. *correspondence: elangop@esn.ac.lk; https://orcid.org/0000-0002-7359-989x received: 03rd march 2019, revised: 29th may 2020, accepted: 25th june 2020 abstract. among the degree based topological indices which are mostly used in the study of chemical graph theory, the zagreb index plays a prominent role in the study of therapies and treatments. there were different versions of the zagreb indices defined in the literature and used for different purposes. in this paper, we calculated the different versions of the zagreb indices for the hexadentate 3hydroxypyridinones-terminated dendrimers which are used in iron binding and anti-microbial activities. these indices may play an important role in determining properties of these compounds. keywords: anti-microbial activities, dendrimers, iron binding, topological index, zagreb index. 1 introduction iron performs a crucial role in the growth and development of living systems. therefore, it is one of the most vital trace elements in biological systems (weizman et al. 1996). iron is an essential element in hemoglobin, myoglobin, cytochromes and important in many biological processes (haas and browlie 2001, zhang and enns 2008, evstatiev et al. 2012). however, both iron deficiency and the excessive iron in the body are harmful to human health. excessive iron in the body is usually removed by iron-specific chelators (saliba et al. 2016). of the few iron specific chelators, macromolecular iron chelators play a major role in the treatment in removing both chronic and acute excessive iron in the body (mahoney et al. 1989). dendrimers are a class of synthetic macromolecules with highly branched structures which built up by repeating the assembly of the constituent layers or generations. these structures have a central core and different functional groups attached to their outer surface (zhu and shi 2013). these dendrimer structures are used as iron chelators for therapeutic purposes. these iron chelators bind to the iron and to the complex which are kinetically inert and non-toxic. furthermore, these complexes are not absorbed by the https://creativecommons.org/licenses/by-nc/4.0/ https://orcid.org/0000-0002-7359-989x https://orcid.org/0000-0002-7359-989x t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 60 gastrointestinal tract of the human due to their large size so will be excreted from the human body. furthermore, the macromolecular iron chelators inhibit the growth of bacteria hence are used in the treatment of wound healings (zhou et al. 2011). zhou et. al synthesized hexadentate 3-hydroxypyridinones-terminated dendrimers that could be used for iron bindings and anti-bacterial purposes (zhou et al. 2018). chemical graph theory is an important tool for studying molecular structures. in theoretical chemistry, molecular descriptors, especially topological indices are used for modelling information of molecules which are used in biology and pharmacology. therefore, determining topological indices of dendrimers will give a suitable correlation between chemical structure and its activity (farahani et al. 2015, soleimani et al. 2016). in this work, different versions of the zagreb indices are calculated for the hexadentate 3-hydroxypyridinones-terminated dendrimers which are used in iron binding and anti-microbial activities. a molecular graph is a simple graph which has neither loops nor multiple edges. the atoms and the chemical bonds between them in molecular graphs, are represented by vertices and edges respectively in the graph theory. a graph g = g(v, e) is a pair of nonempty set of vertices v = v(g) and the set of connected edges e = e(g) if there exists a connection between any pair of vertices in g. the degree dv(g) or dv of a vertex v in the set of vertices v(g) in a connected graph g is the number of vertices which are connected to that vertex by the edges. the concept of degree is closely related to the concept of valence bond in chemistry (gutman 2013). a topological index is a numerical quantity which is derived mathematically in a direct and unambiguous manner from the structural graph of a molecule. the topological index is defined to be a function top: g(v, e) → 𝑅, where 𝑅 is the set of real numbers. if two graphs are isomorphic then their topological indices are same. many properties of a chemical compound are closely related to some of the topological indices of its molecular graph. the wiener index is the first and the most studied distance-based topological index in chemical graph theory. this index was introduced by the chemist h. wiener (wiener 1947) in 1947 to demonstrate the correlations between physicochemical properties of organic compounds and the topological structure of their molecular graphs. the wiener index was defined as the sum of distances between all the carbon atoms in the molecules, in terms of carbon-carbon bonds. among the different types of topological indices, the degree-based topological indices are the mostly studied type of topological indices, which play a prominent role in chemical graph theory. one of the oldest degree-based topological indices is the well-known zagreb index which was introduced by gutman and trinajstić (1972) during the analysis of the structure-dependency of total π-electron energy. currently researchers are interested in calculating the zagreb indices for different types of molecular structures which are used in various applications (aslam et al. 2017, kana et al. 2018, jude et al. 2019). t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 61 2 methods in this paper we considered four types of the hexadentate 3-hydroxypyridinonesterminated dendrimers which are used in iron binding and anti-microbial activities. zhou et al. 2018 shows that the convergent synthesis of a range of novel hexadentate 3-hydroxypyridinone-terminated dendrimers. initially, first generation dendrimeric chelators were synthesized and as its structure shown in figure 1 which contain three hexadentate moieties. the second generation of this dendrimeric chelators was synthesised by the same authors and its structure is shown in figure 2. second generation dendrimer was respectively conjugated with di-acid and tri-acid to form protected structures in figure 3 and figure 4 which contain six and nine hexadentate centres respectively. we calculated four different versions of the zagreb indices using the following formulas. the first zagreb index (gutman and trinajstić 1972) was defined as m1(g) = ∑ (dv) 2 v∈v(g) = ∑ (du + dv) uv∈e(g) and the second zagreb index (gutman and trinajstić 1972) was defined as m2(g) = ∑ dudv uv∈e(g) the second modified zagreb index (hao 2011) was defined as m2(g) m = ∑ 1 dudv uv∈e(g) the reduced second zagreb index which appeared in the literature earlier but so far not been studied in mathematical chemistry was re-presented in the paper (gutman et al. 2015) with its main mathematical properties. this was defined as rm2(g) = ∑ (du − 1)(dv − 1) uv∈e(g) this is of the difference between the two zagreb indices m1(g) and m2(g) and if the graph g is a tree, then this is equal to the number of pairs of vertices at distance 3, which is often referred to as the “wiener polarity index” in mathematical chemistry (gutman et al. 2014). 3 results and discussion by observing the structure of synthetic route of the first generation dendrimeric chelators, we inferred six partitions of the edge set as: t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 62 𝐸1(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 1 𝑎𝑛𝑑 𝑑𝑣 = 2}, 𝐸2(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 1 𝑎𝑛𝑑 𝑑𝑣 = 3}, 𝐸3(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 𝑑𝑣 = 2}, 𝐸4(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 2 𝑎𝑛𝑑 𝑑𝑣 = 3}, 𝐸5(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 2 𝑎𝑛𝑑 𝑑𝑣 = 4}, 𝐸6(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 𝑑𝑣 = 3}. also, we get |𝐸1(𝐺)| = 9, |𝐸2(𝐺)| = 33, |𝐸3(𝐺)| = 42, |𝐸4(𝐺)| = 69, |𝐸5(𝐺)| = 12, |𝐸6(𝐺)| = 27. fig. 1. structure of the first generation dendrimeric chelators (zhou et al. 2018) theorem 1. let g be the first generation dendrimeric chelators. then the first zagreb index 𝑀1(𝐺), the second zagreb index 𝑀2(𝐺), the second modified zagreb index 𝑀2(𝐺) 𝑚 and the reduced second zagreb index 𝑅𝑀2(𝐺) for g are 1. 𝑀1(𝐺) = 906 2. 𝑀2(𝐺) = 1038 3. 𝑀2(𝐺) 𝑚 = 42 4. 𝑅𝑀2(𝐺) = 324 t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 63 proof: using this edge partition of the first generation dendrimeric chelators and by the respective formulas of different versions of the zagreb indices, we get 1. 𝑀1(𝐺) = ∑ (𝑑𝑢 + 𝑑𝑣 )𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 + 2) + |𝐸2(𝐺)|(1 + 3) + |𝐸3(𝐺)|(2 + 2) + |𝐸4(𝐺)|(2 + 3) + |𝐸5(𝐺)|(2 + 4) + |𝐸6(𝐺)|(3 + 3) = 9 × 3 + 33 × 4 + 42 × 4 + 69 × 5 + 12 × 6 + 27 × 6 = 906 2. 𝑀2(𝐺) = ∑ (𝑑𝑢 × 𝑑𝑣 )𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 × 2) + |𝐸2(𝐺)|(1 × 3) + |𝐸3(𝐺)|(2 × 2) + |𝐸4(𝐺)|(2 × 3) + |𝐸5(𝐺)|(2 × 4) + |𝐸6(𝐺)|(3 × 3) = 9 × 2 + 33 × 3 + 42 × 4 + 69 × 6 + 12 × 8 + 27 × 9 = 1038 3. 𝑀2(𝐺) 𝑚 = ∑ 1 𝑑𝑢𝑑𝑣 𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)| 1 1 × 2 + |𝐸2(𝐺)| 1 1 × 3 + |𝐸3(𝐺)| 1 2 × 2 + |𝐸4(𝐺)| 1 2 × 3 + |𝐸5(𝐺)| 1 2 × 4 + |𝐸6(𝐺)| 1 3 × 3 = 9 2 + 33 3 + 42 4 + 69 6 + 12 8 + 27 9 = 42 4. 𝑅𝑀2(𝐺) = ∑ (𝑑𝑢 − 1)(𝑑𝑣 − 1)𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 − 1)(2 − 1) + |𝐸2(𝐺)|(1 − 1)(3 − 1) + |𝐸3(𝐺)|(2 − 1)(2 − 1) + |𝐸4(𝐺)|(2 − 1)(3 − 1) + |𝐸6(𝐺)|(2 − 1)(4 − 1) + |𝐸5(𝐺)|(3 − 1)(3 − 1) = 9 × 0 + 33 × 0 + 42 + 69 × 2 + 12 × 3 + 27 × 4 = 324 remark: .by observing the structure of the second generation dendrimeric chelators, we inferred six partitions of the edge set as: 𝐸1(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 1 𝑎𝑛𝑑 𝑑𝑣 = 2} 𝐸2(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 1 𝑎𝑛𝑑 𝑑𝑣 = 3} 𝐸3(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 2 𝑎𝑛𝑑 𝑑𝑣 = 3} 𝐸4(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 𝑑𝑣 = 2} 𝐸5(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 2 𝑎𝑛𝑑 𝑑𝑣 = 4} 𝐸6(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 𝑑𝑣 = 3} t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 64 also, we get |𝐸1(𝐺)| = 10, |𝐸2(𝐺)| = 25, |𝐸3(𝐺)| = 100, |𝐸4(𝐺)| = 95, |𝐸5(𝐺)| = 16, |𝐸6(𝐺)| = 27 fig. 2: second generation dendrimeric chelators (zhou et al. 2018). theorem 2. let 𝐺 be the second generation dendrimeric chelators. then the first zagreb index 𝑀1(𝐺), the second zagreb index 𝑀2(𝐺), the second modified zagreb index 𝑀2(𝐺) 𝑚 and the reduced second zagreb index 𝑅𝑀2(𝐺) for 𝐺 are 1. 𝑀1(𝐺) = 1263 2. 𝑀2(𝐺) = 1436 3. 𝑀2(𝐺) 𝑚 = 355 6 4. 𝑅𝑀2(𝐺) = 446 proof: using this edge partition of the second generation dendrimeric chelators and by the respective formulas of different versions of the zagreb indices, we get t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 65 1. 𝑀1(𝐺) = ∑ (𝑑𝑢 + 𝑑𝑣 )𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 + 2) + |𝐸2(𝐺)|(1 + 3) + |𝐸3(𝐺)|(2 + 2) + |𝐸4(𝐺)|(2 + 3) + |𝐸5(𝐺)|(2 + 4) + |𝐸6(𝐺)|(3 + 3) = 10 × 3 + 25 × 4 + 100 × 4 + 95 × 5 + 16 × 6 + 27 × 6 = 1263 2. 𝑀2(𝐺) = ∑ (𝑑𝑢 × 𝑑𝑣 )𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 × 2) + |𝐸2(𝐺)|(1 × 3) + |𝐸3(𝐺)|(2 × 2) + |𝐸4(𝐺)|(2 × 3) + |𝐸5(𝐺)|(2 × 4) + |𝐸6(𝐺)|(3 × 3) = 10 × 2 + 25 × 3 + 100 × 4 + 95 × 6 + 16 × 8 + 27 × 9 = 1436 3. 𝑀2(𝐺) 𝑚 = ∑ 1 𝑑𝑢𝑑𝑣 𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)| 1 1 × 2 + |𝐸2(𝐺)| 1 1 × 3 + |𝐸3(𝐺)| 1 2 × 2 + |𝐸4(𝐺)| 1 2 × 3 + |𝐸5(𝐺)| 1 2 × 4 + |𝐸6(𝐺)| 1 3 × 3 = 10 2 + 25 3 + 100 4 + 95 6 + 16 8 + 27 9 = 355 6 4. 𝑅𝑀2(𝐺) = ∑ (𝑑𝑢 − 1)(𝑑𝑣 − 1)𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 − 1)(2 − 1) + |𝐸2(𝐺)|(1 − 1)(3 − 1) + |𝐸3(𝐺)|(2 − 1)(2 − 1) + |𝐸4(𝐺)|(2 − 1)(3 − 1) + |𝐸6(𝐺)|(2 − 1)(4 − 1) + |𝐸5(𝐺)|(3 − 1)(3 − 1) = 10 × 0 + 25 × 0 + 100 + 95 × 2 + 16 × 3 + 27 × 4 = 446 fig. 3. dendrimeric chelators containing six hexadentate centres (zhou et al. 2018) t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 66 by observing the structure of dendrimeric chelators contain six hexadentate centres, we inferred six partitions of the edge set as: 𝐸1(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 1 𝑎𝑛𝑑 𝑑𝑣 = 2}, 𝐸2(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 1 𝑎𝑛𝑑 𝑑𝑣 = 3} 𝐸3(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 𝑑𝑣 = 2}, 𝐸4(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 2 𝑎𝑛𝑑 𝑑𝑣 = 3} 𝐸5(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 2 𝑎𝑛𝑑 𝑑𝑣 = 4}, 𝐸6(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 𝑑𝑣 = 3}. also, we get |𝐸1(𝐺)| = 18, |𝐸2(𝐺)| = 72, |𝐸3(𝐺)| = 118, |𝐸4(𝐺)| = 146, |𝐸5(𝐺)| = 32, |𝐸6(𝐺)| = 56. fig. 4. dendrimeric chelators containing nine hexadentate centres (zhou et al. 2018) t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 67 theorem 3. let 𝐺 be the dendrimeric chelators contain six hexadentate centres. then the first zagreb index 𝑀1(𝐺), the second zagreb index 𝑀2(𝐺), the second modified zagreb index 𝑀2(𝐺) 𝑚 and the reduced second zagreb index 𝑅𝑀2(𝐺) for 𝐺 are 1. 𝑀1(𝐺) = 2072 2. 𝑀2(𝐺) = 2360 3. 𝑀2(𝐺) 𝑚 = 1747 18 4. 𝑅𝑀2(𝐺) = 730 proof: using this edge partition of the dendrimeric chelators contain six hexadentate centres and by the respective formulas of different versions of the zagreb indices, we get 1. 𝑀1(𝐺) = ∑ (𝑑𝑢 + 𝑑𝑣 )𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 + 2) + |𝐸2(𝐺)|(1 + 3) + |𝐸3(𝐺)|(2 + 2) + |𝐸4(𝐺)|(2 + 3) + |𝐸5(𝐺)|(2 + 4) + |𝐸6(𝐺)|(3 + 3) = 18 × 3 + 72 × 4 + 118 × 4 + 146 × 5 + 32 × 6 + 56 × 6 = 2072 2. 𝑀2(𝐺) = ∑ (𝑑𝑢 × 𝑑𝑣 )𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 × 2) + |𝐸2(𝐺)|(1 × 3) + |𝐸3(𝐺)|(2 × 2) + |𝐸4(𝐺)|(2 × 3) + |𝐸5(𝐺)|(2 × 4) + |𝐸6(𝐺)|(3 × 3) = 18 × 2 + 72 × 3 + 118 × 4 + 146 × 6 + 32 × 8 + 56 × 9 = 2360 3. 𝑀2(𝐺) 𝑚 = ∑ 1 𝑑𝑢𝑑𝑣 𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)| 1 1 × 2 + |𝐸2(𝐺)| 1 1 × 3 + |𝐸3(𝐺)| 1 2 × 2 + |𝐸4(𝐺)| 1 2 × 3 + |𝐸5(𝐺)| 1 2 × 4 + |𝐸6(𝐺)| 1 3 × 3 = 18 2 + 72 3 + 118 4 + 146 6 + 32 8 + 56 9 = 1747 18 4. 𝑅𝑀2(𝐺) = ∑ (𝑑𝑢 − 1)(𝑑𝑣 − 1)𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 − 1)(2 − 1) + |𝐸2(𝐺)|(1 − 1)(3 − 1) + |𝐸3(𝐺)|(2 − 1)(2 − 1) + |𝐸4(𝐺)|(2 − 1)(3 − 1) + |𝐸6(𝐺)|(2 − 1)(4 − 1) + |𝐸5(𝐺)|(3 − 1)(3 − 1) = 18 × 0 + 72 × 0 + 118 + 146 × 2 + 32 × 3 + 56 × 4 = 730 t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 68 remark by observing the structure of dendrimeric chelators contain nine hexadentate centres, we inferred six partitions of the edge set as: 𝐸1(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 1 𝑎𝑛𝑑 𝑑𝑣 = 2}, 𝐸2(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 1 𝑎𝑛𝑑 𝑑𝑣 = 3} 𝐸3(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 𝑑𝑣 = 2}, 𝐸4(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 2 𝑎𝑛𝑑 𝑑𝑣 = 3} 𝐸5(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 2 𝑎𝑛𝑑 𝑑𝑣 = 4}, 𝐸6(𝐺) = {𝑒 = 𝑢𝑣𝜖𝐸(𝐺): 𝑑𝑢 = 𝑑𝑣 = 3}. also, we get |𝐸1(𝐺)| = 27, |𝐸2(𝐺)| = 108, |𝐸3(𝐺)| = 174, |𝐸4(𝐺)| = 219, |𝐸5(𝐺)| = 48, |𝐸6 (𝐺)| = 84. theorem 4. let 𝐺 be the dendrimeric chelators contain nine hexadentate centres. then the first zagreb index 𝑀1(𝐺), the second zagreb index 𝑀2(𝐺), the second modified zagreb index 𝑀2(𝐺) 𝑚 and the reduced second zagreb index 𝑅𝑀2(𝐺) for 𝐺 are 1. 𝑀1(𝐺) = 3096 2. 𝑀2(𝐺) = 3432 3. 𝑀2(𝐺) 𝑚 = 869 6 4. 𝑅𝑀2(𝐺) = 1092 proof: using this edge partition of the dendrimeric chelators contain nine hexadentate centres and by the respective formulas of different versions of the zagreb indices, we get 1. 𝑀1(𝐺) = ∑ (𝑑𝑢 + 𝑑𝑣 )𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 + 2) + |𝐸2(𝐺)|(1 + 3) + |𝐸3(𝐺)|(2 + 2) + |𝐸4(𝐺)|(2 + 3) + |𝐸5(𝐺)|(2 + 4) + |𝐸6(𝐺)|(3 + 3) = 27 × 3 + 108 × 4 + 174 × 4 + 219 × 5 + 48 × 6 + 84 × 6 = 3096 2. 𝑀2(𝐺) = ∑ (𝑑𝑢 × 𝑑𝑣 )𝑢𝑣∈𝐸(𝐺) t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 69 = |𝐸1(𝐺)|(1 × 2) + |𝐸2(𝐺)|(1 × 3) + |𝐸3(𝐺)|(2 × 2) + |𝐸4(𝐺)|(2 × 3) + |𝐸5(𝐺)|(2 × 4) + |𝐸6(𝐺)|(3 × 3) = 27 × 2 + 108 × 3 + 174 × 4 + 219 × 6 + 48 × 8 + 84 × 9 = 3432 3. 𝑀2(𝐺) 𝑚 = ∑ 1 𝑑𝑢𝑑𝑣 𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)| 1 1 × 2 + |𝐸2(𝐺)| 1 1 × 3 + |𝐸3(𝐺)| 1 2 × 2 + |𝐸4(𝐺)| 1 2 × 3 + |𝐸5(𝐺)| 1 2 × 4 + |𝐸6(𝐺)| 1 3 × 3 = 27 2 + 108 3 + 174 4 + 219 6 + 48 8 + 84 9 = 869 6 4. 𝑅𝑀2 (𝐺) = ∑ (𝑑𝑢 − 1)(𝑑𝑣 − 1)𝑢𝑣∈𝐸(𝐺) = |𝐸1(𝐺)|(1 − 1)(2 − 1) + |𝐸2(𝐺)|(1 − 1)(3 − 1) + |𝐸3(𝐺)|(2 − 1)(2 − 1) + |𝐸4(𝐺)|(2 − 1)(3 − 1) + |𝐸5(𝐺)|(2 − 1)(4 − 1) + +|𝐸6(𝐺)|(3 − 1)(3 − 1) = 27 × 0 + 108 × 0 + 174 + 219 × 2 + 48 × 3 + 84 × 4 = 1092 4 conclusions in this work, we considered, hexadentate 3-hydroxypyridinones-terminated dendrimers which are used in iron binding and anti-microbial activities. we calculated the four different versions of the zagreb indices of these four novel hexadentate 3hydroxypyridinone-terminated dendrimers which have been demonstrated to have a high affinity towards iron (iii) ions and have the antimicrobial property. these indices are calculated by using the edge set partitions of these structures of dendrimers. acknowledgments we acknowledge prof. robert c. hider and his research team for granting permission to reproduce the figures in zhou et al. 2018. two anonymous reviewers are acknowledged for comments on the initial draft of the manuscript. references aslam a, bashir y, ahmad s, gao w. 2017. on topological indices of certain dendrimer structures. zeitschrift für naturforschung 72: 559–566. evstatiev r, gasche c. 2012. iron sensing and signalling. gut 61: 933–952. farahani mr., gao w, kanna rmr. 2015. the connective eccentric index for an infinite family of dendrimers. indian journal of fundamental and applied life sciences 5 (s4): 766-771. gutman i. 2013. degree-based topological indices. croatica chemical acta 86: 351-361. gutman i, furtula b, elphick c. 2014. three new/old vertex–degree–based topological indices. matchcommunications in mathematical and computational chemistry 72: 617-632. t. pio jude et al. zagreb indices for 3-hydroxypyridinones-terminated dendrimers ruhuna journal of science vol 11 (1): 59-70, june 2020 70 gutman i, trinajstić n. 1972. graph theory and molecular orbitals total f-electron energy of alternant hydrocarbons. chemical physics letters 77: 535-538. haas jd, browlie iv t. 2001. iron deficiency and reduced work capacity: a critical review of the research to determine a causal relationship. journal of nutrition 131: 676s–690s. hao j. 2011. theorems about zagreb indices and modified zagreb indices. matchcommunications in mathematical and computational chemistry 65: 659-670. jude tp, panchadcharam e, masilamani k. 2019. computation of zagreb and randić indices of two biodegradable dendrimers used in cancer therapy. ceylon journal of science 48: 359-366. kang sm., zahid ma., virk ar., nazeer w, gao w. 2018. calculating the degree-based topological. indices of dendrimers, open chemistry 61: 681-688. mahoney jr, hallaway pe, hedlund be, eaton jw. 1989. acute iron poisoning – rescue with macromolecular chelators. journal of clinical investigation 84: 1362–1366. saliba an, el rassi f, taher at. 2016. clinical monitoring and management of complications related to chelation therapy in patients with beta-thalassemia. expert rev hematol. 9, 151–168. soleimani n, mohseni e, maleki n. 2016. connectivity indices of some famous dendrimers. journal of chemical and pharmaceutical research. 8(8): 229-235. weizman h, ardon o, mester b, libman j, dwir o, hadar y, chen y, shanzer a. 1996. fluorescentlylabeled ferrichrome analogs as probes for receptor-mediated, microbial iron uptake. journal of the american chemical society 118: 12368–12375. wiener hj. 1947. structural determination of paraffin boiling points, j. amer. chem. soc. 69: 17-20. zhang as, enns ca. 2008. iron homeostasis: recently identified proteins provide insight into novel control mechanisms. journal of biological chemistry 284: 711–715. zhou t, chen k, kong l-m, liu m-s, ma y-m, xie y-y, hider rc. 2018. synthesis, iron binding and antimicrobial properties of hexadentate 3-hydroxypyridinones-terminated dendrimers. bioorganic & medicinal chemistry letters 28: 2504–2512. zhou t, winkelmann g, dai zy, hider rc. 2011. design of clinically useful macromolecular iron chelators. journal of pharmacy and pharmacology 63: 893–903. zhu j, shi x. 2013. dendrimer-based nanodevices for targeted drug delivery applications. journal of materials chemistry b 134: 4199–4211. ruhuna journal of science vol 11 (1): 71-82, june 2020 eissn: 2536-8400 © faculty of science doi: http://doi.org/10.4038/rjs.v11i1.88 university of ruhuna © faculty of science, university of ruhuna 71 sri lanka morphology and optical properties of cuals2 crystals prepared using the solid-phase al and s precursors j. damisa1*, b. olofinjana2, o. ebomwonyi1, m.a. eleruja2, s.o. azi1 1department of physics, university of benin, benin city, nigeria 2 department of physics & engineering physics, obafemi awolowo university, ile-ife, 220005, nigeria *correspondence: john.damisa@uniben.edu, https://orcid.org/0000-0002-9017-755x received: 05th march 2019, revised: 25th may 2020, accepted: 25th june 2020 abstract copper dithiocarbamate and aluminium dithiocarbamate were prepared and then characterized by infrared spectroscopy. the combination of the prepared precursors in different ratios was deposited on glass substrates using metalorganic chemical vapour deposition (mocvd) technique at 450oc through the pyrolysis of the precursors to yield cu-al-s thin films. compositional, morphological, structural and optical characterizations were then carried out. the compositional analysis revealed that the ratio of cu to al in the precursor is not preserved in the films. morphological study showed that the films are polycrystalline in nature whose homogeneity and grain size distribution decrease with a decrease in al content of the films. the crystallinity of the films was further revealed from the pxrd results with the formation of the cu-al-s crystal structure as the al content increases in the precursor. the energy gap obtained falls between 2.63 and 2.75 ev which decreases as the al content in the films decreases. optical constants such as refractive index and extinction coefficient exhibit a decreasing trend as the al content in the film decreases. keywords: energy gap, infrared spectroscopy, mocvd, semiconductors, thin film. 1 introduction copper aluminium sulphide has gained much attention due to its special properties such as high absorption coefficient of about 105 cm-1 and wide direct energy gap of about 3.49 ev together with its constituent elements (cu, al and s) which are nontoxic and abundant in nature (jaffe and zunger 1983). these properties of cu-al-s have made it a potential material in many different optoelectronic applications such as oxygen gas sensor eyeglass industry, optical detectors, solar cells, light-emitting diodes (led) and nonlinear optics (abaab et al. 2000, reshak and auluck 2008). several methods have been employed to synthesize cu-al-s semiconductors in the form of thin-films. these methods include chemical spray pyrolysis, atomic layer deposition (ald), chemical vapour transport (cvt), thermal evaporation technique, https://creativecommons.org/licenses/by-nc/4.0/ https://orcid.org/0000-0002-9017-755x https://orcid.org/0000-0002-9017-755x j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 72 sulfurization, dip coating (olejniček et al. 2011, chaki et al. 2013, duclaux et al. 2015, kumar et al. 2015, ahmad 2017). other methods reported are metal decomposition (md), chemical bath deposition technique (cbd), vacuum thermal evaporation, spray pyrolysis technique, thermal evaporation technique, and iodine transport (miyake et al. 1995, abaab et al. 2000, brini et al. 2009, mujdat et al. 2008, alwan and jabbar 2011). most of these methods have their disadvantages which range from nonuniformity with different levels of impurities to lack of reproducibility. in our previous study (damisa et al. 2017), we established that the growth of cuals2 from a single source precursor using the mocvd technique is achievable. however, we observed from literature that the growth of cuals2 in thin-film form proceeds via two or more precursors where a stable film of cuals2 was not achieved (miyake et al. 1995, benchouck et al. 1999). this we believe due to the active aluminium in the matrix of cuals2. aluminium is highly sensitive to moisture, and it hydrolyses in air to form its oxide and hydroxide. this property of aluminium poses the difficulty in the growth of stable cuals2. duclaux et al.(2015) studied the effect of the number of als cycles with respect to cu-s cycles on the growth rate of cu-al-s but reported that none of the films grown represented a cu-al-s crystal structure. therefore, the present study was designed to use solid source precursors of cu and al to grow cu-al-s thin films. these solid source precursors are novel to the best of our knowledge and are scarcely reported in the literature. the mocvd technique used in this study had been previously used to synthesize binary and ternary oxides (ajayi et al. 1994, adedeji et al. 2002) and sulphide films (osasona et al. 1997, eleruja et al. 1998, damisa et al. 2017). 2 material and methods 2.1 preparation and characterization of precursor preparation of the intermediate complex had been reported previously by damisa et al. (2017) by modifying the method of ajayi et al. (1994). this was followed by coupling of the intermediate complex to prepare the various precursors. the preparation of copper dithiocarbamate followed the following procedure. dried ammonium morpholino-dithiocarbamate (6.00 g, 0.03 mol) was dissolved in 50:50 (v/v) of acetone – water solvent. copper (ii) chloride (3.00 g, 0.02 mol) was dissolved in 50 cm3 of ethanol. solution of copper (ii) chloride in ethanol solvent was added gradually to the solution of ammonium morpholino-dithiocarbamate on a hot plate and stirred vigorously. upon addition, there was a spontaneous formation of a brown precipitate which was heated for about 30 minutes at a temperature of 50 oc. the product was then filtered, allowed to dry in air for 24 h, and put in an oven maintained at a temperature of 60 oc for 96 h. the yield was found to be 58.40%. the same route was followed in the preparation of aluminium dithiocarbamate where 6.00 g (0.03 mol) of ammonium morpholino-dithiocarbamate was dissolved in 50:50 (v/v) of acetone – water solvent. aluminium (iii) chloride (2.00 g, 0.015 mol) was also dissolved in 30:10 j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 73 (v/v) of ethanol-water solvent. solution of aluminium (iii) chloride in ethanol-water solvent was gradually added to the solution of ammonium morpholino-dithiocarbamate on a hot plate and stirred vigorously. upon addition, spontaneous formation of a white solution was occurred which was then put in a freezer for 24 h to precipitate. the product formed was then filtered, allowed to dry in air for 24 h after which it was put in an oven maintained at a temperature of 60 oc for 96 h and the yield was found to be 44.33%. the copper dithiocarbamate and aluminium dithiocarbamate were ground into smooth powder and characterized by infrared (ir) spectroscopy using a shimadzu ftir-8400 s spectrophotometer. the transmission spectrum was measured in kbr at normal incidence angle of 𝜃 = 900 over the range of 4000 and 500 cm-1 at ambient room temperature. 2.2 film deposition and characterization combination of copper dithiocarbamate and aluminium dithiocarbamate in different proportions were deposited on the substrate (table 1). these precursors were ground into smooth powder, poured into a receptacle and nitrogen gas was bubbled through at a pressure rate of 2.5 dm3/min. the nitrogen-borne particle was then transported into the chamber which consists of a pyrex tube maintained at 450℃. to maintain good and uniform thermal contact, the substrate was supported on a steel block. the deposition process lasted for 2 h. during this process, the precursor sublimed before thermal decomposition on reaching the chamber, resulting in the formation of the films. the substrate, the steel block, and the pyrex tube were cleansed in distilled water, methanol, ethanol, acetone and distilled water before each deposition. to reduce some of the handling problems associated with sulphur, the deposition process was carried out in a fume cupboard. the mocvd technique used in this work has the added advantage of not needing a machine-driven pump to expunge the byproduct as they are carried away by the nitrogen inert gas. after all the deposition, the films were designated as cas1, cas2, and cas3 for easy identification. table 1. different proportions of the precursors used for the deposition of the thin films. thin films precursor combination cas1 80% copper dithiocarbamate + 20% aluminium dithiocarbamate cas2 50% copper dithiocarbamate + 50% aluminium dithiocarbamate cas3 20% copper dithiocarbamate + 80% aluminium dithiocarbamate compositional analysis was carried out using phenom prox energy dispersive x-ray (edx) machine operated with acceleration voltages of 15 kv while the surface morphology was carried out by zeiss ultra plus 55 field emission scanning electron microscope (fe-sem) operated at an accelerated voltage of 1.0 kv. the film was coated with gold in order to avoid charging effect and enhance the sem micrograph. j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 74 the coating of the films was carried out using coater quorum (q150r es). the structural analysis was carried out by d8 advance x-ray diffractometry (xrd) using cukα radiation (λ= 1.5406 å ). the optical properties of the films were investigated at room temperature using janway 6405 uv-visible spectrophotometer in the wavelength range of 300-1500 nm. 3 results and discussion 3.1 infrared spectrophotometry of the precursor fig. 1. ir spectrum of (a) aluminium dithiocarbamate complex and (b) copper dithiocarbamate complex. (a) (b) j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 75 the ir spectrum in figure 1 shows that the two complexes exhibit the basic absorption bands between 4000 and 500 cm-1. the major peaks in figure 1a are o-h vibration at 3421 cm-1, c-h vibration between 3037 and 2862 cm-1 and the three n-h bonds between 2476 and 2177cm-1, the carbonyl stretching, c at 1631cm-1, c-c stretching at 1419, 1219 and 1105 cm-1. al-s bands are below 422 cm-1. in figure 1b, the major peaks are o-h vibration at 3443 cm-1, c-h vibration between 2968 and 2854 cm-1, the carbonyl stretching, c at 1707 cm-1, c-c stretching at 1485, 1232, 1111 and 1026 cm1, and cu-s bands are below 420 cm-1. the peaks are real and plots above 420 and 422 cm-1 showed an insignificant result. 3.2 compositional analysis the spectra as shown in figure 2 revealed the presence of cu, al, s and other elements that can be associated with the glass substrate without any carbon impurity. fig. 2. edx spectrum for cu-al-s thin films, (a) cas1 (b) cas2 and (c) cas3. the figure 2 shows that there is a complete pyrolysis of copper dithiocarbamate and aluminium dithiocarbamate to form cu-al-s thin films. the proportion of elements in the film cu:al:s was found to be 32.93: 35.03: 32.04 for cas1, 42.40: 29.00: 28.60 for cas2 and 42.64: 28.63: 28.73 for cas3. these results are given within the limit of the equipment used. furthermore, it is observed that the ratio of cu to al in the various combinations of copper dithiocarbamate and aluminium dithiocarbamate used as precursor is not preserved in the films. in fact, the aluminium content in the films decreases as the aluminium content in the precursor increases. here, it is believed to be due to the reason that when the precursor first sublime in the hot zone of the mocvd reactor, it results in the breakdown of the metal-metal bonds of the precursor in the vapour phase, and at the deposition temperature, there will be subsequent reconstitution of the bonds. this possible reconstitution of the bonds after the breaking of the metal-metal bonds then leads to the obtained stoichiometry in the various cas films. this hypothesis has been observed by adedeji et al. (2002) and eleruja et al. (1998). j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 76 3.3 morphological analysis fig. 3. sem micrographs of cu-al-s thin films, (a) cas1 (b) cas2 and (c) cas3. figure 3 shows the sem micrograph of the deposited films. for 20% aluminium complex, the films (cas1) are dense and the particle size distribution seems to be broad. it also shows the coalescence of small grains into big grains, which are well distributed over the surface. for the films deposited with 50% of aluminium complex (cas2), the grains are seen to coalesce into short rods which are continuous with wellconnected grains, dense, homogenous and without any visible cracks. as the aluminium complex increases to 80% (cas3), the films show that the particle size distribution of the grains seems to be less broad, less dense and non-homogeneous with the formation of nano-rods. furthermore, the micrographs revealed that all the deposited films are polycrystalline in nature. overall, homogeneity and grain size distribution decrease with a decrease in aluminum content of the films. 3.4 structural analysis the x-ray diffraction of the film deposited at different aluminium content is shown in figure 4. the film deposited with 20% aluminium complex (cas1) shows that there is no cu-al-s crystal structure but a mixture of cu2-xsx polymorphs such as cus2 (chalcocite) with diffraction angle at 29.095o and cu38s28 (spionkopite) with diffraction angle at 27.706o. films deposited with 50% aluminium complex (cas2) show the formation of cu-al-s crystal structure with intense peaks occurring at diffraction angle 29.329 o with a preferred orientation at 112 planes and other peaks observed belong to the chalcocite cu2s polymorphs with diffraction angle at 35.631 o, 32.028o , and 38.686o respectively. also, films deposited with 80% aluminium complex show the formation of cu-al-s crystal structure with prominent peaks at diffraction angle 29.257o with a preferred orientation at 112 planes. other peaks recorded belong to the cu2s (chalcocite) structure at diffraction angle 35.645 o. accordingly, the results showed that as the aluminium content in the precursor is increased, the films deposited j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 77 become more crystalline. there is also an increase in the formation of the cu-al-s structure which is clearly in contrast to the result of duclaux et al. (2015). fig. 4. pxrd pattern of cu-al-s thin films (cas1, cas2 and cas3). 3.5 optical properties figure 5 shows the transmission spectrum in the wavelength range of 300-1500 nm for all the cas thin films deposited at 450 0c. fig. 5. transmittance against wavelength for cu-al-s thin films (cas1, cas2 and cas3). j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 78 it was observed that all the films have high transmittance in the visible and nearinfrared region. however, the transmittance decreases as the content of aluminium in the films decreases with cas3 having the lowest. this is believed to be due to the nonhomogeneous nature of cas3 as observed by the sem micrograph (figure 3c) (smaili 2011). also, the high transparency in the visible region implies that the films absorb less photon of light imposed on it. this property of high transmittance in the vis and ir makes the films suitable material for eyeglass industry as they are capable of transmitting the visible radiation needed for vision (damisa et al. 2017). the absorption co-efficient, 𝛼 is evaluated using equation (1) (harbeke, 1972) ∝= 1 t ln ( 1 t ), (1) where t is the thickness of the films with values 319, 670 and 102 nm for cas1, cas2 and cas3 films respectively. the optical energy gap can be estimated from the tauc plot (tauc, 1974) 𝛼 = 𝐴 ℎ𝑣 (ℎ𝑣 − 𝐸𝑔) 𝑟 , (2) where r is an index which can be assumed to have values, 1/2, 3/2, 2 and 3 depending on the nature of the electronic transition responsible for the absorption. exponent r = 1/2 is for allowed direct transitions, r = 2 for allowed indirect transitions, r = 3 for forbidden indirect transitions and r = 3/2 for forbidden direct transitions. figure 6 shows the plot of square of absorption against photon energy. the extrapolation of the linear portion of the plot to the energy axis gives the energy gap of the various cas thin films. the energy gap obtained is 2.75 for cas1, 2.68 for cas2 and 2.63 ev for cas3. this shows a slight decrease in the band-gap as the aluminium content in the films decreases. fig. 6. square of absorption coefficient against energy for cu-al-s thin films (cas1, cas2 and cas3). j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 79 however, our observed band gap is smaller than the expected value of 3.49 ev for cuals2 but gives a better result when compared with 2.30 – 2.60 ev reported by duclaux et al. (2015). similarly, the observed result in this study is in good agreement with 2.88 – 3.28ev band gap reported in our previous study (damisa et al. 2017) and 2.40 – 2.81 ev obtained by alwan and jabbar (2011). different factors may cause such a reduction in the energy gap as aluminium content in the films decreases. the decrease in energy gap tailing which is triggered by the disorder of the film, defects, residual strain, impurities, and disorders of grain boundaries can cause the reduction (hossain et al. 2017). other important parameters usually considered for materials to be used in optical applications include the refractive index, n and the extinction coefficient k. in the region of inter-band transition that has strong absorption, n and k can be determined by equations (3) and (5) respectively (swanepoel 1983), only when the illuminations of electromagnetic waves are perpendicular to the surface of the film 𝑥 = (𝑛−1)3(𝑛+𝑠2) 16𝑛2𝑠 t, (3) where s is the refractive index of glass and 𝑥 the absorbance given by equation (4) (swanepoel, 1983) 𝑥 = 𝑒 −∝𝑡, (4) 𝑘 = 𝛼𝜆 4𝜋 , (5) where α and λ are the absorption coefficient and wavelength respectively. fig.7. refractive index against energy for cu-al-s thin films (cas1, cas2 and cas3) figures 7 and figure 8 show the plots of refractive index, n versus energy, e and extinction coefficient, k versus energy e, respectively. the refractive index and extinction coefficient first increase at lower energy, then decrease with energy, and finally increase again at higher energy. this fluctuation of n and k can be attributed to j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 80 successive internal reflections or as a result of trapped photon energy within the grain boundaries (ongal et al. 2000). we conclude from our xrd results that the overall increase of n and k with e can be attributed to the degree of crystallinity of the films as the aluminium content in the films decreases. fig. 8. extinction coefficient against energy for cas1, cas2 and cas3 4. conclusions cu-al-s thin films were deposited on the glass substrate by mocvd technique at 450oc and pressure rate of 2.5 dm3/ min using different proportions of copper dithiocarbamate and aluminium dithiocarbamate as the precursor. the elemental composition of the films as determined by edx revealed that the expected element that is copper, aluminium and sulphur are present. we also observed that the ratio of cu to al in the various combinations of copper dithiocarbamate and aluminium dithiocarbamate was not preserved in the films. morphological study shows that the films are polycrystalline in nature whose homogeneity and grain size distribution decreases with a decrease in aluminum content of the films. the crystallinity of the films where further revealed from the xrd results which shows the formation of the cu-al-s crystal structure as the aluminium content increases in the precursor. from the optical characterization, the cu-al-s thin film shows a high transmittance in the visible and nearinfrared region which decreases as the aluminium content of the films decreases. the energy gap obtained falls between 2.63 and 2.75 ev which decreases as the aluminium content in the films decreases. other optical constants such as refractive index and extinction coefficient also exhibit a decreasing trend as the aluminium content in the film decreases. the energy gap values together with other j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 81 optical properties exhibited by the material can find applications in areas like solar cells and other optoelectronic devices. acknowledgements the authors are grateful to dr. remy and the entire staff of ithemba laboratory, south africa for their assistance in the pxrd measurements. the deposition process was carried out at the material science/solid state laboratory of the department of physics and engineering physics, obafemi awolowo university ile ife, nigeria, so we thank the entire staff of the department. we also wish to appreciate two anonymous reviewers for their comments and suggestions. references abaab m, bouazzi as, rezig b. 2000. competitive cuals2 oxygen gas sensor. microelectronic engineering 51: 343–348. https://doi.org/10.1016/s0167-9317(99)00495-5. adedeji av, egharevba go, jeynes c, ajayi eob. 2002. preparation and characterization of pyrolytically deposited (co-v-o and cr-v-o). thin solid films 402: 49-54. doi:10.1016/s00406090(01)01602-9. ahmad sm. 2017. study of structural and optical properties of quaternary cuxag1-xals2 thin films. international journal for light and electron optics 127: 10004-100013. doi:10.1016/j.ijleo. 2016.07.034. ajayi ob, osuntola ok, ojo ioa, jeynes c. 1994. preparation and characterization of mocvd thin films of cadmium sulphide. thin solid films 248:57-62. https://doi.org/10.1016/0040-6090(94) 90211-9. alwan t, jabbar m. 2011. structure and optical properties of cuals2 thin films prepared via chemical bath deposition. turkish journal physics 34: 107-116. doi:10.3906/fiz-1007-14. benchouck k, el moctar c, marsillac s, bernede jc, pouzet j, barreau n, emziane m. 1999. growth and physicochemical characterization of cualte2 films obtained by reaction, induced by annealing, between cu/al/te/al/cu... al/cu/al/te layers sequentially deposited. journal of materials science 34: 1847-1853. doi:10.1023/a:1004575612586. brini r, schmerber g, kanzari m, werckmann j, rezig b. 2009. study of the growth of cuals2 thin films on oriented silicon (111). thin solid films 517: 2191–2194. doi:10.1016/j.tsf.2008.10.086. chaki sh, deshpande mp, mahato ks. 2013. growth and microtopographic study of cuals2 single crystals. american institute of physics conference proceedings 1536: 833–834. doi:10.1063/ 1.4810486. damisa j, olofinjana b, ebomwonyi o, bakare f, azi so. 2017. morphological and optical study of thin films of cuals2 deposited by metal organic chemical vapour deposition technique. materials research express 4: 086412 (1-10). https://doi.org/10.1088/2053-1591/aa851d. duclaux l, donsanti f, vidal j, bouttemy m, schneider n, naghavi n. 2015. simulation and growing study of cu–al–s thin films deposited by atomic layer deposition. thin solid films 594: 232–237. https://doi.org/10.1016/j.tsf.2015.06.014. eleruja ma, adedeji av, ojo iao, djebah a, osasona o, aladekomo jb, ajayi eob. 1998. optical characterization of pyrolytically deposited znxcd1−xs thin films. optical materials 10: 257-263. doi:10.1016/s0925-3467(97)00178-x. harbeke g. 1972. optical properties of semiconductor, amsterdam: north-holland pub.con. hossain ms, kabir h, rahman mm, hasan k, bashar ms, rahman m, gafur ma, islam s, amri a, jiang zt, altarawneh m, b.z. dlugoggorski bz. 2017. understanding the shrinkage of optical absorption edges of nanostructured cd-zn sulphide films for photothermal applications. applied surface science 392: 854-872. https://doi.org/10.1016/j.apsusc.2016.09.095. j. damisa et al. morphology and optical properties of cuals2 crystals ruhuna journal of science vol 11 (1): 71-82, june 2020 82 jaffe je, zunger a. 1983. electronic struture of the theory chalcopyrite semiconductors cuals2, cugas2, cuins2,cualse2 and cuinse2. physical review b 28: 5822-5847. https://doi.org/10.1103/ physrevb.28.5822. kumark, jariwalac, pillair, chauhann, raole pm. 2015. preparation & characterization of sio2 interface layer by dip coating technique on carbon fibre for cf/sic composites. american institute of physics conference proceedings 1675: 020046 (1-4). doi:10.1063/1.4929204. miyake h, yamada m, sugiyama k. 1995. vapor phase epitaxy of cuals2 on cugas2 substrate by the iodine transport method. journal of crystal growth 153: 180-183. doi:10.1016/00220248(95)00199-9. mujdat c, saliha i, yasemin c. 2008. structural, morphological and optical properties of cuals2 films deposited by spray pyrolysis method. optics communications281: 1615–1624. doi:10.1016/ j.optcom.2007.11.031. olejniček j, flannery le, darveau sa, exstrom cl, kment s, ianno nj, soukup rj. 2011. cuin1−xalxs2 thin films prepared by sulfurization of metallic precursors. journal of alloys and compounds 509: 10020– 10024. doi:10.1016/j.jallcom.2011.08.016. ongal hc, dai jy, hung kc, chan yc, ho st. 2000. electronic structure of polycrystalline zno thin films probed by electron energy loss spectroscopy. applied physics letters 77: 1484-1486. doi:00036951/2000/77(10)/1484/3/$17.00. osasona o, djebah a, ojo iao, eleruja ma, adedeji av, jeynes c, ajayi eob. 1997. preparation and characterization of mocvd thin films of zinc sulphide. optical materials 7: 109-115. doi:10.1016/s0925-3467(96)00065-1. reshak ah, auluck s. 2008. electronic properties of chalcopyrite cualx2 (x = s, se, te) compounds. solid state communication 145: 571–6. doi:10.1016/j.ssc.2007.12.034. smaili f. 2011. effect of annealing on the structural and optical properties of cuin1-xalxs2 thin films. materials sciences and applications 2: 1212-1218. doi: 10.4236/msa.2011.29164. swanepoel r. 1983. determination of the thickness and optical constants of amorphous silicon. journal of physics e 16: 1214-1222.https://doi.org/10.1088/0022-3735/16/12/023. tauc j. 1974. amorphous and liquid semiconductor, london: plenum. 159. http://aip.scitation.org/author/kumar%2c+kundan http://aip.scitation.org/author/jariwala%2c+c http://aip.scitation.org/author/pillai%2c+r http://aip.scitation.org/author/chauhan%2c+n http://aip.scitation.org/author/raole%2c+p+m sv-lncs ruhuna journal of science vol 9(2): 78-91, december 2018 eissn: 2536-8400  faculty of science http://doi.org/10.4038/rjs.v9i2.41 university of ruhuna  faculty of science, university of ruhuna 78 an ionic liquid based gel polymer electrolyte to be employed in power generating applications k.w. prasadini, k.s. perera* and k.p. vidanapathirana department of electronics, wayamba university of sri lanka, kuliyapitiya, sri lanka correspondence: *kumudu31966@gmail.com; orcid: 0000-0003-4491-662x received: 21st june 2018, revised: 28th november 2018, accepted: 25th december 2018 abstract. ionic liquid (il) based gel polymer electrolytes (gpes) are being investigated extensively at present as substitutes for conventional gpes based on a polymer, a salt and solvents. the main reason behind this is the drawbacks in usage of solvents. il based gpes have been employed for energy storage devices such as batteries and super capacitors due to their interesting mechanical, physical and electrochemical properties. this study focused on synthesis preparation and characterization of an il based gpe consisting of poly(vinylidene fluoride-co-hexafluoropropylene) (pvdf-co-hfp), zinc trifluoro metha -nesulfonate (zn(cf3so3)2 zntf) and 1-ethyl-3-methylimidazolium trifluoromethanesulfonate (1e3mitf). thin film samples were prepared using solvent casting method. the optimized composition was found to be 1 pvdf-co-hfp: 1 1e3mitf: 3 zntf (by weight basis). this mechanically stable, thin film has the maximum room temperature conductivity of 7.42×10-3 s cm-1. conductivity variation with temperature follows vogel tamman fulcher (vtf) behavior confirming the relation of conductivity mechanism with the free volume theory. the il based gpe is a purely an ionic conductor having a considerable anionic contribution. it shows stability up to 2.5 v which is very much convenient from a practical point of view. oxidation and reduction of zn takes place at the potentials of 0.5 v and –0.5 v, respectively. in addition, zn platting and stripping occurs only on the zn electrodes but not on the stainless steel (ss) electrodes. impedance measurements taken for the gpe continuously for a long period of time exhibited a satisfactory stability with zn electrodes. keywords. solvent casting method, ionic conductor, vtf behavior, poly(vinylidenefluoride-co-hexafluoropropylene),1-ethyl-3-methylimid -azolium trifluoromethanesulfonate. 1 introduction over recent years, there has been a significant breakthrough in the field of solid polymer electrolytes due to their excellent features that are making them k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 79 vol 9(2): 78-91, december 2018 suitable candidates in the arena of electrochemical devices. even they possess attractive features, their low ambient temperature conductivities have set some limitations in applications (wang et al. 2017). various attempts have been made to improve the conductivity and the introduction of gel polymer electrolytes (gpes) is one such outcome. gpes are consisted of a polymer, a salt and solvents. they have been extensively employed in batteries, super capacitors, fuel cells etc due to the blended characteristics of high ionic conductivity and good mechanical properties (hashmi et al. 2005, kuo et al. 2002). recently, ionic liquids (il) based gpes have motivated the attention of researchers owing to some inherent drawbacks of conventional gpes such as reduction of conductivity with time, interfacial instability towards electrodes and some safety issues (pandey et al. 2010). many of those demerits have been identified as due to the presence of solvents. therefore, the necessity to replace solvents by a suitable substitute has been well realized. ils which are room temperature molten salts with bulky asymmetric organic cations and inorganic anions have been recognized as a suitable set of materials to be used in place of solvents (shamsipur et al. 2010). they have attractive physiochemical properties including non-volatility, non-toxicity, wide electrochemical window, negligible vapor pressure, high conductivity and high thermal stability (wang et al. 2017). when they are incorporated in a gpe with a polymer and a salt, conductivity enhancement has also been observed (liu et al. 2014). liew et al have reported that conductivity increase takes place in three ways, i.e., (i) softening polymer backbone, (ii) weakening transient coordination bonds, and (iii) demolishing crystalline structure (liew et al. 2013). these il based gpes prepared using different polymer hosts like poly(etheleneoxide) (peo), poly(acrylonitrile) (pan), poly (vinylidinefluoride) (pvdf) and poly(vinylidenefluoride-cohexafluoropropylene) (pvdf-co-hfp) have been employed for various applications [singh et al. 2017, chaurasia et al. 2011). with the new global insight towards the clean and efficient devices, much attention has been focussed on replacing li which is a very harmful, toxic and expensive material by environmental friendly, non toxic and abundant materials such as zn, mg, na and cu [agrawal et al. 2013, kim et al. 2015, perera et al. 2017). in this regard, various gpes with non li salts have received a substatntial attention from the scientific community. as far as il based gpes are concerned, there are very few reports based on non li based systems. therefore, by considering the above facts, this study is forcused on the synthesis and characterization of an il based gpe comprising 1-ethyl-3methylimidazolium trifluoromethanesulfonate(1e3mitf) as the il and zinc trifluoromethanesulfonate (zn(cf3so3)2 zntf) as the salt entrapped in pvdf-co-hfp. here, pvdf-co-hfp co polymer was selected due to its promising characteristics like low glass transition temperature, low crystallinity and high dielectric constant (tripathi et al. 2013, liu et al. 2014). moreover, it has been identified as a semicrystalline polymer with crystalline k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 80 vol 9(2): 78-91, december 2018 vdf and amorphous hfp phases. crystalline nature is sufficient to maintain mechanical stability while liquid electrolytes are retained well due to amorphous nature. electrochemical impedance spectroscopy (eis), dc polarization test, cyclic voltammetry (cv) and linear sweep voltammetry (lsv) have been used as characterization techniques. 2 experimental methods 2.1 materials poly(vinylidinfluoride-co-hexafluoropropylene) (pvdf-co-hfp) with average molecular weight 400,000 g mol-1, 1-ethyl-3-methylimidazoium trifluoromethane sulfonate (1e3mitf, 98%), zinc trifluoromethane sulfonate (zn(cf3so3)2 zntf, 98%) and acetone (99.5%) were purchased from sigma aldrich and used without further purification. 2.2 preparation of il based gpe gpe was prepared using solvent casting technique. pvdf-co-hfp was first dissolved in acetone by magnetically stirring at room temperature. then appropriate amounts of 1e3mitf and zntf were added to the resulting mixture and it was again magnetically stirred overnight. finally, the resulting homogeneous, viscous solution was poured in to a glass petri dish and allowed to evaporate acetone to obtain a free standing film. different samples were prepared by varying the composition of zntf and pvdf-co-hfp. 2.3 optimization of the gpe composition a circular shaped sample obtained from one of the above samples was sandwiched between two stainless steel electrodes of a brass sample holder. impedance data were collected using a metrohm autolab impedance analyzer m101 in the frequency range from 0.1 hz to 400 khz. by placing the sample holder inside a glass tube furnace, measurements were taken from room temperature to 55c. the thickness of the electrolyte was measured using a micrometer screw gauge before starting the measurements. above procedure was repeated for all the above samples prepared with different compositions. the sample that showed the highest conductivity was taken for further characterizations. k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 81 vol 9(2): 78-91, december 2018 2.4 transference number measurements a circular shaped, thin film gpe sample obtained from the electrolyte which showed the highest conductivity at room temperature was sandwiched between two stainless steel (ss) electrodes in a brass sample holder. a dc potential of 1 v was applied between the two stainless steel electrodes and the variation of the current flowing through the circuit was monitored with the time at room temperature. 2.5 linear sweep voltammetry (lsv) study a circular shaped gpe sample was sandwiched between a zn electrode and a ss electrode. lsv study was performed in the potential range from 0 v to 3 v at the scan rate of 10 mv s-1 using a three electrode electrochemical cell. ss electrode was used as the working electrode whereas zn electrode was used as reference and counter electrodes. a computer controlled metrohm autolab potentiostat m101 was used to obtain the current variation with voltage. 2.6 cyclic voltammetry (cv) study using the same setup as in section 2.5, cv study was carried out for the cells with configurations of zn/ il based gpe/ zn and ss/ il based gpe/ ss in the potential range from -1.5 v to 1.5 v at the scan rate of 10 mv s-1. 2.7 impedance measurements to investigate the stability a circular shaped gpe was loaded inside a brass sample holder in between two zn electrodes. impedance measurements were gathered at different time intervals for the cell configuration zn / il based gpe / zn for 450 min. using a metrohm autolab impedance analyzer m101 at room temperature. bulk electrolyte resistance (rb) variation was monitored with time. 3 results and discussion 3.1 impedance data for determining conductivity one of the resultant impedance (nyquist) plots is shown in figure 1(a). generally, for a symmetric cell with the configuration, ss/ il based gpe/ ss, impedance plot contains two semi circles at high and medium frequency ranges and a tilted spike at low frequency range. the first semi-circle at the high frequency region represents bulk electrolyte and the other semi-circle k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 82 vol 9(2): 78-91, december 2018 corresponds to charge transfer phenomena at the electrode and the electrolyte interface while the tilted spike represents the diffusion process taken place. fig.1 (a) ac-impedance plot for pvdf-co-hfp:zntf:1e3mitf polymer electrolyte (inset – expansion of high frequency region), and (b) equivalent circuit of ss/il based gpe/ss cell in figure 1(a), the semi-circle at high frequency range is absent and it may be due to absence of required high frequency. the low frequency spike is also not visible due to the insufficient low frequency range. only in the mid frequency region, a part of the corresponding semi-circle appears. this may be also due to the absence of required frequency range. the impedance plot can be described by the equivalent circuit shown in the figure 1(b). in the figure, rb and rct correspond to bulk resistance and charge transfer resistance respectively. cg stands for geometric capacitance and q represents the (a) r b r ct q c g (b) k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 83 vol 9(2): 78-91, december 2018 constant phase element. in the present study, conductivity values were calculated using the values of rb obtained at different temperatures for various salt and polymer concentrations using the following equation,  = (1/rb)(l/a) (1) where σ is the ionic conductivity, is the thickness of the il based gpe film, a is the surface area of the electrolyte and rb is the bulk resistance of the electrolyte. 3.2 variation of isothermal conductivity at different salt concentrations figure 2 illustrates the variation of conductivities of all samples with the salt concentrations at different temperatures. fig. 2. conductivity isotherm curves for il based gpe with different salt concentrations conductivity has initially increased up to the salt concentration ratio of 3 showing the maximum conductivity and then started to decrease. this feature is seen repeatedly at each temperature. it is a well-known fact that conductivity is depending on both charge carrier concentration and mobility of the charge carriers. then the initial enhancement of the conductivity can be due to the increment of mobile charge ions with increment of salt concentration (jayathilake et al. 2014). but with the further increment of the salt concentration, viscosity increases disturbing the mobility of the charge k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 84 vol 9(2): 78-91, december 2018 carriers in the medium which lowers conductivity greatly. on the other hand, at the higher concentrations of charge carriers, aggregation and the formation of ion pairs which are in neutral state are also taking place (ramesh et al. 2012, jayathilake et al. 2015). both phenomena may be responsible for the reduction of ionic conductivity with further increment of salt concentration. 3.3 variation of isothermal conductivity at different polymer concentrations the ionic conductivity variation with the polymer concentration is shown in figure 3. fig. 3. conductivity isotherm curves for il based gpe with different polymer concentrations at lower concentrations of pvdf-co-hfp (less than 1 g), it was not possible to fabricate a free standing gpe. thus in this study, polymer concentration was increased from 1 (by weight ratio). according to the present study, the maximum conductivity has been reported at the polymer ratio of 1. when increasing the pvdf-co-hfp concentration, paths for ion conduction may become tortuous and thereby it limits the mobility of the charge careers leading to decrease the ionic conductivity (rajendran et al. 2002). some have reported that molecular level blending of polymer with salt/il takes place with increasing polymer concentration and that will lead the conductivity to reduce very much (xu et al. 2005). sample with the composition of 1 pvdf-co-hfp:1 1e3mitf : 3 zntf was found to be showing the maximum room temperature conductivity of 7.42 × k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 85 vol 9(2): 78-91, december 2018 10-3 scm-1 and hence it was selected for further studies. apart from the optimum conductivity, it exhibited good mechanical integrity blended with translucent, flexible properties. 3.4 temperature dependence of conductivity figure 4 shows the conductivity-temperature correlation of the il based gpe having the optimized composition. as it shows a curvature, the conductivity and temperature correlation can be explained by vogel – tamman – fulcher (vtf) behavior (juan et al. 2015), =at-1/2exp(-ea/kb(t-t0)) (2) where a is the pre-exponential factor, ea the activation energy, kb the boltzmann constant, t the temperature and t0 the equilibrium glass transition temperature. fig. 4 . variation of the natural logarithm of conductivity with 1000 / t for the optimized il based gpe composition this proves that conductivity mechanism takes place in association with the free volume theory. according to that, the mobility of charge careers is governed by polymer segmental motion by providing sufficient free spaces. when the polymer network expands, more free space is created between adjacent interstitial sites allowing migration from one site to another (wang et al. 2017). this polymer network expansion may be supported by increasing temperature of the system. therefore, this might be one of the reasons for observing a noticeable conductivity increase with increasing temperature. in k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 86 vol 9(2): 78-91, december 2018 addition, when temperature increases, mobile charges may become energetically rich. thus, they speed up the motion in the medium and as a result, conductivity increases. also, the viscosity reduction that is possible with increasing temperature may assist ion mobility favorably for conductivity. 3.5. transference number measurements figure 5 shows the current variation with time for the electrolyte with the electrode configuration of ss/ il based gpe/ ss. fig. 5. current variation with time for the electrolyte with electrode configuration ss/ il based gpe/ ss under 1 v dc potential as can be seen from the figure, the initial current has dropped quickly and reached to a stable state. the presence of initial current is due to ions. since the ss electrodes were of ion blocking nature, the resulting current drops very fast. thereafter, a constant current flows through the cell due to electrons. ionic transference number (ti) was calculated by the following equation ti = (it-ie) / it (3) here is the total current due to ions and electrons and is the current due to electrons. according to the equation, calculated value was 0.98 and it suggests that il based gpe is predominantly an ionic conductor (deraman et al, 2013). variation of current with time for the cell with configuration zn/ il based gpe/ zn is shown in figure 6. as in the previous case, current drops initially as the zn electrodes block the anion movement. the following near constant k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 87 vol 9(2): 78-91, december 2018 fig. 6. current variation with time for the electrolyte with electrode configuration zn/ il based gpe/ zn under 1 v dc potential current is due to cations. calculated cationic transference number ( ) was 0.08 according to the equation tcation = ication / it (4) where is constant current and is total current due to anions and cations. accordingly, anionic transference number was 0.92. this is an indication for the anionic nature of the conductivity of the sample. dominant anionic nature may be due to the bulky cationic group in the il which is less mobile and a large amount of similar anions in the salt and the il (kumar et al. 2009). fig. 7. linear sweep voltammogram of zn/ il based gpe/ ss cell at a scan rate 10 mv s-1 k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 88 vol 9(2): 78-91, december 2018 3.6 determination of the electrochemical stability window the electrochemical stability window is defined as the potential range in which a significant faradic current is absent. a better gpe should possess a wide electrochemical stability. the resulted liner sweep voltammogram is shown in figure 7. the gpe membrane is stable up to a wide range (2.5 v). this value is very suitable for energy generating devices from a practical point of view. 3.7 plating / stripping effect of zn ions figure 8 shows cyclic voltammograms (cvs) obtained between the potential window of-1.5 v to 1.5 v for the cell configurations, ss/ il based gpe/ ss and zn/ il based gpe/ zn at a scan rate 10 mv s-1. here, the ss and zn electrodes are serving as blocking and non-blocking electrodes respectively. fig. 8 cyclic voltammogram for cells zn/ il based gpe/ zn and ss/ il based gpe/ ss at a scan rate 10 mv s-1 there are no peaks in the respective cyclic voltammogram when blocking electrodes are used. but, with non-blocking electrodes, oxidation and reduction peaks can be seen. this elucidates the fact that zn plating and stripping is possible only on zn electrodes. according to the cv, oxidation and reduction take place at 0.5 v and -0.5 v respectively as reported by xu et al. (2005). the result suggests the highly reversible zn plating/ stripping which confirms satisfactory zn ion conductivity (kumar et al. 2010). k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 89 vol 9(2): 78-91, december 2018 3.8 investigation of the stability of the gpe figure 9 shows the variation of the bulk electrolyte resistance (rb) with time. fig. 9. variation of bulk electrolyte resistance (rb) with time according to that, rb does not change very much with time showing that it is quite stable with zn electrodes. this feature makes the il based electrolyte investigated in the present study more suitable for practical applications. if it varies with time, there can be distortions for the proper and efficient operation of any device due to the deterioration of the electrolyte. in addition, this result well confirms the fact that presence of il assists the electrolyte to retain its properties unlike solvents. 4 conclusions by varying the salt and polymer concentrations, it was found out that an electrolyte comprised with 1e3mitf ionic liquid exhibiting the highest room temperature of 7.42 × 10-3 s cm-1 can be obtained by using the chemical composition of 1 pvdf-co-hfp : 1 1e3mitf : 3 zntf by weight basis. a free standing, bubble free, mechanically stable thin film could be obtained by this composition. the conduction mechanism is associated with the free volume theory. transference number studies reveal that the il based gpe is a purely ionic conductor having a major anionic contribution. the sample is stable up to 2.5 v. cv studies describes that the oxidation and the reduction of zn ions take place at 0.5 v and -0.5 v respectively on zn electrodes. moreover, this proves plating and stripping of zn is possible only k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 90 vol 9(2): 78-91, december 2018 on zn electrodes but not on ss electrodes. also, the sample is very stable with zn electrodes. based on these findings, it can be concluded that the electrolyte with the above optimized chemical composition can be employed in energy generating devices such as batteries and super capacitors. further studies are being done to employ this il based gpe in different types of batteries and super capacitors. acknowledgements authors wish to thank for the support extended by national science foundation, sri lanka under the grant numbers, rg/2017/bs/02 and rg/ 2015/eq/07. also, wayamba university of sri lanka is greatly acknowledged for assistance in numerous ways. comments on the initial manuscript from two anonymous reviewers are acknowledged. references agrawal rc, sahu dk. 2013. mg2+ ion conducting polymer electrolytes: materials characterization and all solid state battery performance studies. journal of physical science and applications 3(1): 9-17. chaurasia sk, singh rk, chandra s. 2011. structural and transport studies on polymeric membranes of peo containing ionic liquid emim-ty : evidence of complexation. solid state ionics 183: 32-39. deraman sk, mohamed ns, subban rhy. 2013. conductivity and electrochemical studies on polymer electrolytes based on poly vinyl (chloride)-ammonium triflate-ionic liquid for proton battery. international journal of electrochemical science 8: 1459-1468 hashmi sa, kumar a, tripathi sk. 2005. investigations on electrochemical super capacitors using polypyrrole redox electrodes and pmma based gel electrolytes. european polymer journal 41: 1373-1379. jayathilake ymcd, bandara lrak, vidanapathirana kp, perera ks. 2014. optimization of conductivity of the gel polymer electrolyte based on polymethylmethacrylate (pmma). sri lankan journal of physics 30: 27-33. jayathilake ymcd, perera ks, vidanapathirana kp. 2015. preparation and characterization of a polyacrylonitrile-based gel polymer electrolyte complexed with 1 methyl-3 propylimidazolium iodide. journal of solid state electrochemistry 19(8): 2199-2203. juan p, tafur g, antonio j, fernandez r. 2015. interaction between zn2+cations and n-methyl2-pyrrolidone in ionic liquid-based gel polymer electrolytes for zn batteries. electrochimica acta 176: 14471453. kim b, winslow r, lin i, gururangan k, evans j, wright p. 2015. layer-by-layer fully printed zn-mno2 batteries with improved internal resistance and cycle life. journal of physics 660: 012009012013. kumar d, hashmi sa. 2010. ionic liquid based sodium ion conducting gel polymer electrolytes. solid state ionics 181: 416-423 kuo hh, chen wc, wen tc, gopalan a. 2002. a novel composite gel polymer electrolyte for rechargeable lithium batteries. journal of power sources 110: 27-33. liew cw, ong ys, lim jy, lim cs, teoh kh and ramesh s. 2013. effect of ionic liquid on semi crystalline poly(vinylidenefluoride-co-hexafluoropropylene) solid copolymer electrolytes. international journal of electrochemical science 8: 7779-7794. k.w. prasadini et al. an ionic liquid based gel polymer electrolyte ruhuna journal of science 91 vol 9(2): 78-91, december 2018 liu z, abedin sze, endres f. 2014. electrodeposition and stripping of zinc from an ionic liquid polymer gel electrolyte for rechargeable zinc-based batteries. journal of solid state electrochemistry 18: 2683-2691. pandey gp, kumar y, hashmi sa. 2010. ionic liquid incorporated polymer electrolytes for super capacitor applications. indian journal of chemistry 49a: 743-751. perera ks, vidanapathirana kp, jayamaha b, wewagama l, dissanayake makl, senadheera gkr, vignarooban k. 2017. peo based nano composite polymer electrolyte for redox capacitor. journal of solid state electrochemistry 21: 3459-3465. rajendran s, mahendran o, kannan r. 2002. characterization of [(1-x) pmma – x pvdf] polymer blend electrolyte with li+ ion. fuel 81: 1077-1081. ramesh s, lu sc, morris e. 2012. towards mg ion conducting polyvinylidenefluoride based solid polymer electrolytes with great prospects : ionic and di electric behaviours. journal of taiwan institute of chemical engineers 43: 806-812. shamsipur m, beigi aam, teymouri m, pourmortazavi sm, irandoust m. 2010. physical and electrochemical properties of ionic liquids 1-ethyl-3-methylimidazoliumtetrafluoroborate, 1-butyl-3-methylimidazolium trifluoromethanesulfonate and 1-butyl-1methylpyrrolidinium bis(trifluoromethylsulfonyl)imide. journal of molecular liquids 157: 43–50. singh r, bhattacharya b, gupta m, khan zh, tomar sk, singh v, singh pk. 2017. structural and electrical properties of ionic liquid doped polymer gel electrolyte for dual energy storage devices. international journal of hydrogen energy 42(21): 14602-14607. tripathi sk, gupta a, jain a, kumari m. 2013. electrochemical studies on nano composite polymer electrolytes. indian journal of pure and applied physics 51: 358-361. wang a, xu h, zhou q, liu x, li z, gao r, liu x, zhang l. 2017. electrochemical performances of a new solid composite polymer electrolyte based on hyper branched star polymer and ionic liquid for lithium ion batteries. journal of solid state electrochemistry 10: 2355-2364. wang j, song s, muchakayala r, hu x, liu r. 2017. structural, electrical and electrochemical properties of pva based biodegradable gel polymer electrolyte membranes for mg-ion battery applications. ionics 23(7): 1759-1769. xu jj, ye h, huang j. 2005 novel zinc ion conducting polymer gel electrolytes based ionic liquids. electrochemistry communications 7: 1309-1317. sv-lncs ruhuna journal of science vol 9: 32-43, june 2018 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v9i1.36 university of ruhuna  faculty of science, university of ruhuna sri lanka 32 ascorbic acid retention of freshly harvested seven nigerian green leafy vegetables after soaking in water s.a. akande 1*, m.e. inana 2, e.e. ugama 2, a.e. azeke 2, f. owuno3, c.f. oledibe1, m.n. adindu 2, a.o. adedokun 2, s.i. roberts 2 and o.u. nze-dike 2 1nigerian stored products research institute, km3 asa dam-road, p.m.b.1489 ilorin, nigeria 2nigerian stored products research institute, mile 4, ikwere road, rumueme, p.m.b.5063, port-harcourt, nigeria. 3department of food science and technology, rivers state university of science and technology, nkpolu, p. m. b. 5080, port-harcourt. nigeria *correspondence: desayoakande@yahoo.com ; orcid 0000-0002-6514-0368 received: 9th march 2018, revised: 22nd june 2018, accepted: 30th june 2018 abstract. vitamins are micronutrients needed in the body for important biologic functions. the current study examined the influence of steeping on vitamin c retention of seven nigerian vegetable leaves after soaking for 8 h. one kilograme each was purchased and 6 lots of 100 g were sorted out, cleaned and treated as follow; whole leaf 1, whole leaf 2, sliced leaf 1, sliced leaf 2, sliced and salted leaf 1 and sliced and salted leaf 2. each of the treatments 1 was soaked in 1 litre of distilled water while each of treatments 2 was soaked in 2 litres of distilled water. all treatments were kept for 8 h while monitoring the trend of reduction in vitamin c contents at 2 h intervals. moisture (%) was determined following aoac (2002) methods while dry matter content was estimated from moisture by calculating the difference. ascorbic acid content (mg/100 g) was determined following the method of ndawula et al. (2004). result showed that; moisture, dry matter, and ascorbic acid contents of raw leaves ranged from 67.63–86.70%, 13.30–32.37%, and 103.00–1199.23 mg/100g respectively. during soaking, ascorbic acid retained by the seven green vegetables reduced as follows; 73.39–24.26% (amaranthus viridis), 100.26–19.62% (gnetum africanum), 129.05–27.72% (gongronema latifolium), 66.84–7.55% (ocimum gratissmum), 42.59–4.14% (piper guinense), 77.38–10.26% (pterocapus mildbedii) and 120.02–17.97% (telfaria occidentalis). the study showed that ascorbic acid retention (%) of seven nigerian green vegetable leaves decreased with increasing soaking duration. keywords: nigeria, ascorbic acid, micronutrients, vegetable, soaking 1 introduction vitamin c (ascorbic acid, aa) is a significant vitamin in fruits and vegetables for human diets. it is a water dissoluble mineral demanded for a mailto:desayoakande@yahoo.com s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 33 vol 9: 32-43, june 2018 number of biologic activities and a strong reducing agent which easily oxidizes discretionally to dehydroascorbic acid (bello and fowoyo, 2014). when the cells of fruits and vegetables are cut opened, chopped or crushed, the vitamin c in them may become oxidized by enzyme called oxidases which are present within the cells (fox and cameron, 1980). this and other factors may suggest reasons why vitamin c is regarded as the most easily destroyed of all vitamins (babalola et al., 2010). some animals can make vitamin c from glucose and galactose whereas certain groups of primates and the cavy could not do so (bello and fawoyo, 2014). hence, their requirements for vitamin c must be satisfied through external sources especially through fruits and vegetables. to this end, lee and kader (2000) suggested a high recommendation of vitamin c intake of above 100 mg/day due to the fact that pressure of present day activities have placed high demands for this very important micronutrients. different preparation methods in which vegetable leaves undergo before and during cooking; including hot water dipping, squeezed washing, boiling and probably sun drying largely affect the availability of nutrients in them especially vitamin c. whereas, vitamin c present in fruits are much more available because they are eaten in raw forms, those in vegetable leaves are not (babalola et al., 2010). in addition, large portions of vitamin c is lost in the water during washing and processing of vegetables since it’s a water dissoluble vitamin (babalola et al., 2010). one of the basic needs of man is the provision of sufficient and nutritionally adequate food (yakubu et al., 2012). however, the tremendous post-harvest treatments they are subjected to often resulted in unequal food quality (oboh, 2003). several studies in this regard has established these facts. for instance, favell (1998) stated that green vegetable leaves are prone to vitamin c damage. domestic cooking also affects the resultant nutrients derived by the consumer particularly that of labile, water dissoluble c vitamin (babalola et al., 2010). aside that, various means through which fruits and green vegetables which consist vitamin c are apportioned reduced their retentivity. for example, the acts of contusion, peeling, slicing or cutting into pieces which introduce the materials into air also reduced vitamin c retentivity (bello and fawoyo, 2014). ariahu and egwujeh (2009) studied the results of hot water dipping and drying conditions on vitamin c retention of two green vegetable leaves. akande et al. (2015) also examined the influence of steeping duration and amount of water on vitamin c concentrations of three green vegetable leaves from nigeria. it is a common knowledge and scientifically proven facts that soaking affects ascorbic acid concentration during processing, therefore the need for extensive study of most common leafy vegetables in nigeria for the purpose of documentations and general knowledge. hence, this study is an attempt to s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 34 vol 9: 32-43, june 2018 investigate the percentage vitamin c retention of freshly harvested seven nigerian green leafy vegetables after soaking in water for eight hours. 2 material and methods 2.1 sample collection and treatments analytical grade chemicals and reagents were used throughout this work. trichloroacetic acid (tca) crystal was obtained from loba chemie india, lascorbic acid and 2,6-dichlorophenolindophenol (dcpip) were obtained from kem light laboratories pvt ltd. india. seven green leafy vegetables commonly consumed by the people of south-south and south-east zones of nigeria; including amaranthus viridis (african spinach, green, tete, inene), gnetum africanum (ukazi), gongronema latifolium (utazi), ocimum gratissimum (scent leaf, efinrin-nla, nchanwu), piper guinense (uziza), pterocarpus mildbreadii (oha) and telfaria occidentalis (fluted pumpkin leaf, ugu) were used for the present work. one kilogram (1 kg) each of the above named freshly harvested green leafy vegetables was purchased from farmers and/or major holders in mile 3 market, diobu area port-harcourt, nigeria. each sample was sorted, cleaned and 6 portions of 100 g each were separated out and treated as follows. for soaking, distilled water (dist. water) was used while for salting, nacl (royal salt ltd, apapa, lagos, nigeria) was used. whole leaf 1 (whl1): a portion of untreated leaves soaked in 1 l dist. water whole leaf 2 (whl2): a portion of untreated leaves soaked in 2 l dist. water sliced leaf 1 (sll1): a portion of sliced leaves soaked in 1 l dist. water sliced leaf 2 (sll2): a portion of sliced leaves soaked in 2 l dist. water sliced and salted leaf 1 (ssl1): a portion of sliced leaves soaked in 2% salt solution (1 l) sliced and salted leaf 2 (ssl2): a portion of sliced leaves soaked in 2% salt solution (2 l) the set up was kept on the laboratory bench under ambient conditions for 8 h while the ascorbic acid content was being determined every 2 h. 2.2 determination of moisture and dry matter contents moisture (%) was carried out following aoac (2002) methods. a known weight (5 g) of sample was dried first at 80°c for 4 h in hot-air oven and subsequently at 105°c for 2 h until constant weight was obtained. percentage dry matter (dm) was determined from the moisture (%) of the leafy vegetables as, dry matter (%) = 100 miosture (%). s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 35 vol 9: 32-43, june 2018 2.3 determination of ascorbic acid content vitamin c content was estimated following the method of ndawula et al. (2004). the method was slightly modified and used as follow; 2 g of leaf sample was homogenized in a mortar containing 10 ml of 5% tca (extraction medium). the homogenized mixture was transferred into 100 ml standard flask and made up with more tca solution to the mark and then filtered with whatman filter paper number 1. exactly 10 ml of the filtrate was titrated with standardized 2,6-dichlorophenolindophenol solution. blank determination was carried out at the same time with 10 ml of trichloroacetic acid solution. 2.4 determination of ascorbic acid retention (%) the degree of reduction in aa was monitored every two hours of soaking and the amount of aa retained after each determination was used to estimate the percentage ascorbic acid retention using the relation 2.5 analysis of results results of moisture, dry matter and vitamin c contents of raw leaves were expressed as mean ± sd. ascorbic acid retention was expressed as percentage. data were pooled and analyzed with one-way anova by using spss for windows version 20.0.0 (ibm spss statistics, ibm corporation armonk ny, usa). means were separated with new duncan’s multiple range f-tests (dmrt) as described by duncan (1955). significance was tested at 95% confidence limit (p<0.05). 3 result and discussion the results of moisture, dry matter and ascorbic acid contents of seven green vegetable leaves (table 1) indicated that moisture contents (mc) of all the vegetables ranged from 67.63% (gnetum africanum) to 86.70% (piper guinense) and the dry matter (dm) contents of all the vegetables ranged from 13.30% (p. guinense) to 32.37% (g. africanum). the mc of piper guinense was significantly high (p<0.05) when compared to other green vegetable leaves examined in this study while there was no significant difference (p<0.05) between the mc of amaranthus viridis, gongronema latifolium and s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 36 vol 9: 32-43, june 2018 telfaria occidentalis. the dm content of g. africanum was significantly (p<0.05) higher than the other vegetables while there was no significant difference (p<0.05) between the dm contents of a. viridis, gongronema latifolium and t. occidentalis. the trends in the moisture and dry matter contents of the seven green leafy vegetables examined in this study is a direct reflection of their texture (hardness and/or softness). the same trend was reported (akande et al., 2015) for freshly harvested three green vegetable leaves in nigeria namely; heinsia crinata, talinum triangulare and vernonia amygdlina. however, some reported values of moisture in the current work were lower in comparison with available data most especially the moisture contents of t. occidentalis (79.60%) for instance, solanke and awonorin (2002) reported the moisture content of t. occidentalis as 85.8%. also edeh et al. (2013) reported 87.21% and 83.95% for t. occidentalis and amaranthus caudatus respectively. the reason for lower moisture content of t. occidentalis in the present study might probably be due to many factors ranging from environmental to varietal differences and post-harvest handlings. table 1: moisture, dry matter and ascorbic acid contents of freshly harvested seven nigerian vegetable leaves (mean ± sd, n=3 samples, # common name not available). botanical name local /common name moisture (%) dry matter (%) ascorbic acid (mg/100 g ) amaranthus viridis ‘tete’ (african spinach) 80.71c ± 1.08 19.29c ± 1.08 575.75b ±1.34 gnetum africanum ‘ukazi’# 67.63e ± 0.54 32.37a ± 0.54 322.14d ± 1.98 gongronema latifolium ‘utazi’ (amaranth globe) 80.55c ± 0.83 19.45c ± 0.83 227.40e ± 5.42 ocimum gratissimum ‘nchianwu, ‘efinrin’ (scent leaf) 84.40b ± 0.64 15.60d ± 0.64 103.30g ± 5.99 piper guinense ‘uziza’(climbing black pepper) 86.70a ± 0.59 13.30e ± 0.59 360.12c ± 8.01 pterocarpus mildbreadii ‘oha’# 77.20d ± 0.20 22.80b ± 0.20 190.50f ± 1.08 telfaria occidentalis ‘ugu’ (fluted pumpkin) 79.60c ± 2.01 20.40c ± 2.01 1999.23a ± 0.24 means within the same column with unshared superscript letters are significantly different (p<0.05). the ascorbic acid (aa) contents of seven nigerian green leafy vegetables ranged from 103.30 mg/100 g dry weight (fresh ocimum gratissimum) to 1999.23 mg/100 g dry weight (fresh t. occidentalis). the aa content of t. occidentalis was significantly (p<0.05) higher than the other vegetables examined in this study while the aa content of o. gratissimum was significantly (p<0.05) low compared to the other vegetables examined. the observed values of aa were higher than most available literature reports. for s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 37 vol 9: 32-43, june 2018 instance, solanke and awonorin (2002) reported the aa concentration of fresh t. occidentalis as 221.1 mg/100 g. the aa content of fresh t. occidentalis and amaranthus hybridus were reported as 158.2 mg/100 g and 155.1 mg/100 g respectively (ariahu and egwujeh, 2009). babalola et al. (2010) reported the aa content of t. occidentalis as 62.50 mg/100 g fresh weight and 24.00 mg/100 g fresh weight for amaranthus viridis. also in a report published by edeh et al. (2013) the ascorbic acid contents of t. occidentalis and amaranthus caudatus were 63.00 mg/100 g and 39.08 mg/100 g respectively. the reason for high values recorded for ascorbic acid contents in the present report might be attributed to the facts that; in our report, we calculated the ascorbic acid contents of seven green vegetable leaves on dry matter basis whereas other reports were estimated on fresh or wet weight basis. other reasons may be attributed to environment and varietal differences as well. diminution in ascorbic acid contents (mg/100 g) was observed in all the seven freshly harvested nigerian vegetable leaves after soaking for two, four, six and eight hours as presented in table 2. generally, whole leaf 1 (whl1) and/or whole leaf 2 (whl2) had significantly (p<0.05) high ascorbic acid content (mg/100 g) during all the sampling periods especially in telfaria occidentalis, pterocarpus mildbredii, piper guinense, ocimum gratissimum, gnetum africanum and amaranthus viridis. the interaction between the ascorbic acid retention (%) of seven nigerian green leafy vegetables and time (h) was generally a reverse order relation while the effects of pre-processing treatments (slicing, slicing with 2% salt solution) and volume of water varied from one leaf to the other (figures 1–7). soaking for 2 h resulted in increased aa concentrations in whl1 (0.26%) of gnetum africanum (figure 2), whl1 and whl2 (29.05% each) of gongronema latifolium (figure 3) and whl1 (20.02%) of t. occidentalis (figure 7). similar result was recorded when talinum triangulare (whole leaves) was soaked in water for 2 h, the ascorbic acid content increased up to 161.93% (akande et al., 2015). a good number of treatments retained more than 50% of their original aa concentration after 8 h soaking time; these include, whl2 and ssl2 (57.26% and 51.93% respectively) both in amaranthus viridis (figure 1), whl1 (56.13%) in ocimum gratissimum (figure 4) and whl1 (68.18%) in telfaria occidentalis (figure 7). salt solution (2%) resulted into 14.36% increase in the aa retention (%) of sliced amaranthus viridis compared to the same sample soaked in ordinary water. similar results were obtained when amaranthus hybridus was blanched in 0.2% sodium metabisulphite solution before drying; it resulted into 1.02 times more aa retention in solar dried sample than the sun dried sample within 8 h according to ariahu and egwujeh (2009). s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 38 vol 9: 32-43, june 2018 table 2: ascorbic acid content (mg/100 g) of freshly harvested seven nigerian vegetable leaves after soaking in water for 2, 4, 6, and 8 h (mean ± sd, n=3 samples, results expressed in dry matter basis). after 2 hours whl1 whl2 sll1 sll2 ssl1 ssl2 amaranthus viridis 452.31b± 7.09 392.65a± 2.89 230.99e± 1.73 321.65c± 0.57 256.82d± 6.64 347.98b± 8.89 gnetum africanum 285.41a± 4.89 247.52b± 2.50 153.99e± 4.61 188.14d± 2.60 195.55c± 2.15 148.32f± 5.50 gongronema latifolium 193.80a± 1.73 108.25c± 1.98 174.43b± 5.29 171.38b± 0.00 107.59c± 0.82 107.87c± 2.79 ocimum gratissimum 68.32a± 8.44 69.05a± 8.04 37.12c± 3.23 33.79c ± 4.50 52.32b± 1.96 49.73b± 4.24 piper guinense 147.99a± 6.64 97.08b± 3.10 55.09e± 0.36 68.60cd± 12.10 58.14de± 5.11 45.05f± 1.17 pterocarpus mildbredii 134.28b± 3.64 147.36a± 6.15 118.74c±3.28 120.85c± 6.05 97.22d± 5.81 95.87d± 5.33 telfaria occidentalis 2354.19a ± 49.65 1788.66b± 28.67 1519.04d± 11.39 1396.29e± 43.79 1576.03c± 15.18 1503.70d± 21.14 after 4 hours amaranthus viridis 352.65b ±13.28 383.31a ±6.81 274.32d ±23.18 255.66e ±4.62 326.65c ±8.08 349.65b ±8.08 gnetum africanum 322.98a±2.30 88.94f ±1.72 159.15e ±2.37 163.23d ±1.87 183.14bc ±3.25 180.39c ±3.25 gongronema latifolium 293.45a±11.21 293.44a±11.21 107.82b±1.00 88.20c±2.56 87.31c±2.80 85.83c±0.24 ocimum gratissimum 51.59b ±2.82 69.79a ±7.18 13.73c ±3.23 20.40c±2.82 45.66b ±7.74 50.06b±3.92 piper guinense 103.28b ±5.64 48.78d±5.28 77.65c±4.24 131.13a ±5.76 83.17c±9.27 48.78d±5.28 pterocarpusmildbredii 104.75b±3.08 136.49a±6.34 80.58d±4.03 89.46c±1.00 56.81e±6.15 87.69cd±3.41 telfariaoccidentalis 1907.02a±28.67 1595.76b±21.14 957.97e±26.58 907.48e±19.73 1205.59c±20.09 1117.71d±29.81 after 6 hours amaranthus viridis 251.65d ±2.31 327.98a ±5.20 269.66bc ±6.51 262.32cd ±12.66 261.32cd ±8.50 280.98b ±5.00 gnetum africanum 164.90b ±2.75 157.99c ±c2.30 143.39d±3.87 147.33d ±4.16 176.65a ±4.26 168.07b ±2.84 gongronema latifolium 63.13d ±1.98 104.97a ±3.71 63.03d ±2.15 83.55c ±3.71 99.97bc ±12.37 71.41d ±12.37 ocimum gratissimum 59.39ab ±2.82 43.79b ±4.20 44.92b ±3.23 46.79b ±7.80 47.52b ±6.73 50.12b±7.74 piper guinense 69.45b±3.79 153.38a±3.15 39.63d ±5.28 55.17c±8.72 52.12c ±4.79 51.83c ±5.28 pterocarpus mildbredii 116.53a ±1.13 122.51a ±0.00 85.26b ±4.77 89.46b ±1.00 69.24c ±0.00 116.31a ±8.57 telfaria occidentalis 1828.11a±43.12 1168.33b±16.55 556.76e±27.37 907.48c±19.73 747.46d ±7.59 865.83c ±77.06 after 8 hours amaranthus viridis 139.66d ± 1.53 329.65a± 13.28 141.32d ± 3.51 216.32c ± 12.10 222.32c ± 7.09 298.99b ± 0.00 gnetum africanum 90.70b± 0.00 85.44c± 5.25 126.42a± 0.00 63.21d ± 0.00 66.44d ± 3.85 92.27b ± 2.73 gongronema latifolium 149.96a ± 21.42 164.24a ± 32.72 86.64c± 1.64 109.02b± 3.31 71.21cd± 12.02 65.22d± 1.64 ocimum gratissimum 51.99a± 9.01 13.73c± 3.23 7.80c± 0.00 10.40c± 4.50 21.93b± 5.74 10.40c ± 4.50 piper guinense 62.12a± 9.57 36.59cd ± 9.15 14.92f ± 2.92 24.96ef ± 10.07 42.69bc ± 10.56 34.11de ± 4.29 pterocarpus mildbredii 26.63b ± 0.00 19.53b ± 3.07 40.62a± 6.35 40.62a± 6.35 38.41a ± 3.65 37.28a ± 0.00 telfaria occidentalis 1343.69a ± 36.22 975.43b ± 32.44 401.13f ± 74.11 469.08e ± 42.28 734.34c ± 3.82 611.56d ± 47.42 results are expressed as mean ± sd of triplicate readings (n=3). mean values with different superscripts across the row for each interval are significantly different (p<0.05). whl1=whole leaves in 1 l of water; whl2= whole leaves in 2 l of water; sll1=sliced leaves in 1 l of water; sll2=sliced leaves in 2 l of water; ssl1=sliced leaves in 1 l of 2% salt solution; ssl2= sliced leaves in 2 l of 2% salt solution. s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 39 vol 9: 32-43, june 2018 fig. 1. changes in ascorbic acid retention (%) of amaranthus viridis with soaking time, pre-processing treatments and volume of water. values showed percentage of ascorbic acid retained. whl1=whole leaves in 1 l of water; whl2= whole leaves in 2 l of water; sll1=sliced leaves in 1 l of water; sll2=sliced leaves in 2 l of water; ssl1=sliced leaves in 1 l of 2% salt solution; ssl2= sliced leaves in 2 l of 2% salt solution. fig. 2. variation of ascorbic acid retention (%) of gnetum africanum with soaking time, pre-processing treatments and volume of water (see fig 1 for the description on the key). another category was leaves that did not retain up to 50% of initial aa concentration in any treatment (whl1, whl2, sll1, sll2, ssl1 and ssl2) after 8 h soaking time include gnetum africanum (figure 2), gongronema latifolium (figure 3) and pterocarpus mildbredii (figure 6). the case of piper guinense was exceptional; more than 50% of the original aa concentration was lost in all the treatments even within 2 h soaking time s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 40 vol 9: 32-43, june 2018 (figure 5). the reason for lower aa retention time in piper guinense may be due to its high residual moisture content. since water serve as a medium that facilitates most chemical and biochemical reactions, its increase would be expected to aid chemically and biochemically mediated degradation losses of ascorbic acid as suggested by ariahu and egwujeh, (2009). fig. 3. variation of ascorbic acid retention (%) of gongronema latifolium with soaking time, pre-processing treatments and volume of water (see fig 1 for the description on the key). fig. 4. variation of ascorbic acid retention (%) of ocimum gratissimum with soaking time, pre-processing treatments and volume of water (see fig 1 for the description on the key). s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 41 vol 9: 32-43, june 2018 fig. 5. variation of ascorbic acid retention (%) of piper guinense with soaking time, preprocessing treatments and volume of water (see fig 1 for the description on the key). fig. 6. variation of ascorbic acid retention (%) of pterocarpus mildbredii with soaking time, pre-processing treatments and volume of water. (see fig 1 for the description on the key). s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 42 vol 9: 32-43, june 2018 fig 7. variation of ascorbic acid retention (%) of telfaria occidentalis with soaking time, pre-processing treatments and volume of water (see fig 1 for the description on the key). 4 conclusion the study has shown that generally, the interaction between ascorbic acid retention (%) with respect to soaking time (h) is a reverse order relation. it is a common in-house practice in nigeria to soak leafy vegetables in water probably after slicing in order to remove sand and other extraneous materials attached to them. this study has shown that it is better to wash or soak (if any need) a vegetable before slicing it. this measure would help reduction of oxidative degradation of ascorbic acid. soaking for a maximum of 2 h and/or 4 h caused a significant increase in ascorbic acid content of gnetum africanum, gongronema latifolium, and telfaria occidentalis. however, subsequent processing such as steaming or cooking may cause more damage or adverse effect on the other nutritional components. these areas are still open for more investigations. it is important to note that leafy vegetables are not being consumed majorly for the vitamin c they contained because an appreciable amount of it is lost during in-home processing and cooking. it is therefore advisable to take more fruits to complement the need for ascorbic acid. acknowledgements authors wish to acknowledge the management of nigerian stored products research institute, port-harcourt for the use of laboratory facilities and equipment during the course of this research work. two anonymous reviewers are acknowledged for comments on the initial manuscript. s.a. akande et al. ascorbic acid retention of green leafy vegetables ruhuna journal of science 43 vol 9: 32-43, june 2018 references akande sa, azeke ae, adedokun ao and israel du. 2015. effect of soaking time and volume of water on the ascorbic acid content of three nigerian green leafy vegetables. food science and quality management 44: 23–27. aoac 2002. official method of analysis of aoac international, 17th ed. published by the association of official analytical chemists international, suite 400 2200, wilson boulevard, arlington, virginia 2220 –3301, usa. ariahu cc and egwujeh sid. 2009. effect of blanching and drying conditions on the sensory quality, chlorophyll and ascorbic acid retention of leafy vegetables. nigerian food journal 28(2): 96–101. babalola oo, tugbobo os and daramola aa. 2010. effect of processing on the vitamin c content of seven nigerian green leafy vegetables. advance journal of food science and technology 2(6): 303-305. bello aa and fowoyo pt. 2014. effect of heat on the ascorbic acid content of dark green leafy vegetables and citrus fruits. african journal of food science and technology 5(4): 114– 118. duncan pb 1955. new multiple range and multiple f-tests. biometrics 11: 1-42. edeh ri, adeniji po and salau ba. 2013. the effects of household processing on ascorbic acid and moisture content of some selected nigerian vegetables. iosr journal of environmental science, toxicology and food technology 6(4): 52–54. favell dj. 1998. a comparison of the vitamin c content of fresh and frozen vegetables. food chemistry 62: 59-64. fox ba and cameron a g. 1980. food science-a chemical approach. 3rd ed. hodder and stoughton educational. 265-268 pp. lee sk and kader aa. 2000. pre-harvest and postharvest factors influencing vitamin c content of horticultural crops. post-harvest biology and technology 20: 207–220. ndawula j, kabasa jd and byaruhaanga yb. 2004. alteration in fruit and vegetable β-carotene and vitamin c content caused by open sun drying, visqueen-covered and polyethylenecovered solar dryers. african health science 4(2): 125-130. oboh g. 2003. hemolytic effect of saponin extract from vernonia amygdalina (bitter leaf) on human erythrocyte. applied natural science research 1(14): 25-29. solanke oe and awonorin so. 2002. kinetics of vitamin c degradation in some tropical green leafy vegetables during blanching. nigerian food journal 20: 24–32. yakubu n, amuzat ao and hamza ru. 2012. effect of processing methods on the nutritional contents of bitter leaf (venonia amygdalina). american journal of food & nutrition 2(1): 26–30 ruhuna journal of science vol 11 (2): 131-142, december 2020 eissn: 2536-8400 © faculty of science http://doi.org/10.4038/rjs.v11i2.92 university of ruhuna © faculty of science, university of ruhuna sri lanka 131 antioxidant and anti-inflammatory potential, and chemical composition of fractions of ethanol extract of annona muricata leaf i. u. nwaehujor*1, g. a. olatunji2, o. a. fabiyi3 and s. a. akande1 1biochemistry/chemistry unit, perishable crop research department, nigerian stored products research institute. p.m.b. 1489, ilorin, kwara state, nigeria 2department of chemistry, faculty of physical science, university of ilorin, p.m.b. 1515, ilorin, nigeria 3department of crop protection, faculty of agriculture, university of ilorin, p.m.b. 1515, ilorin, nigeria *correspondence: idorenyinugochi@gmail.com; orcid: https://orcid.org/0000-0001-5851-5283 received: 15th march 2020, revised: 20th july 2020, accepted: 29th november 2020 abstract serious health challenges have been associated with inflammation which is a major cause of mortality in the world. this study evaluated the antioxidant, anti-inflammatory potential, and chemical compositions of fractions of ethanol extract of annona muricata leaf. the leaves were dried at room temperature, blended and extracted in sequential with solvents of varying degree of polarities, i.e., n-hexane, ethyl acetate and ethanol. ethanol extract was fractionated via solvent-solvent partitioning into five fractions, i.e., n-hexane fraction (f1), dichloromethane fraction (f2), dichloromethane/ methanol (1:1) fraction (f3), methanol fraction (f4), and ethanol fraction (f5). these fractions were examined for their in-vitro antioxidant activities on dpph, abts and h2o2 while the antiinflammatory activities were investigated using lipoxygenase inhibition, proteinase inhibition and membrane stabilization assays. the f4 being the most active fraction was further analyzed with gcms to determine its chemical compositions. the results showed that f4 had the highest h2o2 scavenging activity at 10–100 µg/ml. the activity of f4 at 50 µg/ml was significantly higher (p<0.05) than that of other treatments including the standard (vitamin c). activity of f4 also showed significantly higher (p<0.05) membrane stabilization than other fractions at 50-100 µg/ml. f4 exhibited higher antioxidant and anti-inflammatory activities than the other fractions. the activity of this fraction could be attributed to the synergetic effect of various antioxidant compounds present in the fraction. some of the bioactive compounds identified in the gc-ms of f4 were coumaran, tyrosol, phytol, tetracosanol, elaidic acid methyl ester and β-sitosterol. keywords: bioactivity, concentration, inhibition, radical scavenging. 1 introduction the reaction of free radicals with molecular oxygen generates reactive oxygen species (ros), which causes an imbalance between the oxidizing molecules and the http://doi.org/10.4038/rjs.v11i2.92 https://creativecommons.org/licenses/by-nc/4.0/ mailto:idorenyinugochi@gmail.com https://orcid.org/0000-0001-5851-5283 https://orcid.org/0000-0002-4761-4558 i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 132 antioxidant system of the body that results in inflammation. most serious health challenges have been associated with inflammation, and it is a major cause of mortality in the world (krishnamoorthy et al. 2016). steroid drugs, non-steroidal anti-inflammatory drugs (nsaids) and immuno-suppressants which are commonly used for the treatment of inflammatory conditions are associated with side effects, such as gastrointestinal bleeding, suppressed function of the immune system and peptic ulcers (amri et al. 2018). the risk involved in the use of these synthetic drugs has necessitated the search for safe and natural based anti-inflammatory and antioxidants agents. plant secondary metabolites have been found to be effective in the treatment of inflammation and related diseases. annona muricata is a species of annonaceae family that has been widely studied due to its therapeutic potential. different parts of annona muricata plant are used in traditional medicine in the tropics including the bark, leaves, root, fruit and seeds. the plant has ethnomedicinal uses such as antispasmodic, sedative, hypoglycemic, hypotensive, and smooth muscle relaxant (moghadamtousi et al. 2015). the unripe fruit, seeds, leaves, and roots are used as biopesticides and insect repellents (coria-tellez et al. 2018). different parts of annona muricata have been found to contain phytochemicals such as alkaloids, flavonoids, phenolic compounds, glycosides, saponins, tannins, terpenoids and phytosterols (usunobun et al. 2015, aku and okolie 2017). in a previous study (nwaehujor et al. 2020), the antioxidant and anti-inflammatory activities of n-hexane extract, ethyl acetate extract and ethanol extract were compared, and the ethanol extract was found to be most active. the aim of this study was to separate ethanol extract of annona muricata into various fractions of different polarity and to test the bioactivity of the various fractions to ascertain the most active fraction and further determine the compounds that confer the bioactivity to the fraction. 2 material and methods 2.1 preparation and extraction of plant material the plant materials were collected from a local garden in ilorin, kwara state, nigeria. the sample was identified and documented at the herbarium of plant biology department, university of ilorin, nigeria with voucher number uih001/1106.the leaves of annona muricata were dried at room temperature and ground to powder using a mill. cold extraction method was used for the extraction of secondary metabolites from the plant. about 2.5 kg of the leaf powder was extracted with n-hexane for three days. the crude extract solution was decanted, filtered with whatman no. 1 filter paper, and concentrated in vacuo. the n-hexane crude extract (52 g; 2.08%) was coded amh and stored in the refrigerator until further analyses. ethyl acetate was used to extract the remaining plant material for three days. the crude extract was decanted, filtered, and concentrated. the ethyl acetate extract was i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 133 coded amea and weighed 110 g (4.4%). it was stored in refrigerator until further analyses. the residual plant material was extracted with ethanol for three days. the resulting ethanol crude extract which weighed 66g (2.4%) was coded ame and stored in the refrigerator until further analyses. the crude extracts, i.e., n-hexane, ethyl acetate and ethanol were dark green oils. as our previous study has found the ethanol extract to be the most active (nwaehujor et al. 2020), the ethanol extract was further separated into fractions of various polarities using solvent-solvent partitioning to test the bioactivity of those fractions. 2.2 fractionation of ethanol extract concentrated ethanol extract of a. muricata was fractionated into five major fractions using solvent-solvent partitioning. two grams of the concentrated ethanol extract was separated into n-hexane fraction (f1: 25 mg, 1.25 %), dichloromethane fraction (f2: 280 mg, 14%), dichloromethane/ methanol (1:1) fraction (f3: 1.061 g, 53.05%), methanol fraction (f4: 102 mg, 5.10%), and ethanol fraction (f5: 37 mg, 1.85%). 2.3 antioxidant assay of fractions of ethanol extract of a. muricata leaf f1, f2, f3, f4 and f5 were tested for their antioxidant activities. the parameters that were considered include 2,2-azinobis-3-ethylbenzothiozoline-6-sulfonic acid (abts) radical scavenging, 1,1-diphenyl-2-picryl hydroxyl (dpph) radical scavenging and hydrogen peroxide (h2o2) scavenging activities. the experiments were carried out as described by nishaa et al. (2012). abts radical scavenging activity the abts cation radical was produced by the reaction between 5 ml of 14 mm abts solution and 5 ml of 4.9 mm potassium persulphate (k2s2o8). the solution prepared was stored in a dark place at room temperature (280c). the solution was then diluted with ethanol to get an absorbance of 0.7 ± 0.02 at 734 nm. the samples at different concentrations were homogenized with 1 ml of abts solution and its absorbance was recorded at 734 nm. ethanol blank was run in each assay. the activity of the samples was compared with a standard (vitamin c). the percentage abts radical scavenging was calculated using the equation (1) below. abts radical scavenging activity (%) = a0 – a1 / a0 x 100 (1) where a0 is the absorbance of the blank and a1 is the absorbance of the sample. dpph radical scavenging activity one milliliter of the sample solution was added to 1ml of a dpph solution (0.2 mm in ethanol). after 30 minutes of reaction at room temperature (280c), the absorbance of the solution was measured at 517 nm. the free radical scavenging activity of each i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 134 sample was determined by comparing its absorbance with that of a blank solution (ethanol). the free radical scavenging activity was calculated using the equation (2). dpph scavenging activity (%) = a0 – a1 / a0 x 100 (2) where a0 is the absorbance of the blank and a1 is the absorbance of the sample. hydrogen peroxide scavenging activity a solution of h2o2 (43 mm) was prepared in phosphate buffer (0.1m, ph 7.4). the samples at different concentrations were mixed with 3.4 ml phosphate buffer and added to 0.6 ml h2o2 solution. the absorbance value of the reaction mixture was recorded at 230 nm. h2o2 scavenging activity (%) = a0 – a1 / a0 x 100 (3) where a0 is the absorbance of the blank and a1 is the absorbance of the sample. 2.4 anti-inflammatory assay of fractions of ethanol extracts of a. muricata leaf anti-inflammatory studies were carried out with the various fractions (f1, f2, f3, f4, and f5) of ethanol extract. the anti-inflammatory assays include lipoxygenase inhibition, proteinase inhibition and membrane stabilization. the experiments were carried out using the procedures described by leelaprakash and das (2011). inhibition of lipoxygenase activity inhibition of lipoxygenase was studied using linoleic acid as substrate and lipoxidase as enzyme. test samples were dissolved in 0.25 ml of 2m borate buffer at ph 9.0 and added to 0.25 ml lipoxidase enzyme solution and incubated for 5 minutes at 250c. one milliliter of linoleic acid solution (0.6 mm) was then added and mixed thoroughly. the absorbance of the solution was then measured at 234 nm. the analysis was carried out in triplicate. the percentage inhibition was calculated from the equation (4) below. % lipoxygenase inhibition = a0 – a1 / a0 x 100 (4) where a0 is the absorbance of the blank and a1 is the absorbance of the sample. proteinase inhibitory activity the reaction mixture consisted of 0.06 mg trypsin, 1 ml of 20 mm tris hcl buffer (ph 7.4) and 1 ml of test sample of different concentrations. the mixture was incubated at 370c for 5 minutes and then 1 ml of 0.8 % (w/v) casein was added. the mixture was incubated for additional 20 minutes. two milliliters of 70 % perchloric acid was added to terminate the reaction. a cloudy suspension was obtained which was centrifuged and the absorbance of the supernatant was recorded at 210 nm. a i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 135 buffer was used as blank. the experiment was performed in triplicate. the percentage proteinase inhibitory activity was calculated using the equation (5) below. % proteinase inhibition = a0 – a1 / a0 x 100 (5) where a0 is the absorbance of the blank and a1 is the absorbance of the sample. membrane stabilization activity the blood sample used for the experiment was collected from the university of ilorin teaching hospital, ilorin, kwara state, nigeria. the blood sample was mixed with an equal volume of alsever solution (2% dextrose, 0.8% sodium citrate, 0.5% citric acid and 0.42% nacl) and centrifuged at 3000 rpm (rate per minute). the packed cells were washed with isosaline and a 10% suspension was made. various concentrations of extracts were prepared using distilled water. to each concentration, 1 ml of phosphate buffer, 2 ml of hyposaline and 0.5 ml of human red blood cell (hrbc) suspension was added. the solution was incubated at 370c for 30 minutes and centrifuged at 3000 rpm for 20 minutes. the hemoglobin content of the supernatant solution was determined with spectrophotometer at 560 nm. a blank sample was also prepared without the extract. the experiment was carried out in triplicate. the percentage of hrbc membrane stabilization was calculated using the equation (6) below. % membrane stabilization = a0 – a1 / a0 x 100 (6) where a0 is the absorbance of the blank and a1 is the absorbance of the sample. 2.5 gc-ms analysis of f4 based on the results of the antioxidant and anti-inflammatory studies which showed f4 as the most active fraction, f4 was analyzed using gas chromatography/ mass spectrometry (gc-ms) in order to identify the actual compounds present in the fraction and to determine the specific compounds that confer bioactivity on it. 2.6 statistical analysis data were subjected to analysis of variance (anova) and tested for significance difference among treatments by new duncan’s multiple range f-test (dmrt) at (p<0.05) using spss software package version 20.0.0 (ibm spss statistics 2011, ibm corporation armonk ny usa). i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 136 3 results and discussion 3.1 antioxidant potential of fractions of ethanol extract of annona muricata leaf abts radical scavenging activity the results of the abts radical scavenging showed that the activities of all the fractions were comparable with the standard (vitamin c) at 10–50 µg/ml (figure 1). there was no significant difference (p<0.05) in the activities of the various fractions. however, at 100-150 µg/ml the activity of the standard was significantly (p<0.05) higher than that of all the fractions. baskar et al. (2007) reported 90.05% of abts radical scavenging activity for ethanol extract of annona muricata leaf at 500 µg/ml concentration. fig 1. in-vitro abts radical scavenging activity of fractions of ethanol extract of annona muricata leaf. each point on the line graph represents mean of triplicate readings (n=3) while error bars represent standard error (se) of the means. the dpph radical scavenging activity all fractions showed dpph radical scavenging potentials within the range of 20–50 µg/ml. the activities of the fractions were comparable with that of the standard (quercetin) at 10 and 20 µg/ml (figure 2). there was no significant difference (p<0.05) in the activities of the various fractions. at 50-150 µg/ml, the activity of the standard was significantly (p=0.05) higher than that of all the fractions. in a study by leon-fernandez et al. (2017), ethanol extract, chloroform fraction and isolated acetogenins were found to have high dpph activity which was attributed to acetogenin concentration and phenolic compounds in the extract. i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 137 fig 2. the in-vitro dpph radical scavenging activity of fractions of ethanol extract of annona muricata leaf. each point on the line graph represents mean of triplicate readings (n=3) while error bars represent standard error (se) of the means. hydrogen peroxide (h2o2) radical scavenging activity the results showed that f4 had the highest h2o2 radical scavenging activity at 10– 100 µg/ml. the activity of f4 at 50 µg/ml was significantly (p<0.05) higher than that of other treatments including the standard (vitamin c) (figure 3). the activity of f4 could be attributed to the synergetic effect of the compounds present in the fraction. phenolic compounds and flavonoids have been reported to convey antioxidant properties to plant extracts (orak et al. 2019). f1 had the lowest hydrogen peroxide scavenging activity. fig 3. in-vitro hydrogen peroxide scavenging activity of fractions of ethanol extract of annona muricata leaf. each point on the line graph represents mean of triplicate readings (n=3) while error bars represent standard error (se) of the means. i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 138 3.2 anti-inflammatory potential of fractions of ethanol extract of annona muricata leaf lipoxygenase inhibition activity all the fractions showed good lipoxygenase inhibition activities. the activities of all the fractions were comparable with that of the standard (vitamin c) at concentrations 100 and 150 μg/ml (p<0.05). however, the activity of various fractions was not dose dependent (figure 4). the fluctuation in the activity of the various fractions might be due to antagonistic interactions between the compounds in the samples and lipoxygenase and this is not a desirable quality for a drug. in contrast, the activity of the standard increases steadily with concentration and this is desirable property of a good drug. evidence from studies in human cancer cells show that lipoxygenase catalyzed metabolites influence the development and progression of human cancers, agents that block lypoxygenase activities are effective in preventing cancers (steele et al. 1999). fig 4. the in-vitro lipoxygenase inhibition activity of fractions of ethanol extract of annona muricata leaf. each point on the line graph represents mean of triplicate readings (n=3) while error bars represent standard error (se) of the means. proteinase inhibition activity all the fractions tested showed low proteinase inhibition activity compared with the standard (2-[2-{2,6-dichloroanilino}phenyl]acetic acid commonly called diclofenac). the activities of all the fractions decreased with increase in concentrations (figure 5). 2-[2-{2,6-dichloroanilino}phenyl]acetic acid is a nonsteroidal anti-inflammatory drug. apart from inhibiting the production of proteinase, the drug also decreases the i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 139 production of prostaglandin which is a pro-inflammatory agent. even though this drug is highly potent, it is reported to be associated with side effects such as gastrointestinal tract complications, ulcers and cardiovascular problems (perumal et al. 2017). fig 5. the in-vitro proteinase inhibition activity of fractions of ethanol extract of annona muricata leaf. each point on the line graph represents mean of triplicate readings (n=3) while error bars represent standard error (se) of the means. fig 6. the in-vitro membrane stabilization activity of fractions of ethanol extract of annona muricata leaf. each point on the line graph represents mean of triplicate readings (n=3) while error bars represent standard error (se) of the means. i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 140 membrane stabilization activity all the fractions showed good red blood cell membrane stabilization activity (figure 6). even though the activity of f4 was significantly higher (p<0.05) than other fractions at 50 100 µg/ml, it was less than the activity of the standard. the vitality of the cells depends on the viability of their cell membrane. exposure of the cells to injuries leads to oxidation of haemoglobin and secondary damage through free radical induced lipid peroxidation. damage to the cell membrane leads to the release of lysosomal enzymes which cause different disorders. the membrane stabilizing agents inhibit the release of these enzymes (sumathi and anuradha 2016). 3.3 results of gc-ms analysis of f4 table 1: compounds identified in the fraction f4. s/n name of compound % a mf mm (g/mol) 1 levomenthol 0.28 c10h20o 156 2 n,n-dimethyl-1-leucine 2.20 c8h17no2 159 3 coumaran 2.39 c8h8o 120 4 p-vinylguaiacol 1.13 c9h10o2 150 5 syringol 0.62 c8h10o3 154 6 tyrosol 1.11 c8h10o2 138 7 3-hydroxybenzylhydrazine 1.36 c7h10n2o 138 8 stevioside 2.34 c38h60o18 804 9 sorbitol 3.13 c6h14o6 182 10 nonanoic acid 5.11 c9h18o2 158 11 methyl m-hydroxycinnamate 2.02 c10h10o3 178 12 n-hexadecanol 0.78 c16h34o 242 13 5-hydroxy-4,7,7-trimethyl bicyclo[2.2.1] heptan-2-one 0.83 c10h16o2 168 14 1,2-epoxyundecane 0.67 c11h22o 170 15 hexadecanoic acid, methyl ester 3.43 c17h34o2 270 16 1-nonadecene 3.41 c19h38 266 17 2-cyclohexylnonadecane 1.78 c25h50 350 18 elaidic acid methyl ester 8.61 c19h36o2 296 19 phytol 1.73 c20h40o 296 20 methyl stearate 2.68 c19h38o2 298 21 n-tetracosanol 5.56 c24h50o 354 22 palmitic acid beta monoglyceride 5.02 c19h38o4 330 23 alpha-monostearin 3.80 c21h42o4 358 24 3-dodecyl-2,5-furandione 4.46 c16h26o3 266 25 β-sitosterol 4.44 c29h50o 4i4 26 cycloartanyl acetate 1.23 c32h54o2 470 a = abundance; mf = molecular formula; mm = molecular mass as the fraction f4 was found to be most active in some of the antioxidant and antiinflammatory assays, it was subsequently analyzed with gc-ms to determine its chemical composition. the results of the gc-ms analysis (table 1) showed that elaidic acid methyl ester was the most abundant compound present in the sample i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 141 (8.61%). the β-sitosterol, cycloartanyl acetate, tyrosol, p-vinylguaiacol, phytol and coumaran were also identified in the sample. the p-vinylguaiacol is an aromatic compound used as a flavoring agent. tyrosol is a natural antioxidant present in several foods such as wines and green tea. the presence of this antioxidant in foods protects cells and tissues from oxidative injuries thereby preventing diseases such as cancer and several heart related diseases. phytol has been reported to have antimicrobial, anticancer and anti-inflammatory activities (jothi and geetha 2017). the presence of these compounds in f4 may account for the relatively high antioxidant and anti-inflammatory activities. 3.4 limitations appropriate methods for identification of the active constituents in f4 such as hplc, column chromatography and nmr were not available as such gc-ms was employed as it was the only available alternative. 4 conclusions f4 exhibited higher antioxidant and anti-inflammatory activities than the other fractions. the activity of this fraction could be attributed to the synergetic effect of the various antioxidant compounds present in the fraction. some of the bioactive compounds identified in the gc-ms of f4 were coumaran, tyrosol, phytol, tetracosanol, elaidic acid methyl ester and β-sitosterol. acknowledgments two anonymous reviewers are acknowledged for valuable comments on the initial draft of the manuscript. references agu kc, okolie pn. 2017. proximate composition, phytochemical analysis and in-vitro antioxidant potentials of extracts of annona muricata (soursop). food science and nutrition 5:1029-1036. amri o, zekhnini a, bouhaimi a, tahrouch s, hatimi a. 2018. anti-inflammatory activity of methanolic extract from the pistacia atlantica desf. leaves. pharmacognosy journal 10(1): 71-76. baskar r, rajeswari v, kumar ts. 2007. in-vitro antioxidant studies in leaves of annona species. indian journal of experimental biology 45(5): 480 – 485. cora-tellez av, montalvo-gonzalez e, yahia em, obledo-vazquez en. 2018. annona muricata: a review on its traditional medicinal uses, phytochemicals, pharmacological activities, mechanism of action and toxicity. arabian journal of chemistry 11(5): 662-691. jothi ba, geetha b. 2017. gcms and ftir analysis of bioactive compounds on methanol extract of ipomoea aquatic. international journal of fisheries and aquatic research 2(3): 12 – 16. i. u. nwaehujor et al. bioactivity of ethanol extract of annona muricata leaf ruhuna journal of science vol 11 (2): 131-142, december 2020 142 krishnamoorthy k, senguttuvan j, gurunathan a, arjunan v, krishnaswamy t, subramanian p. 2016. evaluation of anti-inflammatory and antioxidant properties of crude extract and forskolin from solena amplexicaulis leaf. indian journal of pharmaceutical sciences 78(3): 377-387. leelaprakash g, das sm. 2011. in-vitro anti-inflammatory activity of methanol extract of enicostemma axillare. international journal of drug development and research 3(3): 189 – 196. leon-fernandez ae, sayago-ayerdi sg, velazquez-estrada rm, zepeda-vallejo lg, yahia e, montalvo-gonzalez e. 2017. in-vitro antioxidant capacity of crude extracts and acetogenin fraction of soursop fruit pulp. pharmaceutica analytica acta 8(6): 550 – 556. doi: 10.4172/21532435.1000550. moghadamtousi sz, fadaeinasab m, nikzad s, mohan g, ali hm, kadir ha. 2015. annona muricata (annonaceae): a review of its traditional uses, isolated acetogenins and biological activities. international journal of molecular sciences 16: 15625-15658. nishaa s, vishnupriya m, sasikumar jm, hephzibah pc, gopalakrishnan vk. 2012. antioxidant activity of ethanolic extract of maranta arundinacea l. tuberous rhizomes. asian journal of pharmaceutical and clinical research 5(4): 85 – 88. nwaehujor i, olatunji g, atolani o, akande s. 2020. studies on in-vitro anti-inflammatory and antioxidant potentials of the leaf extracts of annona muricata. croatian journal of food science and technology 12(2). doi: 10.17508/cjfst.2020.12.2.xx in press. orak hh, bahrisefit is, sabudak t. 2019. antioxidant activity of the extracts of soursop (annona muricata l.) leaves, fruits pulps, peels and seeds. polish journal of food and nutrition sciences 69(4): 359 – 366. doi: 10.31883/pjfns/112654. perumal ss, ekambaram sp, dhanam t. 2017. in-vivo antiarthritic activity of the ethanol extracts of stem bark and seeds of calophyllum inophyllum in freund’s complete adjuvant induced arthritis. pharmaceutical biology 55(1): 1330 – 1336. steele ve, holmes ca, hawk et, kopelovich l, lubet ra, crowell ja, sigman cc, kelloff gj. 1999. review: lipoxygenase inhibitors as potential cancer chemopreventives. cancer epidemiology, biomarkers and prevention 8: 467 – 483. sumathi s, anuradha r. 2016. in-vitro anti-inflammatory activity of flower extract of couroupita guianensis aubl. international journal of herbal medicine 4(5): 05 – 08. usunobun u, okolie np, anyanwu og, adegbegi aj, egharevba me. 2015. phytochemical screening and proximate composition of annona muricata leaves. european journal of botany plant science and phytology 2(1): 18-28. ruhuna journal of science vol 13 (2): 141-149, december 2022 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v13i2.121 university of ruhuna faculty of science, university of ruhuna sri lanka 141 nutritive and sensory value assessment of smoked dried catfishes from two indigenous markets in benue state, nigeria ikyo benjamin chiaaondo1, okoroafor chinedu henry*1 and umawam samuel fanyam2 1 department of perishable crops research, nigerian stored products research institute, nigeria 2 department of fish post-harvest technology, benue state university, nigeria. *correspondence: okoroaforchinedu@gmail.com, orcid: https://orcid.org/0000-0001-7333-179x received: 5th november 2021, revised: 28th september 2022, accepted: 20th november 2022 abstract nutritional assessment of processed fish is needed to ensure that such products sold in the local communities are safe for human consumption. present study aimed at evaluating the nutritional quality of selected catfishes, namely, clarias gariepinus, heterobranchus spp and synodontis spp from abinsi and wadata fish markets in benue state, nigeria. samples of smoked catfish each weighing approximately 500 g were sourced from abinsi and wadata fish markets, and packaged using foil paper and polythene bags. the smoked fish samples were analysed to determine organoleptic qualities, proximate and mineral composition. the results of the study revealed a significant (p<0.05) variation in proximate and mineral composition of the three smoked catfishes between the two markets. the percentage moisture content for the three species of catfish ranged from 8.19 ± 0.031 to 10.34 ± 0.035%. synodontis spp. from abinsi market had the least moisture content (8.19 ± 0.031%) while heterobranchus spp. from wadata market had the highest moisture content (10.34 ± 0.035%). protein content of the three species ranged from 55.80 ± 0.060 to 68.97 ± 0.125%. c. gariepinus from wadata market had the highest protein content (68.97 ± 0.125%). the lipid content of the fish ranged from 10.37 ± 0.023 to 22.68 ± 0.035. the most abundant mineral was potassium (k) and was more abundant in c. gariepinus from wadata market which had the highest k content (410.15 ± 0.895 mg/100 g). heterobranchus spp. had the highest calcium (ca) content (395.48 ± 0.499 mg/100 g), sodium (na) (39.84 ± 0.045 mg/100 g) and zinc (zn) (0.80 ± 0.015 mg/100 g) while the highest concentration (12.0 ± 0.093 mg/100 g) of iron (fe) was recorded in synodontis spp. the concentrations of cu, zn, and fe in the three species across the markets were well above who permissible limits of 20-30 ppm, 30-100 ppm and 50-100 ppm respectively. the smoked catfishes retained good scores for taste, appearance, texture, and odour. however, there was significant (p < 0.05) differences in taste, appearance, texture, and odour for heterobranchus spp., c. gariepinus and synodontis spp in abinsi and wadata markets. keywords: mineral composition, organoleptic qualities, proximate analysis, smoked fish https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v13i2.121 https://creativecommons.org/licenses/by-nc/4.0/ https://orcid.org/0000-0001-7333-179x b. c. ikyo et al. nutritive and sensory assessment of smoked dried catfish ruhuna journal of science vol 13 (2): 141-149, december 2022 142 1 introduction fish and seafood products are very essential commodities in the international trade market. they constitute an essential part of a healthy meal due to their high protein content, low fat content and the presence of omega-3 fatty acidan essential amino acid (rhea 2009, pal 2010). about two-third of the world’s protein is derived from fishes and seafood products (emikpe et al. 2011). processing of fish and seafood products aims at inhibiting the growth of microorganisms, extending product shelf-life, and ensuring that these products have acceptable quality. several methods such as refrigeration, salt curing or smoking have gained wide application in preservation of fish and sea products. in nigeria, smoking is the oldest and commonest method of preservation, especially by rural inhabitants. it is a simple and affordable preservation method. it reduces the fish moisture content and inhibits microbial activity. smoke-drying technology gives fish a characteristic colour and flavour (alasalvar et al. 2011). it also extends the shelf life of the fish (eyo 2000, kumolu-johnsoh et al. 2010). the type of wood used during the smoke drying of fish is very important as this affects the quality of the fish (huong 2014). the main objective of food processing is to meet safety and acceptability standards, especially for consumers (pal 2012). to actualize this, high-level hygienic practices during processing and preservation are very essential. nutritional assessment of fish is an integral part of the fish industry and should be routinely carried out to meet the quality demands of fishmongers. most countries have quality standards that meet their local market needs. nutritional assessment of fish products will ensure that only smoke-dried fish of good quality is marketed to consumers. smoking of fish possibly elevates mineral compositions in the fish to beneficial or toxic levels. the method of smoking and length of exposure to the smoking process have been identified as essential factors that affect the quality and acceptability of the product (indranesa et al. 2000). hence, the need to access not only the physical properties (taste, odour, flavour etc.) of the fish product but also the mineral and proximate composition to ensure overall product quality. the aim of this study is to carry out a nutritional assessment of some selected catfishes, namely, clarias gariepinus, heterobranchus spp. and synodontis spp. from abinsi and wadata fish markets in benue state nigeria. 2 materials and methods 2.1 sample collection smoked dried catfish samples were randomly purchased from three different locations at abinsi and wadata markets. a total of eighteen samples were collected, six samples from each species of clarias gariepinus, synodontis spp. and heterobranchus spp. b. c. ikyo et al. nutritive and sensory assessment of smoked dried catfish ruhuna journal of science vol 13 (2): 141-149, december 2022 143 weighing approximately 500 g were sourced. the samples were collected and packaged separately using foil paper and polythene bags. thereafter, they were taken to the laboratory for analysis. 2.2 proximate analysis the proximate composition of the samples was determined according to association of official analytical chemists methods (aoac 2010). these analyses were conducted at nigeria stored product research institute (nspri) ilorin kwara state, nigeria. 2.3 organoleptic assessment a panel of ten (10) judges was selected from centre for food technology and research (cefter) community at random to assess the smoked dried catfish samples. a 9point hedonic scale ranging from 1(dislike extremely) to 9 (like extremely) was used (olayemi et al. 2011) for the assessment. the qualities assessed were appearance, odour, texture, and taste. 2.4 mineral determination for wet digestion of samples, the powdered sample was weighed (1 g) into a digesting glass tube, 12 ml of hno3 was then added and the mixture was left to stand overnight at 25 – 30 °c. perchloric acid (4 ml) was added to this mixture and then digested in the fume block. the temperature was gradually raised from 50°c to 300°c. the appearance of white fumes after about 90 min of digestion, brought the digestion to an end. after cooling, the digested samples were transferred to 100 ml volumetric flasks and made up to 100 ml with distilled water. this was then used for mineral determination. 2.5 determination of mineral elements determination of minerals by atomic absorption spectrometry (aas) was carried out at university of ilorin central research laboratories using model buck scientific accusys-211-aas. the standards for each element under investigation were prepared in parts per million (ppm) and the limit standard concentration for each element was carefully followed according to the standard operating procedure. the standard solutions were aspirated, and the graph was obtained. the concentrations of various metal elements in the samples were read and calculated using the equation below. b. c. ikyo et al. nutritive and sensory assessment of smoked dried catfish ruhuna journal of science vol 13 (2): 141-149, december 2022 144 concentration of element= 𝑐𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑠𝑎𝑚𝑝𝑙𝑒(𝑝𝑝𝑚) 𝑊𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑠𝑎𝑚𝑝𝑙𝑒 × 𝑑𝑖𝑙𝑢𝑡𝑖𝑜𝑛 𝑓𝑎𝑐𝑡𝑜𝑟 (100) 2.6 statistical analysis the data were subjected to analysis of variance (anova) using genstat (discovery version) and a significant test for differences between samples means was done using duncan multiple range (dmrt) test at 5% level of significance. 3 results and discussion 3.1 proximate compositions the highest moisture content (10.34 ± 0.035%) was recorded in heterobranchus spp. the highest protein (68.97 ± 0.125%) and lipid content (22.68 ± 0.035%) were recorded in clarias gariepinus and synodontis spp respectively (table 1). significant (p<0.05) differences between the two markets occur in c. gariepinus for moisture contents, in all three species for ash content and lipid content, in c. gariepinus for protein content, and in heterobranchus spp. for carbohydrate content. the protein content values of the heterobranchus spp. c. gariepinus, and synodontis spp. from both markets showed high values and variation in the markets (table 1). table 1: proximate compositions (%) of smoked catfish obtained from abinsi and wadata markets in benue state, nigeria market sample moisture ash protein lipid carbohydrates abinsi heterobranchus spp. 10.16 ± 0.012d 12.74±0.049d 58.98±0.110b 16.42±0.081d 1.71±0.018b clarias gariepinus 9.09 ± 0.015c 10.10±0.029a 66.11±0.029c 13.94±0.041b 0.76±0.041a synodontis spp 8.19 ± 0.031a 11.07±0.035b 55.80±0.060a 22.68±0.035f 2.27±0.054cd wadata heterobranchus spp. 10.34 ± 0.035d 11.89±0.055c 59.48±0.067b 15.71±0.029c 2.59±0.032d clarias gariepinus 8.88 ± 0.049b 11.39±0.032b 68.97±0.125d 10.37±0.023a 0.38±0.015a synodontis spp 8.36 ± 0.026a 12.19±0.061c 56.35±0.033a 21.30±0.044e 1.81±0.041bc means ± standard error on the same column with different superscripts differs significantly at p < 0.05 level. 3.2 elemental composition of smoked catfishes from abinsi and wadata markets significant (p<0.05) difference was recorded for k, ca, zn and fe content of heterobranchus spp, clarias gariepinus and synodontis spp across the two markets (table 2). the highest potassium content (410.15 ± 0.895) was recorded in c. gariepinus from wadata market while the least potassium content (240.33 ± 0.050) was recorded in heterobranchus spp from abinsi market. heterobranchus spp from b. c. ikyo et al. nutritive and sensory assessment of smoked dried catfish ruhuna journal of science vol 13 (2): 141-149, december 2022 145 abinsi market had the highest calcium content (395.48 ± 0.499) while the same species from wadata market had the lowest calcium content (127.88 ± 0.792) (table 2). the highest fe content (12.0 ± 0.093) was recorded in synodontis spp from wadata market while the least fe content (4.95 ± 0.052) was recorded in heterobranchus spp from wadata market. table 2: elemental composition of smoked fish species obtained from abinsi and wadata fish markets, benue state, nigeria means ± standard error on the same column with different superscript differs significantly at p<0.05 level 3.3 organoleptic assessment from the organoleptic scores, heterobranchus spp. and clarias gariepinus from wadata market were preferred over heterobranchus spp. and c. gariepinus from abinsi market while synodontis spp. from abinsi were preferred over synodontis spp. from wadata market (table 3). table 3: sensory evaluation of smoked-dried catfish from abinsi and wadata markets, benue state, nigeria market samples taste appearance texture odour abinsi heterobranchus spp. 6.23 ± 0.032c 6.67 ± 0.021c 6.97 ± 0.040c 6.43 ± 0.032c clarias gariepinus 5.38 ± 0.020a 6.22 ± 0.053b 6.92 ± 0.024bc 6.23 ± 0.032b synodontis spp. 8.50 ± 0.026f 7.67 ± 0.020e 7.52 ± 0.031e 6.93 ± 0.052e wadata heterobranchus spp. 7.63 ± 0.046e 7.33 ± 0.035d 7.43 ± 0.035d 6.92 ± 0.053e clarias gariepinus 7.38 ± 0.021d 7.63 ± 0.050e 7.33 ± 0.047d 6.77 ± 0.040d synodontis spp. 5.73 ± 0.050b 5.58 ± 0.032a 6.63 ± 0.041a 5.92 ± 0.367a means ± standard error on the same column with different superscript differs significantly at p < 0.05 level. market sample mineral concentration in mg/100g k ca na cu zn fe abinsi heterobranchus spp 240.33 ± 0.050a 395.48 ± 0.499e 32.44 ± 0.035a 0.03 ± 0.009a 0.80 ± 0.015e 5.37 ± 0.021a clarias gariepinus 270.30 ± 0.240c 278.05 ± 0.954d 32.14 ± 0.113a 0.21 ± 0.012b 0.30 ± 0.012a 5.06 ± 0.055a synodontis spp. 330.05 ± 0.421e 257.78 ± 0.093c 34.24 ± 0.032c 0.24 ± 0.017bc 0.40 ± 0.029b 6.98 ± 0.058c wadata heterobranchus spp 250.30 ± 0.643b 127.88 ± 0.792a 39.84 ± 0.045d 0.39 ± 0.186d 0.68 ± 0.030d 4.95 ± 0.052a clarias gariepinus 410.15 ± 0.895f 159.33 ± 0.891b 33.03 ± 0.064b 0.22 ± 0.015b 0.60 ± 0.023c 5.96 ± 0.070b synodontis spp. 300.45 ± 0.480d 295.48 ± 1.24d 33.93 ± 0.057c 0.30 ± 0.038c 0.71 ± 0.029d 12.0 ± 0.093d b. c. ikyo et al. nutritive and sensory assessment of smoked dried catfish ruhuna journal of science vol 13 (2): 141-149, december 2022 146 a significant (p<0.05) difference occurred between heterobranchus spp., c. gariepinus and synodontis spp. in abinsi and wadata for taste, appearance, texture, and odour and this might be due to variations among individuals in responding to the same level of stimuli such as appearance and taste. 4 discussion the percentage moisture content of synodontis spp. from abinsi market had the least value of moisture content while heterobranchus spp. from wadata market had the highest moisture content. the lower moisture content recorded in synodontis spp. shows that it may have a good keeping quality, as a study on the influence of traditional smoke drying on the quality of fish by ali et al. (2011) revealed that smoked fish with good keeping quality had the least moisture content. fish is an essential part of a balanced diet because it contains complete protein and some vital polyunsaturated fatty acids (polak-juszczak and adamczyk 2009, polakjuszczak and komar-szymczak 2009). protein content was slightly higher in clarias gariepinus from both markets, each recording 66.11 ± 0.029% and 68.97 ± 0.125% respectively. these values are higher than the value of 53.10% reported by ogbonna and ibrahim (2009) but concur with 68.40% reported by olayemi et al. (2011). the fat content values of the fishes were within the range of 10.37 ± 0.023 to 22.68 ± 0.035% and this concurs with the findings of ogbonna and ibrahim (2009), and olayemi et al. (2011). these values indicate that smoking had no adverse effect on the fat content of the catfish species examined for the study. the relatively high ash content values of the fishes could be because of the high drying temperature and enclosed system of drying which is associated with high ash values (olayemi et al. 2011). this can also reflect the high mineral constituent of the fish (liu 2019). very low carbohydrate content was recorded in the fishes, with heterobranchus spp. having the highest value of 2.59 ± 0.032% and clarias gariepinus having the least value (0.38 ± 0.015 %), both from wadata markets. the significantly high composition of certain minerals (k, ca, na) reveals that these smoke-dried fishes are a very good source of these vital minerals. these results concur with the findings of adewoye and omotosho (1997), prapasri et al. (1999), and ricardo et al. (2002) who also reported high concentrations of these minerals in fish. generally, varying concentrations of minerals in fishes have been reported in similar studies and this could be because of differences in chemical characteristics of water from which the fishes were sampled, rate of absorption, the season in which the studies were conducted as well as different dietary patterns (yilmaz 2003, papagiannis et al. 2004, ahmed et al. 2010, opaluwa et al. 2012, emurotu et al. 2014, mohammed and osman 2014). freshwater fish is a valuable source of iron, copper, and zinc (fao 2014). the concentrations of cu, zn and fe in the three fishes were well above who permissible limits of 20-30 ppm, 30-100 ppm and 50-100 ppm respectively (fao/who 1989, b. c. ikyo et al. nutritive and sensory assessment of smoked dried catfish ruhuna journal of science vol 13 (2): 141-149, december 2022 147 who 2000, european community 2005, mokhtar 2009). however, the concentration of cu in heterobranchus spp from abinsi market was within fao/who acceptable limit. excess copper in the human blood causes hematemesis, hypotension, melena, coma, jaundice, and gastrointestinal distress (oti-wilberforce et al. 2016). zinc although essential for normal growth and development can be deleterious at high concentrations (oti-wilberforce et al. 2016). high toxicity of this element in the blood can lead to gross deficiencies in copper and iron, abdominal pain, bloody enteritis, paralysis of extremities, and lowered leukocyte count (atsdr 1994). excess iron in human blood may lead to hemochromatosis which increases the susceptibility of the individual to cancer and heart disease (atsdr 1999). this raises alarm as regards the potential danger of continual consumption of these fishes without proper screening by relevant stakeholders in the aquatic industry. reports of sensory evaluation from the panelists revealed that the smoked catfishes from both markets retained very good scores for taste, appearance, texture, and odour. this shows that the products were accepted or liked, since the least score for all the organoleptic indices examined was > 5.38. however, there was a significant difference between taste, appearance, texture, and odour of the smoked catfish samples that were subjected to organoleptic assessment. the preference in taste, texture and appearance could be attributed to the processing method (smoking). the result of the sensory evaluation reveals that high concentrations of cu, zn and fe in the fishes may not have affected the sensory properties of the fishes. this makes it more dangerous to the consumer as elevated levels of these minerals are not easily noticed by the sensory organs, albeit very dangerous to human health when ingested. 5 conclusions the proximate and mineral composition of the smoked catfishes showed variations from market to market, however, the three catfish species from this study were found to be rich in protein, potassium, calcium, and sodium with low presence of lipid and carbohydrate contents. high concentrations of copper, zinc and iron was recorded in all three species across the markets, revealing the potential risk of these products to consumers. the sensory evaluation of hetrobraunchus spp, clarias gariepinus and synodontis spp from both markets had acceptable quality from the panelists. key stakeholders in the aqua feeds industry should constantly screen fishery products sold in local markets to ensure that these products meet international standards and are safe for consumption. acknowledgments the authors are profoundly grateful to m.k. abdulbaki for analysing the results. we also want to thank the anonymous reviewers for their immense contribution to the improvement of the quality of the manuscript. b. c. ikyo et al. nutritive and sensory assessment of smoked dried catfish ruhuna journal of science vol 13 (2): 141-149, december 2022 148 references adewoye so, omotosho js. 1997. nutrient composition of some freshwater fishes in nigeria. bioscience research community 11(4): 333-336. agency for toxic substances and disease registry (atsdr). 1999. toxicological profile for lead. agency for toxic substances and disease registry, us department of health and human services, public health services. 205: 593-606. agency for toxic substances and disease registry (atsdr), 1994. ahmed a, dodo a, bouba am, clement s, dzudie t. 2010. determination of minerals and heavy metals in water, sediments and three fish species (tilapia nilotica, silurus glanis and arius parkii) from lagdo lake, cameroun. journal of fisheries international 5(3): 54-57. doi: 10.3923/jfish.2010.54.57. akinwumi fo. 2014. effects of smoking and freezing on the nutritive value of african mud catfish, clarias gariepinus burchell, 1822. the journal of agricultural science 6(11): 143 – 149. doi: 10.5539/jas.v6n11p143. alasalvar c, miyashita k, shahidi f, wanasundara u. 2011. handbook of seafood quality, safety and health applications page, john wiley & sons. isbn 9781444347760 ali a, amodou d, mohamadou ba, saidou c, tenin d. 2011. influence of traditional drying and smoke drying on the quality of three fish species (tilapia niloticus, silurusglanis sp. and arisparkii sp.) from lagdo lake, cameroon. journal of animal and veterinary advances 10(3): 301306. doi: 10.3923/java.2011.301.306. aoac. official methods of analysis (18th edition). 2010. association of official analytical chemists international maryland, u.s.a. ekeocha ca, agbabiaka la, amadi sa. 2010. storage and organoleptic qualities of clarias gariepinus smoked with woods from oil been and oil palm trees. animal production research advances 6(3): 231–233. emikpe bof, adebisi t, adedeji ob. 2011. bacteria load on the skin and stomach of clarias gariepinus and oreochromis niloticus from ibadan, southwest nigeria: public health implications. journal of microbiology and biotechnology research 1(1): 52-59. emurotu, j.e., yahaya, a. and adegbe, a.a. 2014. determination of some mineral elements in fresh and smoked fish of tilapia (oreochromis noliticus) and catfish (clarius gariepinus) from ibaji, kogi state, nigeria. scientia africana 13 (2): 30-35. ec (european community) 2005. commission regulation no 78/2005 (pp: 16/43eli6/45). official journal of european union [20.1.2005]. eyo aa. 2000. fish processing technology. national workshop on fish processing storage, marketing and utilization techniques, 1-5. fao/who. 1989. evaluation of certain food additives and the contaminants mercury, lead and cadmium. who technical report series 505. fao 2014. sustainable fisheries and aquaculture for food security and nutrition, rome: food and agriculture organization,. 119. huong, dtt. 2014. the effect of smoking methods on the quality of smoked mackerel. united nations university fisheries training programme, iceland [final project]. http://www.unuftp.is/static/fellows/document/huong13prf.pdf indranesa wm, hansen lt, gill ta. 2000. effect of cold-smoking and drying on the textural properties of farmed atlantic salmon (salmo salar). journal of aquatic food. production technology 9(1): 4764. doi: 10.1300/j030v09n0105. kumolu-johnsoh ca, aladetohun nf, ndimele pe. 2010. the effects of smoking on the nutritional qualities and shelf-life of clarias gariepinus. african journal of biotechnology 9(1): 73-76. liu k. 2019. effects of sample size, dry ashing temperature and duration on determination of ash content in algae and other biomass. agal research 40: 101486. doi: 10.1016/j.algal.2019.101486. b. c. ikyo et al. nutritive and sensory assessment of smoked dried catfish ruhuna journal of science vol 13 (2): 141-149, december 2022 149 mohammed eha, osman ar. 2014. heavy metals concentration in water, muscles, and gills of oreochromis niloticus collected from the sewage-treated water and the white nile. international journal of aquaculture 4(6):36-42. doi: 10.5376/ija.2014.04.0006. mokhtar m. 2009. assessment level of heavy metals in penaeus monodon and oreochromis spp. in selected aquaculture ponds of high densities development area. european journal of scientific research 30(3): 348e60. pal m. 2010. fish hygiene. msc lecture notes. addis ababa university, faculty of veterinary medicine, debre zeit, ethiopia. 1-11. pal m. 2012. food spoilage. ph.d. lecture notes. addis ababa university, college of veterinary medicine, debra zeit, ethiopia. l-8. polak-juszczak l, adamczyk m. 2009. quality and amino acid composition of protein of fish from the vistula lagoon. food. science technology quality 16(3): 75–83. polak-juszczak l, komar-szymczak k. 2009. fatty acid profiles and fat contents of commercially important fish from vistula lagoon. polish journal of food and nutrition sciences 59(3): 225–229. ogbonna c and ibrahim mb. 2009. effect of drying methods on proximate compositions of catfish (clarias gariepinus). world journal of agricultural sciences 5(1): 114-116. olayemi ff, adedayo mr, bamishaiye ei, awagu ef. 2011. proximate composition of catfish (clarias gariepinus) smoked in nigerian stored products research institute (nspri): developed kiln. international journal of fisheries and aquaculture 3 (5): 96 – 98. onyia lu, milam c, manu jm, allison ds. 2010. proximate and mineral composition in some freshwater fishes in upper river benue, yola, nigeria. continental journal of food science and technology 4:1–6. doi: 10.5281/zenodo.834679. opaluwa od, aremu mo, ogbo lo, magaji ji, odiba ie. 2012. assessment of heavy metals in water, fish, and sediments from uke stream, nasarawa state, nigeria. current world environment 7(2):213-220. doi: 10.12944/cwe.7.2.04. osibona a, kusemiju k, akande gr. 2006. proximate composition and fatty acids profile of the african catfish clarias gariepinus. journal of the life and physical sciences 3: 85-89. oti-wilberforce jo, nworie f, oti-wilberforce ro. 2016. comparative assay of heavy metals in gills and muscles of common fishes in afikpo, ebonyi state, nigeria. world applied science journal 34(2): 268-272. doi: 10.5829/idosi.wasj.2016.34.2.15657. papagiannis i, kagalou i, leonardos j, petridis d, kalfakaou v. 2004. copper and zinc in four freshwater fish species from lake pamvotis (greece). environment international 30(3): 357-362. doi: 10.1016/j.envint.2003.08.002. prapasri p, kunchit j, eakkarach k, kriengkrai v, yupaporn n, lalita b. 1999. proximate composition of raw and cooked thai freshwater and marine fish. journal of food composition and analysis 12: 916. doi: 10.1006/jfca.1998.0800. rhea f. 2009. microbiology handbook: fish and seafood. leatherhead food international ltd. surrey, uk. ricardo cm, cyrino jep, portz l, trugo lc. 2002. effect of dietary lipid level on nutritional performance of the surubim, pseudoplatystoma coruscans. aquaculture 209: 209-218. sadiku soe, oladimeji aa. 1991. relationship of proximate composition of lates niloticus (l). synodontis schall. research communications 3(1): 29-40. talabi o. 1995. the influence of traditional handling methods on the quality of processed fish in nigeria. proceeding of the conference on the handling, processing, and marketing of tropical fish. tropical product institute, london. yilmaz ab. 2003. levels of heavy metals (fe, cu, ni, cr, pb and zn) in tissues of mugil cephalus and trachurus mediterraneus from iskenderun bay, turkey. environmental research 92(3):277-281. doi: 10.1016/s0013-9351(02)00082-8. world health organization. 2000. the state of world fisheries and aquaculture. rjs-vol-1-sept-2006-12.dvi ruhuna journal of science vol. 1, september 2006, pp. 113–124 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. inferencing design styles using bayesian networks aruna lorensuhewa*, binh pham, and shlomo geva centre for information technology innovation, faculty of information technology, queensland university of technology, brisbane, australia, s.lorensuhewa, b.pham, s.geva@qut.edu.au, reasoning with uncertain knowledge and belief has long been recognized as an important research issue in artificial intelligence (ai). several methodologies have been proposed in the past, including knowledge-based systems, fuzzy sets, and probability theory. the probabilistic approach became popular mainly due to a knowledge representation framework called bayesian networks. bayesian networks have earned reputation of being powerful tools for modeling complex problem involving uncertain knowledge. uncertain knowledge exists in domains such as medicine, law, geographical information systems and design as it is difficult to retrieve all knowledge and experience from experts. in design domain, experts believe that design style is an intangible concept and that its knowledge is difficult to be presented in a formal way. the aim of the research is to find ways to represent design style knowledge in bayesian networks. we showed that these networks can be used for diagnosis (inferences) and classification of design style. the furniture design style is selected as an example domain, however the method can be used for any other domain. key words : bayesian networks, classification, design style, svm, c4.5, data mining 1. introduction uncertain and fuzzy knowledge exists in domains where it is difficult to retrieve all knowledge and experience from experts. in the area of design, some experts believe that design style is an intangible concept and that its knowledge is difficult to be presented in a formal way. so far there has been no computer supported automatic techniques to assist novice designers in learning to distinguish design styles, to judge how similar a design is to a specific style or to learn suitable features for a required design by given a subset of features. machine learning and data mining techniques have been used to automatically extracted knowledge from unstructured information sources. commonly-used algorithms include c4.5 (quinlan 1993), support vector machines (svm) (joachims 1999), nearest neighbor and neural networks. in our previous paper (lorensuhewa 2003), we showed how these techniques can be used for classification task. reasoning with uncertain knowledge and belief has long been recognized as an important research issue in ai. several methodologies have been proposed in the past, including knowledge-based systems, fuzzy sets, and probability theory. the probabilistic * permanent address: department of computer science, university of ruhuna, matara, sri lanka. 113 lorensuhewa, pham, and geva: inferencing design styles... 114 ruhuna journal of science 1, pp. 113–124, (2006) figure 1 web based questionnaire of the experiment. approach became popular mainly due to a knowledge representation framework called bayesian networks (bns). bns provide a graphical representation of independencies amongst random variables (korb 2003). a bn is a directed acyclic graph (dag) with nodes representing random variables and arc representing direct influence. the independence that is encoded in a bn means that each variable is independent of its non-descendents given its parents. to quantify the strengths of these associations, each node has a conditional-probability table that captures these associations among those random variables. bns have been introduced in 1980s as formalism for representing and reasoning with models of problems involving uncertainty, adopting probability theory as a basic framework (pearl 1988). since the beginning of 1990s, researchers are exploring its capabilities for developing medical applications. the bn formalism offers a natural way to represent the uncertainties involved in medicine when dealing with diagnosis, treatment selection, planning, and prediction of prognosis. similarly, we can use bn technique in design style domain. the aim of this research is to develop a generic framework and methodologies that will enable the extraction of domain knowledge and information. we then represent this knowledge in bn and integrate with other expert knowledge and inference with the resulting model according to user requirements. we collected an experimental data set for six different styles: chippendale, chippendale, jacobean, queen anne, sharon and william and marry. a web based questionnaire (figure 1) is used to collect data from users and domain experts. in total, sixteen different features were examined and seven features were selected, including leg type, back shape, foot, etc. we also make use of the connectedline furniture design style guide database (connectedlines 1998) which is a commercially available database. this guide identifies and dates about 20 furniture styles and their distinctive features. this database is used as a source for expert knowledge. lorensuhewa, pham, and geva: inferencing design styles... ruhuna journal of science 1, pp. 113–124, (2006) 115 the remainder of the paper is organized as follows: section 2 provides the background for bayesian network methodology. section 3 describes how to construct bayesian network from data for design style domain and how to use that for style classification. the method of integrating expert knowledge and how to increase the total classification accuracy is presented in section 4. section 5 discusses how to use bayesian networks to inference about design style domain. the conclusion is given in the final section. 2. bayesian networks bayesian networks (or belief networks, casual networks) provide a graphical model for probabilistic relationships among a set of variables allows us to represent and reason about uncertain domains. the nodes in a bn represent a set of random variables from the domain. random variables can be discrete or continuous. a set of directed arcs (or links) connects each pair of nodes, representing direct dependencies between variables (or in other word, casual relationships between variables). an arrow from node x to node y means x is the parent of y and x has a direct influence on y. each node has a conditional probability (for discrete variables) table to represent the strength of the relationships between variables. each root node has a prior probability table. the only constraint on the arcs allowed in a bn is that there must not be any directed cycles. 2.1. structure of the bayesian networks most commonly, bns are considered to be the representations of joint probability distributions. there is a fundamental assumption that there is a useful underlying structure to the problem being modeled that can be captured with a bn, i.e., not every node is connected to every other node. if such a domain structure exists, a bn gives a more compact representation than simply describing the probability of every joint instantiation of all variables. the bn with few arcs or less parents for each node (sparse bns) represents probability distributions in a computationally tractable way. consider a bn containing the n nodes, x1 to xn, taken in that order. a particular value in the joint distribution is represented by p (x1 = x1, x2 = x2, . . . , xn = xn) , or more compactly, p (x1, x2, . . . , xn). the chain rule of probability theory allows us to factorize joint probabilities so that p (x1, x2, . . . , xn) = p (x1) × p (x2|x1) × . . . × p (xn|x1, . . . , xn−1) = πip (xn|x1, . . . , xn−1) the structure of a bn implies that the value of a particular node is conditional only on the values of its parent nodes. this reduces to: p (x1, x2, . . . , xn) = πip (xi|p arents(xi)) provided that p arents(xi) ⊆ {x1, . . . , xi−1} lorensuhewa, pham, and geva: inferencing design styles... 116 ruhuna journal of science 1, pp. 113–124, (2006) 2.2. building a bayesian network there are a number of steps that a knowledge engineer must undertake when building a bayesian network. these steps are: identifying the variables and its relevant values, finding the topological structure and specifying the conditional probability tables (kord (2003)). identifying nodes and values: firstly the variables of interest in the domain must be identified. in other words, the knowledge engineer needs to identify what the nodes represent and what values they take. in the case of discrete nodes, the following types can exist: boolean nodes, ordered values or integer values. finding the structure: the structure of the network represents causes and effects relationships. it captures qualitative relationships between variables. in particular, two nodes should be connected directly if one affects or causes the other, with the arc indicating the direction of the effect. usually expert knowledge is used to build the structure. for example, in the case of the medical diagnosis example, the knowledge engineer might ask the question “what factors affect a patient’s chance of having cancer?” if the answer is “pollution and smoking”, then arcs are added from the pollution and smoker node to the cancer node. if the expert knowledge is not available, the structure should learn in another way. learning the structure from data is a commonly used method for such a situation. specifying conditional probabilities: once the structure of the bn is determined, the next step is to quantify the relationship between connected nodes. in the case of discrete variables, this is done by specifying conditional probability table (cpt) for each node. 2.3. learning structure and parameters bn were originally developed as knowledge representation formalism, with human experts their only resource. during the late 80’s, people realized that the statistical foundations of bayesian makes it possible to learn from data rather than from experts (korb (2003)), and started to use both sources. learning a bn means to learn the graphical model of dependencies (structure) and the conditional probability distributions (parameters). there are two general approaches for constructing a graphical probabilistic model learning from data: the search and scoring methods and the dependency analysis methods (cheng 1997). in the first approach, the algorithms view the learning problem as to search for a structure that can fit the data best. these algorithms start without any edges, and then use some search technique to add edges to the graph. after each change, they use some scoring scheme to compare the new and old structures. if the new structure is better than old one, they keep the newly added edge and another one. this process continues until the best structure is found. several scoring methods have been used, namely bayesian scoring method (cooper (1992)) and heckerman (1995)), entropy based method and minimum description length method (bouckaert (1994)). most of these algorithms need node ordering as the search space is very large. lorensuhewa, pham, and geva: inferencing design styles... ruhuna journal of science 1, pp. 113–124, (2006) 117 in the second approach, the learning problem is viewed differently. the bn structure encodes a group of conditional independence relationships among the nodes, according to the concept of d-separation (pearl (1988)). this suggests a way of learning the bn structure by identifying the conditional independence relationships among the nodes. these algorithms try to discover dependencies from the data, and then use these dependencies to infer the structure. the dependency relationships are measured by using a statistical test such as chi-squared and mutual information tests. this type of algorithms are refereed to as ci-based algorithms or constraintbased algorithms (cheng (1997)). the parameters for a given structure can be learned by simply using the empirical conditional frequencies from the data by using corpus of complete data. 2.4. inference in bayesian networks the basic task for any probabilistic inference system is to compute the posterior probability distribution for a set of query variables, given some observed events-that is, some assignment of values to a set of evidence variables. this task is called belief updating or probabilistic inference. inference in bayesian network is very flexible and useful for most of the application domains. in bayesian inference, evidence can be entered about any node while beliefs in any other nodes are updated. there are mainly two classes of existing inference algorithms, exact and approximate inference. different algorithms are suited for different network structures and performance requirements. we have selected the most popular exact inference algorithm called junction tree (jensen 1994). the junction tree algorithm is a process in two steps, transformation and propagation. the transformation builds an undirected junction tree from the bayesian network. the second step, propagation, is where we propagate received evidence and make inference about variables using only the junction tree. propagation is done by using message passing. this is implemented in bayes net matlab toolbox (murphy (2002)) which we have used to conduct our experiments. 3. construction of bayesian networks for style classification this section describes how bns are used as a knowledge representation tool to represent furniture design style knowledge. it also explains how bns can be used for classification as well as for inference. the bayes net matlab toolbox (murphy (2002)) was used for implementation. the furniture design style domain is selected as an example domain for this experiment and two different datasets were used for the experiment. the first dataset is collected from experts and users and the second dataset is artificially generated from connectedline database. 3.1. data preparation data is selected for six different styles: chippendale, classical, jacobean, sharon, queen anne and william and marry. a web-based questionnaire (figure 1) was used to collect data from users and domain experts. in total, sixteen different features lorensuhewa, pham, and geva: inferencing design styles... 118 ruhuna journal of science 1, pp. 113–124, (2006) table 1 listing of all the features and their encoded numbers feature encoded numbers for each feature type appearance 1,2,3,4,5 proportion 1,2,3 chair arms 1,2,3,4,5,6,7,8 back material 1,2,3,4,5 leg type 1,2,3,4,5,6,7 seat shape 1,2,3,4 foot 1,2,3,4,5,6,7,8,9,10,11,12,13,14 back shape 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20 seat material 1,2,3,4,5 upholstery 1,2,3,4 underbracing 1,2,3 leg shape 1,2 leg carved 1,2 leg curved 1,2 leg fluted 1,2 style 1,2 ,3,4,5,6 table 2 the features of furniture design styles as described in connecetdline database feature jacobean william an marry queen anne chippendale hepplewhite sharon leg type 4,3 1,2,3 1 1,3 3,5 3,4,5,6,7 seat shape 1 1 2,3 1,3,4 3 2,3 foot type 1,7,14 9,10,14 2,11,12,13,14 11,8,7.2 8,7,3 1,2,3 back shape 1,11 12 16,18,19 2,13,17 4,3 1,2,5,6,12 leg shape 1,2 1 1 1 1,2 1,2 leg fluted 1 1 1 1 2 2 were examined. these features are listed in table 1, together with their encoded numbers. each number in this table has a symbolic meaning. for example, foot type 1 is lion leg, back shape 16 is fiddle back, etc. only subset of most significant features was selected for the experiment. in this paper, we will be labeling this data set as the ‘observed’ dataset. the connectedline furniture design style guide database (connectedlines (1998)) is commercially available software for the windows platform. this guide identifies and dates about 20 furniture styles and their distinctive features. this database is used to create an expert bn for this research. in the connectedline database features of each style is described as in table 2. there are many features described in the database, but we only selected the most important features which intersect with the features in the experimental dataset. from this feature table, all the combinations of features in a specific style were artificially generated and stored in a database. some design styles generated less amount of data because of a smaller number of corresponding combination of features. to generate an equal size of samples from each style, smaller samples were increased by duplicating each data instances in the samples several times. in this paper, we will call this data set the ‘expert’ data set. lorensuhewa, pham, and geva: inferencing design styles... ruhuna journal of science 1, pp. 113–124, (2006) 119 figure 2 an algorithm to search for the best node order by using k2 algorithm and bayesian score. 3.2. learning bayesian network for furniture style domain we used the algorithm given by the cooper and herskovits (cooper (1992)) to find the network structure from the observed dataset. cooper and herskovits showed that the best model, which maximizes the posterior probability of the network structure given the data, p (bs|d) also maximizes the joint probability, p (bs|d). they derived a function for this joint probability by using frequency of variable instantiation in the dataset. this function gives a direct way of measuring the goodness of fit between the model structure and the dataset and therefore it defines the fitness function (bayesian score) for the model. they use this fitness function to search the model space to find the good network structure. k2 requires fixed ordering of nodes, such that only the ‘upstream’ nodes of a node are considered as its candidate parents. as we do not know the node order for the case of furniture design domain, we need to find out the correct node order to use the k2 to find the structure. we permutated all possible node order sequence and then sequentially search for the best order. we used the k2 algorithm to find a best structure for a given order sequence. the bayesian score is used to compare two consecutive structures to find the best one. this simple technique of searching the best structure is given in algorithm 1. different methods such as genetic algorithms can be used to find the best structure efficiently when the search space is bigger. after learning the structure, other parameters are also learned from the dataset. two bayesian networks have learned from this method for the research experiments. the first bn has learned from the ‘observed’ data and it is labeled the ‘observed’ bn. the second bn has learned from ‘expert’ dataset and it is called the ‘expert’ bn. lorensuhewa, pham, and geva: inferencing design styles... 120 ruhuna journal of science 1, pp. 113–124, (2006) figure 3 bayesian network structure for fold 1. 3.3. experimental details and summary of results experiments were conducted with some subsets of features (seven features). the size of the ‘observed’ dataset is 120 instances. the ‘observed’ bn is constructed from ‘observed’ dataset and its accuracy is measured from the classification power of the bn. the target variable is the style node and the remaining six nodes are evidence nodes. cross validation is often used to estimate the generalization ability of a classifier where the amount of data is insufficient. under cross validation, the available data is firstly divided into k disjointed sets. k models are then trained (in this case, k bns and k cpds), each one with different combination of (k-1) partitions and tested on the remaining partition. throughout our experiments, we used 10-fold cross validation. training dataset in each fold was used to find the best structure from algorithm 1 and parameters (cpds) for the selected structure. instances in the test dataset which are relevant to the same fold are classified using the relevant bn. the bayesian network structure relevant for fold one is given in figure 2. the ‘expert’ bayesian network is constructed from the ‘expert’ dataset by a similar method. this bn is tested with the ‘observed’ dataset. the structure of the ‘expert’ bayesian network is shown in figure 3. 3.3.1. comparison with the other techniques the experiments were conducted with three other techniques, svm, nearest neighbour and c4.5 with the same ‘observed’ dataset using all the sixteen attributes. the original dataset with nominal attributes was tested with the nearest neighbour and c4.5 classifiers. the nearest neighbour classifier used exact matching as a distance metric. a windowsbased software implementation of c4.5 (see 5) was used as a decision tree classifier. lorensuhewa, pham, and geva: inferencing design styles... ruhuna journal of science 1, pp. 113–124, (2006) 121 figure 4 bayesian network structure for ‘expert’ bn. in both cases 10-fold cross validation is used for validation. the svm classifier was used with the binary coded dataset. the original categorical (nominal) dataset with sixteen attributes is converted to a binary dataset with 100 binary attributes excluding the class attribute which has 6 different class numbers. in this case, six different binary classifiers were trained to classify six different styles separately. each classifier determined if a given style attributes belonged to the corresponding style or not (binary classification). the classification decision for the entire ensemble of classifiers was based on the classifier giving the maximum output value (largest margin). in all three cases, ten-fold cross validation was used for validation. the summary of results is given in the table 3. table 3 comparison of classification accuracies of different techniques classifier data type classification accuracy observe bayesian network categorical (7 attributes) 73.72±14.21 expert bayesian network categorical (7 attributes) 68.31±10.32 nearest neighbour categorical (16 attributes) 85.59±9.49 svm binary coded (100 attributes) 88.75±5.73 c4.5 categorical (16 attributes) 76.50±3.70 4. combining two bayesian networks to improve overall classification accuracy there are several methods of combining multiple classifiers (in this case two bn classifiers). most commonly used technique is voting. voting counts the number of classifiers that agree in their decision and accordingly decides the class to which the input pattern belongs. in this case, no accuracy or expertise is considered. according to our experimental results, the ‘observed’ bn is more accurate than expert bn. the ‘observed’ bn obtained 73% accuracy and the expert bn 68% accuracy. the table 4 shows how both these networks behaved when they were tested separately using same test dataset. from the table 4, we can observe that, 36% (20.95±15.24) of the time both these network classifiers do not agree each other and only one is correct. it is difficult to lorensuhewa, pham, and geva: inferencing design styles... 122 ruhuna journal of science 1, pp. 113–124, (2006) figure 5 an algorithm to combine two bayesian networks according to past experience. table 4 behaviour of the two classifiers while classifing unseen test data percentage of cases both ‘observed’ bn and ‘expert’ bn are predicting correctly 52.38% percentage of cases bath ‘observed’ bn and ‘expert’ bn are not predicting correctly 11.43% percentage of cases only ‘observed’ bn is predicting correctly 20.95% percentage of cases only ‘expert’ bn is predicting correctly 15.24% use the simple voting method to increase the overall accuracy by combining both bayesian networks as the ‘observed’ bn is more accurate than the ‘expert’ bn. by analyzing the past results in table 4, we construct following procedure (algorithm 2) to combine the two classifiers to get a joint decision while classifying particular instances. the task of finding a suitable threshold value t for maximum classification accuracy is an important one. to determine a suitable threshold value t, we conducted experiments with various values of t and measured the overall accuracy of the combined bn classifier using algorithm 2. t lies between 0∼1. we plotted a graph of threshold value (t) vs classification accuracy of combined bayesian network (fig 4) and found that a threshold value range 0.35∼0.4 for t gives the best accuracy of 83.64 ±8.35 for the combine network classifier. 5. using bn to obtain domain knowledge given a complete bn model and defining both the structure and the conditional probabilities, we can begin to make prediction for any variable. if the values of some variables known (‘observed’), then the probabilities of the remaining variables (‘target’) can be calculated. in previous section we used bn for design style classification. in the style classification, we used the style variable as the target variable and all the other variables as observation variables. the same bn can be used further to learn about the furniture design style domain. for example, it can be used to learn the correct combination of parts for a specific design, or to learn possible suggestion for a missing (unknown) part of a particular design. in the ‘observed’ or ‘expert’ bayesian networks, we can define any node as a target node and all the other nodes as evidence (observe) nodes. for example, to complete a lorensuhewa, pham, and geva: inferencing design styles... ruhuna journal of science 1, pp. 113–124, (2006) 123 figure 6 the graph of threshold value vs classification accuracy of combined bayesian networks. table 5 prediction accuarcy of target variables (node) for the ‘observed’ bayesian network observe nodes target node prediction accuracy seat shape, foot, back shape, leg shape, leg type, fluted, style leg type 81.36±14.36 foot, back shape, leg shape, leg type, leg fluted, style seat shape 62.12±17.10 seat shape, back shape, leg shape, leg type, leg fluted, style foot 53.18±13.22 seat shape, foot, leg shape, leg type, leg fluted, style back shape 44.85±18.78 seat shape, foot, back shape, leg type, leg fluted, style leg shape 90.91± 8.57 seat shape, foot, back shape, leg shape, leg type, style leg fluted 96.36± 4.69 seat shape, foot, back shape, leg shape, leg type, leg fluted style 73.72±14.21 particular design, a novice designer needs to find demonstrated deut the correct foot type. in this case, we select the foot type as the target node and all the other nodes as evidence nodes and update the posterior probability of the target node by giving evidence for all other evidence nodes. from the maximum posterior probability of foot node, we can suggest the most suitable foot type for that particular design. we have tested the prediction accuracy of each node in the ‘observed’ bayesian network by giving evidence to all other nodes. the experimental results are shown in table 3. ten-fold cross validation was used throughout the experiment. 6. conclusions bayesian networks provide a powerful tool and are able to represent and reason with uncertain knowledge. we have demonstrated that it can be used for diagnosis (inference) and classification of design style. we have found that the furniture design style can be classified with 73% of accuracy from the ‘observed’ bayesian network. lorensuhewa, pham, and geva: inferencing design styles... 124 ruhuna journal of science 1, pp. 113–124, (2006) the classification accuracy is slightly lower than the accuracies of svm, c4.5 and nearest neighbour classifiers. the higher classification accuracy has been obtained by integrating the ‘observed’ bn with the ‘expert’ bn by using a simple algorithm based on the past performances of the two bns. bayesian networks can be used not only for classification but also for inference on domain knowledge. this is the advantage of using bayesian network over the other classifiers. we showed that higher accuracy can be achieved for most of the prediction nodes (leg type, leg fluted and leg shape) other than the style node. the furniture design style is selected as an example domain, however the method can be used for other application domains such as medicine. references bouckaert, r.r., 1994. probabilistic network construction using the minimum description length principle, technical report ruu-cs-94-27, department of computer science, utrecht university. cheng, j., d. bell, and w. liu., 1997. learning bayesian networks from data: an efficient approach based on information theory. in sixth acm international conference on information and knowledge management. new york, usa. connectedlines, the on-line furniture style guide. 1998. http://www.connectedlines.com/styleguide/index.htm. cooper, g.f. and e. herskovits, a bayesian method for the induction of probabilistic networks from data. machine learning, 1992. 9: p. 309-347. heckerman, d., 1995. a tutorial on learning with bayesian networks. in twelfth international conference on machine learning. tahoe city, california, usa: morgan kaufman. jensen, f., 1994. implementation aspects of various propergation algorithms in hugin. reserach report r-94-2014, department of mathematics and computer science, aalborg university: denmark. joachims, t., 1999. svm light super vector machine. korb, k.b. and a.e. nicholson, eds. bayesian artificial intelligence. 2003, chapman & hall/ crc press uk. lorensuhewa, a., b. pham, and s. geva., 2003. application of machine learning techniques to design style classification. in the 8th australian and new zealand intelligent information systems conference, macquarie university, sydney, australia. murphy, k., 2002. bayes net toolbox. http://www.ai.mit.edu/ murphyk/software/bnt/bnt.html. pearl, j., 1988. probabilistic reasoning in intelligent systems., san mateo, california: morgan kaufman. quinlan, j.r., 1993. c4.5: programs for machine learning: morgan kaufmann. ruhuna journal of science vol 10(1): 18-31, june 2019 eissn: 2536-8400 faculty of science doi: http://doi.org/10.4038/rjs.v10i1.54 university of ruhuna  faculty of science, university of ruhuna sri lanka 18 preliminary study on diversity of intertidal gastropods in barangay day-asan, surigao city, philippines vince r. abarquez*, noe p. mendez and gloria l. galan department of biology, college of arts and sciences, central mindanao university, university town, musuan 8710 bukidnon, philippines *correspondence: vinceromanoabarquez@gmail.com; https://orcid.org/0000-0002-2814-0780 received: 2 nd august 2018, revised: 11 th january 2019, accepted: 12 th march 2019 abstract unregulated harvesting and habitat degradation of marine gastropods are major anthropogenic activities done by the local people in barangay day-asan, surigao city, philippines. this study was undertaken to determine the diversity of marine gastropods found in the intertidal of dapya island. collection of specimens was done through hand picking and beachcombing methods in the three established study stations. data revealed a total of 184 individuals belonging to 37 species in 5 orders, 15 families and 25 genera. the order neogastropoda obtained the highest number of species with 15 species, followed by order mesogastropoda with 10 species, order archaeogastropoda with 7 species, order neritopsina with 4 species and order caenogastropoda with one species. among the 37 species, nerita undata was the most abundant. shannon-weiner diversity index revealed that station 2 (h=1.193) obtained the highest species diversity, followed by station 3 (h= 1.103) and station 1 (h=1.063). the bray-curtis analysis of species composition showed two different clusters of habitat. cluster 1 is composed of two related habitats joined by stations 1 and 2 with si= 54.26% and cluster 2 composed of stations 3 and 1 with si= 40.65%. furthermore, among the 37 species, six of which were concordant in all study stations viz., angaria delphinus, astralium calcar, conus miles, lambis lambis, euprotomus bulla and canarium labiatum. the results of the study showed a low diversity of intertidal gastropods in barangay day-asan. the area is known to be the most gleaned by locals and it is been already disturbed due to over-harvesting of gastropod species. keywords: angaria delphinus, conus miles, dapya island, nerita undata, surigao del norte. 1 introduction gastropoda is the largest molluscan class, containing more than half of all living mollusc species. gastropods, commonly known as snails and slugs, are distributed among marine, freshwater and terrestrial environments, and occupy diverse habitats, including rivers, trees, deserts, marine intertidal zone vinceromanoabarquez@gmail.com https://orcid.org/0000-0002-2814-0780 abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 19 and the deep sea (pechenik 2011). gastropods are an important source of food in the filipino diet for its delectable taste. gastropod shells are widely collected, traded, bought and sold because of the beauty and attractiveness of each variety. they are also used for the preparation as decorative and household items which find their way to both local and foreign markets (pcaarrd 1988). on the other hand, some species of gastropods act as intermediate host for some medically important trematode parasites of human (miller and harley 2002). one of the major places where people harvest gastropods in the philippines is the dapya island. this island is located in surigao city, the gateway to mindanao and is one of the many cities endowed with extraordinary gifts. according to barrio folks, during older days, various untamed fishes appeared and came along the sea water because of the presence of mangrove trees which grow in swampy places that surrounds barangay day-asan. dapya island has a large tidal flat and it is abundant with marine life including gastropods. since these gastropods are harvested not only for food but also sold for livelihood, there might be possible danger of losing some gastropod species through this action. furthermore, these gastropods have also not been documented and listed, and there is no regulated policy in the area which will limit the people from over-gleaning the gastropod species. although the philippines is considered as a mega-diverse country in terms of molluscs, studies on mollusc diversity in mindanao is considerably scarce (jumawan et al. 2015). additionally, molluscs had the highest number of documented extinctions among the major taxonomic groups in the world (lydeard et al. 2004). thus, this study was undertaken to determine the diversity of intertidal gastropods in barangay day-asan, surigao city. similarity indices and physico-chemical parameters were also tested in the study sites. 2 material and methods 2.1 study site the study was conducted in dapya island, brgy. day-asan, surigao city, philippines (figure 1). day-asan is just 15 km away from the surigao city hall and can be accessed either by motorboat ride or land vehicle. according to edera (2010), day-asan is located at arellano district with a total land area of 3,040.447 ha, of which 554 ha are mangrove forests. there are 1,572 inhabitants and a total of 341 households. abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 20 fig 1. map of dapya island showing the three study stations (inset: philippine map). three study stations were selected in the study site, and the exact position and elevation of the three study stations designated as study station 1, station 2 and station 3 were recorded using the global positioning system (gps) apparatus. a 50 m belt transect line was laid in each station from landward to seaward with the width of 1 m (total area of 50 x 1 m 2 ). a total of 50 quadrats were established per station with an area of 1m 2 , and were numbered from 1 to 50 (q1–q50). three established station plots had a distance of approximately 100 m apart from each other. this method followed that of henderson (2003), which stated that the sampling can be based on the number of quadrates in order to come up with the requirements for determining the species diversity. station 1 faces the open water of surigao with a rocky shoreline (big boulders of rocks), mixed sandy and coralline with some rocks in the substratum. it was abundant in seaweeds and sea grasses, and in marine fauna including gastropods. station 2 was 100 m away from station 1 with almost the same habitat type but closer to the mangrove area. the fauna associates in station 2 like the sea stars were lower compared to station 1. station 3 was established near a mangrove ecosystem which has a rocky shoreline (big abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 21 boulders of rocks). the substratum is mixed rocky and coralline, seaweeds and sea grasses were also abundant but fauna associates were lower compared to the other stations. 2.2 collection of specimens collection of gastropods was done within the established quadrats through hand picking and beach combing from november 2012 – january 2013. the gastropod species were hand-picked or removed using sharp objects like a knife. stone and debris were removed and a hand shovel was used to dig out substrate cover and transferred to a sieve to extract smaller gastropods. the specimens were properly placed in labeled cellophane bags for identification. representative specimens were brought to the biology laboratory room of the college of arts and sciences, central mindanao university, musuan, bukidnon for further processing. 2.3 classification and identification of specimens the specimens were classified and identified using standard taxonomic keys and books of garcia (1986) and laureta (2008) and literature (e.g., tabugo et al. 2013; sosa et al. 2014; dolorosa, dangan-galon 2014; cuadrado 2015; jumawan et al. 2015; ramos et al. 2018). identification was made as far as possible to order, family, genus and species. 2.4 diversity indices the following biodiversity indices were determined using biopro software version 2.0 of mcaleece (2001). the raw data from the line transect were entered into a spread sheet application such as microsoft excel and open orgcalculation. the spread sheet file was loaded to the biodiversity professional application software which processed data and interpreted as plots and tables (mac aleece 2001). the diversity indices were obtained using the formulas of eskandar and kotonegara (1993) as follows: shannon-weiner index across study stations h’= [ ∑ (pi/ ln pi)] where: h= shannon index ‘pi’ = represents the proportion of total sample belong to each species ‘ln pi’ = represents the natural logarithm abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 22 similarity indices of species composition using bray-curtis across study stations by, si= c/(a+b+c) x 100 where: si= species composition c= total species common to all stations a = study station 1 b= study station 2 c= study station 3 2.5 physico-chemical parameters the following environmental conditions were reported following odum (1971): a. temperature: three reading of the water temperature was taken by submerging the laboratory thermometer in the water surface for five minutes. the average temperature was calculated. b. substrate: the substrate of each study station was classified as sandy, rocky, muddy, or coralline. associated flora and fauna were also noted. c. transparency: a secchi disc was submerged in the water until the disc disappeared. reading was recorded and then lifted slowly until the white disc reappeared for the second reading. the averages of the first and second readings were measured for turbidity. this was done three times and the average was then calculated for each station. d. ph: determination of water ph was done using ph paper submerged in the water for three times and referred to a standard color chart and calculated. e. tidal level: tidal determination was based on the tidal calendar record. 2.6 preservation of specimens specimens were secured in the labelled cellophane bags and the foot and visceral mass were carefully extracted from the shell. the shells were then cleaned using soap and detergent and carefully brushed to remove any debris. they were cleaned further using a 10% gleam muriatic acid. after which, the shells were air-dried, packed and labelled for identification. voucher specimens were deposited at the zoology section of central mindanao university museum, musuan, bukidnon, philippines. abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 23 3 results and discussion 3.1 species composition a total of 184 individuals belonging to 37 species were found in the area. these belong to five orders, 15 families and 25 genera (fig. 2). fig 2. representative species collected in the study sites. a) angaria delphinus, b) lambis lambis, c) conus miles, d) euprotomus bulla, e) astralium calcar, f) canarium labiatum. scale bars = 2 cm. abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 24 the order neogastropoda had the highest number of families with six families, followed by orders mesogastropoda and archaeogastropoda with four families each and orders neritopsina and caenogastropoda with one family each. family cypraeidae had the highest number of species recorded with six species and lowest in numbers were the families chilontidae, trochidae, potamidae, angariidae, buccinidae, turbenillidae and volutidae represented with only one species (table 1). table 1. species composition of intertidal gastropods in the three study stations. family/ species number of individuals per station total number of individuals per species s1 s2 s3 i. order archeogastropoda family chilontidae euchelus atratus (gmelin, 1791) 3 0 0 3 family tegulidae tectus pyramis (born, 1778) 2 0 2 4 tectus fenestratus (gmelin, 1791) 7 0 6 13 family trochidae trochus masculatus 0 1 0 1 family turbinidae astralium calcar (linnaeus, 1758) 2 2 6 10 turbo bruneus (roding, 1798) 2 0 3 5 lunella cinerea (born, 1778) 0 1 0 1 ii. order caenogastropoda family potamididae terebralia palustris (linnaeus, 1767) 0 0 11 11 iii. order mesogastropoda family angariidae angaria delphinus (linnaeus, 1758) 10 4 6 20 family cerithidae cerethium nodulosum (bruguiere, 1792) 0 0 1 1 family cyperaeidae cypraea sp. 1 0 3 1 4 cypraea sp. 2 0 1 0 1 cypraea sp. 3 0 1 0 1 erosaria errones (linnaeus, 1758) 0 1 0 1 erronea erosa (linnaeus, 1758) 0 1 0 1 mauritia eglantina (duclos, 1833) 0 0 2 2 family naticidae natica fasciata (roding, 1798) 0 1 0 1 polinices lacteus (guilding, 1834) 0 1 0 1 iv. order neogastropoda family buccinidae cantharus sp. 3 0 0 3 abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 25 table 1 continued. family/ species number of individuals per station total number of individuals per species s1 s2 s3 family conidae conus capitaneus (linnaeus, 1758) 0 1 3 4 conus marmoreus (linnaeus, 1758) 1 0 0 1 conus miles (linnaeus, 1758) 3 7 5 15 conus sp. 1 0 1 0 1 conus sp. 2 0 1 0 1 family muricidae chicoreus torrefactus (sowerby, 1841) 0 0 1 1 chicoreus brunneus (link, 1807) 0 0 1 1 family strombidae canarium maculatum (sowerby, 1842) 0 2 0 2 canarium labiatum (röding, 1798) 3 1 1 5 euprotomusbulla (roding, 1798) 2 2 1 5 gibberulus gibberulus (linnaeus, 1758) 0 2 0 2 lambis lambis (linnaeus, 1758) 2 3 3 8 family turbinellidae vasum turbinellus (linnaeus, 1758) 0 0 2 2 family volutidae cymbiola vespertilio (linnaeus, 1758) 2 1 0 3 v. order neritopsina family neritidae nerita undata (linnaeus, 1758) 16 15 0 31 nerita polita (linnaeus, 1758) 8 5 0 13 nerita plicata (linnaeus, 1758) 2 0 0 2 nerita albicilla (linnaeus, 1758) 0 3 0 3 total number of individuals 16 24 17 184 s1 = station 1; s = station 2; s3 = station 3 the time span of the collection of berdach (1981) was 37 years ago. it could be attributed to many environmental changes like habitat degradation due to anthropogenic activities like the establishing of fish pens, over gleaning and overharvesting, gathering of mangrove trees for firewood and for recreational activities. most of the marine gastropods are economically important. they are harvested by the town folks as food, sold in markets for livelihood and collected for shell ornaments. this probably results in low number of species found in the area. the major species harvested in the area were from the genus lambis (pers. comm.). as observed, this species was few in number compared to other species collected. habitat degradation, unregulated harvesting, cutting of mangrove forest and improper waste disposal in the study sites are possible threats that cause gastropod depletion in the area similar to the report of dolorosa and dangan abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 26 galon (2014) in iwahig river-estuary of palawan. as reported by earlier studies, gleaning or collection of edible molluscs is a common practice in the philippines since these organisms are used either for personal consumption or as additional earnings (nieves et al. 2010, napata and andalecio 2011, masangcay and lacuna 2017). it is also important to conserve the gastropod species in the area since some species of family strombidae are the usual species collected by the local people. 3.2 species diversity shannon-weiner diversity index the diversity of gastropod species was highest in station 2 (h’ = 1.193), followed by station 3 (h’ = 1.103) and lowest in station 1 (h’= 1.063) (table 2). table 2. results of shannon-weiner diversity index. index s1 s2 s3 shannon h' log base 10. 1.063 1.193 1.103 shannon hmax log base 10. 1.204 1.38 1.23 shannon j' 0.883 0.864 0.896 legend: s1 = station 1; s = station 2; s3 = station 3 fig 3. k-plot of species abundance of marine gastropods in barangay day-asan, surigao city. abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 27 the diversity of gastropods in barangay day-asan is low which also support the philippine studies on gastropods by berdach (1981) in puerto galera, and flores and zafaralla (2012) in mananga river in cebu. since the values obtained for shannon-weiner diversity index are less than 2.5, this study implies that the three sites have low diversity index as none of them obtained a value higher than 2.5 (flores and zafaralla 2012). in the species accumulation curve, the study shows that the data is sufficient for the diversity index analysis (figure 3). however, as mentioned by cuadrado (2015), this low diversity index value of the three sites would make sense if the sampling was done in a longer period of time. similarity index the results showed clustering of marine gastropod species in stations 1 and 2 maybe because these two stations were on the same side of the island. species in station 3 in the dendrogram (figure 4) appeared as a different cluster and branched to other stations because this station was established in other side of the island specifically nearer to the mangrove area. the species in station 3 were seemed to be unique, which supported by the dendrogram. fig 4. dendrogram of similarity species composition of marine gastropods in barangay day-asan, surigao city. 3.3 species distribution among the species collected, four species were observed only in station 1. these species were nerita plicata, euchelus atratus, conus marmoreus and abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 28 cantharus sp. eleven species were found only in station 2 viz., nerita albicilla, lunella cinerea, trochus masculatus, conus sp. 1, conus sp. 2, cypraea sp. 2, cypraea sp. 3, cypraea errones, cypraea errosa, natica fasciata and polinices lacteus and six species were only found in station 3 viz., terebralia sp., vasum turbinellus, chicoreus torrefactus, chicoreus brunneus, cypraea eglantina and cerithium nodulosum. species found in stations 1 and 2 had similarity in the species composition, while the species in station 3 differed from the other stations. there were six species which had the widest range of distribution concordant in all study stations viz., angaria delphinus, astralium calcar, conus miles, lambis lambis, euprotomus bulla and canarium labiatum. the diversity of marine gastropods in barangay day-asan showed similarity with two clusters in the dendrogram. cluster 1 showed the two closely related habitats namely: station 1 and station 2 with si = 54.26% of species composition, while cluster 2 consists of stations 1 and 3 with si = 40.65%. the most abundant species in station 1 was n. undata. nerites prefer brackish water systems and supported the recorded water ph which was 9. moreover, they were found on rocky substrates where seaweed was abundant. the tide is an important environmental factor influencing the mobility of the organisms and their exposure to air and water. the tidal levels also affect the species distribution. also, station 1 favored the presence of many sea stars of which gastropods are one of its preys. for the distribution of n. undata, since it is attached in rocks on the shoreline, it is not been preyed by the sea stars which also reflects to the abundance of the species in the station. for the species frequency, n. undata was only seen in 1 out of 50 quadrates, since they were distributed and can be seen in clumps or living together because they are said to be gregarious animals (tan and clements 2008). angaria delphinus on the other hand is the frequently seen species since it was found in 7 out of 50 quadrats found scattered along the study station areas. in station 2, still n. undata was the most abundant species. same observation was done with study station 1 which was established on the same part of the island. the presence of fauna associates like sea stars was lesser than compared to station 1. this reflects to number of species found, of which this station is diverse in gastropods. same environmental factors affected the species in station 1. for its distribution, they were also seen in 1 quadrate only since they were in clumped distribution. conus miles was the most frequently seen species in station 2 which was recorded from 7 out of 50 quadrates. in study station 3, terebralia sp. was the most abundant species, since this station was established in the other side of the island, near to a mangrove and exposed to different kinds of environmental factors. this species was present in mangrove area, and since they are mobile, they can be found on the shores of station 3. furthermore, this species is not consumed by local people and hence it is the least disturbed among the species in station 3. still, a. delphinus was the most frequently seen species, widely distributed in the abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 29 station which was found in 6 out of 50 quadrates. flora and fauna associates in the station were less which reflected the disturbance in the area. 3.4 physico-chemical parameters the mean water temperature in the three stations during the study was 27°c. the mean water ph was 9 in the three stations, mean of water salinity was 34 ppt and substrate cover were all mixed with sandy, coralline, rocky and some with boulders of rocks. flora associates were mainly sea grasses and seaweeds, while the fauna associates were sea stars, sea urchins, sea anemone, hermit crabs and synapta sp. sampling in all study stations was done during low tide (table 3). table 3. summary of physico-chemical parameters in the three study stations. parameters s1 s2 s3 temperature 26°c 28°c 27°c substrate cover (nature of substrate) mixed sandy and rocky mixed sandy and rocky rocky (small to moderate) substrate with less sandy substrate transparency clear clear clear salinity 33 34 34 ph 8 9 9 tidal measurement low tide low tide low tide legend: s1 = station 1; s = station 2; s3 = station 3 strzelec and krolczyk (2004) stated that many gastropod species are tolerant to most physico-chemical parameters and their occurrence is affected by the quality of bottom sediments and abundance of vegetation. in this study, the physico-chemical factors in the three sites are closely the same. the substrate cover at stations 1 and 2 were similar with mixed sandy and rocky areas. these types of substrate possibly support many gastropods and hence the higher abundance at these two stations. in contrast, gastropods that thrive on rocky substrata prefer station 3, which mainly consists of rocky (small to moderate) substrate. moreover, the water salinity and ph in station 3 may be influenced by the mangrove environment. the physico-chemical parameters like the water temperature, water ph, water salinity and water substratum in the study stations did not very much. thus, these parameters may fall within the favorable environmental conditions suitable for the growth of gastropods. abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 30 5 conclusions and recommendations this study revealed that among the 37 species, nerita undata was the most abundant with a total of 31 individuals. shannon-weiner index revealed that station 2 (h= 1.193) obtained the highest species diversity, followed by station 3 (h = 1.103) and station 1 (h= 1.063). the bray-curtis analysis of species composition showed two different clusters of habitat. cluster 1 is composed of two related habitats joined by stations 1 and 2 with si= 54.26% and cluster 2 composed of stations 3 and 1 with si= 40.65%. furthermore, among the 37 species, six were concordant in all study stations viz., a. delphinus, a. calcar, c. miles, l.lambis, e. bulla and c. labiatum. among the species, n. plicata, e. atratus, c. marmoreus and cantharus sp.1 were the unique species in station 1. mixtures of sandy and rocky substrata are the preferred areas of gastropods to inhabit. the environment in station 3 was different from the other two stations due to its proximity to the mangrove. it is recommended that a yearly survey should be done for a more comprehensive record on gastropods present in the area. these data will be useful for conservation and culture of highly economically important species (e.g., some species from family strombidae). acknowledgments the authors are grateful to prof. marilou m. ediza and prof. victoria t. quimpang for their suggestions and comments in this paper. due acknowledgments are given to marife s. servasques for helping in identification and processing of specimens; to mr. and mrs. edgardo abarquez for financially supporting this study; to the anonymous reviewers for their constructive comments on this paper. references berdach jt. 1981. inventory of marine gastropods in the man and the biosphere (mab) reserve area, puerto galera, oriental mindoro, philippines. philipp. j. biol. 10(1): 95–108. cuadrado jt. 2015. preliminary assessment of freshwater gastropods in the selected rivers in esperanza, agusan del sur, philippines. the journal of zoology studies 2(4): 13–20. issn 2348-5914. dolorosa rg, dangan-galon f. 2014. species richness of bivalves and gastropods in iwahig river-estuary, palawan, the philippines. international journal of fisheries and aquatic studies 2(1): 207–215. issn: 2347-5129. edera dc. 2010. barangay socio-economic profile, surigao city planning and development office. eskandar j, kontonegara r. 1993. regional study of diversity: concepts and frameworks and method proceedings of seasian university agro ecosystem network (suan and program on environment lenv).west central pei shengii and precy sigisie yunnam. university press. flores mjl, zafaralla mt. 2012. macroinvertebrates composition, diversity and richness in relation to the water quality status of mananga river, cebu, philippines. philippine science letters 5(2): 103–113. garcia rf, natividad vs. 1986. guide to philippine flora and fauna. vol. vi. gastropods, polecypods, annelids. natural resources management center, ministry of natural abarquez et al. intertidal gastropods in dapya island, philippines ruhuna journal of science vol 10(1): 18-31, june 2019 31 resources and university of the philippines. jmc press, inc. quezon city, philippines. 356 pp. henderson pa. 2003. practical methods in ecology. 1117 pp. blackwell publishing. usa. hickman jr, roberts l, keen s, larson a, eisenhour d. 2003. animal diversity. 6th edition. mc graw hill company. usa. 171–182 pp. jumawan jh, tripoli ffd, boquia ees, niez klm, veronilla jah, dellomes sa, udtie rm, seit nk, hasim na, gatinao mjo. 2015. species diversity and spatial structure of intertidal mollusks in padada, davao del sur, philippines. aacl bioflux 8(3): 301–309. laureta lv. 2008. compendium of the economically important seashells in panay, philippines. university of the philippines press, quezon city, philippines. lydeard c, cowie rh, ponder wf, bogan ae, bouchet p, clark sa, cummings ks, frest tj, gargominy o, herbert dg, hershler r, perez ke, roth b, seddon m, strong ee, thompson fg. 2004. the global decline of nonmarine mollusks. bioscience 54(4): 321–330. masangcay sig, lacuna mldg. 2017. molluscan (gastropoda and bivalvia) diversity and abundance in rocky intertidal areas of lugait, misamis oriental, northern mindanao, philippines. journal of biodiversity and environmental sciences 11(3): 169–179. accessed at: http://www.innspub.net. mc aleece. 1997. biodiversity professional version 2.0 devised by pgd lambshade, glj paterson and d.j gage. programme software. copyright by the natural history museum and scottish association for marine saero. miller sa, harley jp. 2002. zoology. 5th edition. 170–179 pp. mc graw hill company. usa. napata rp, andalecio mn. 2011. exploitation and management of brown mussel. natural history museum journal 6(2): 113–128. nieves pm, de jesus sc, macale amb, pelea jmd. 2010. an assessment of macroinvertebrate gleaning in fisheries on the albay side of lagonoy gulf. kuroshio science 4(1):27–35. odum e. 1971. fundamentals of ecology. w. b. saunders co. london. pechenik ja. 2011. biology of the invertebrates. 6th edition. 215–229 pp. mc graw hill company. usa. pcaarrd. 1988. oyster and mussel research. department of science and technology. los banos, laguna. ramos dae, batomalaque ga, anticamara ja. 2018. current status of philippine mollusk museum collections and research, and their implications on biodiversity science and conservation. philippine journal of science 147(1): 123–163. issn 0031-7683 strzelec m, królczyk a. 2004. factors affecting snail (gastropoda) community structure in the upper course of the warta river (poland). biologia, bratislava 59(2):159–163. tabugo srm, pattuinan jo, sespene njj, jamasali aj. 2013. some economically important bivalves and gastropods found in the island of hadji panglima tahil, in the province of sulu, philippines. international research journal of biological sciences 2(7): 30–36. tan sk, clements r. 2008. taxonomy and distribution of neritidae (mollusca: gastropoda) in singapore. zoological studies 47: 481–494. rjs-vol-1-sept-2006-8.dvi ruhuna journal of science vol. 1, september 2006, pp. 67–81 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. validation of taguchi design using a quality parameter of cosmetic soap manufactured in a local industry c. p. s. pathirana department of mathematics, university of ruhuna, matara, sri lanka, macpsp@maths.ruh.ac.lk r. a. dayananda, p. dias department of statistics and computer science, university of sri jayewardenepura, gangodawila, nugegoda, sri lanka, radaya@sjp.ac.lk, dias@sjp.ac.lk s. b. nawarathne harishchandra mills (pvt.) ltd., matara, sri lanka the objective of this study is to investigate the applicability of taguchi method to the production process of a local industry. harishchandra mills (pvt.) ltd, matara, which is engaged in soap manufacturing process has been selected as the local industry. this process is entangled with different kinds of variables. four critical variables two types of oil, coconut oil (a1) and palm oil (a0), on the basis of high and low saponification values, two levels of caustic soda (b0, b1), and two different saponification temperatures, 70�(c0) and 80�(c1), with and without stirring process for saponification (d1, d0) were adopted. the results obtained for total fatty matter (tfm) was used to calculate optimum value (yopt) for each treatment with and without replicates. the results for yopt for tfm with and without replicates were 75.98%, and 76.66%, respectively. the best treatment combination under this method was a1b1c1d1 for the same sequence of quality parameters given above. the same best treatment combinations were also obtained for the validation test, fourfactor experiment design and block design with and without replicates. key words : soap manufacture, statistical quality control, taguchi method, factorial experiment design, and block design. 1. introduction sri lanka is a developing country, which enjoys a wide spectrum of production processes using abundant natural resources. with the emergence of global village concept and free market economy a substantial amount of business establishments have mushroomed throughout the country. though there are many kinds of business establishments’ function in the country, most of these do not posses an adequate advanced technical know how to cope with the issues encountered by them during their day today activities. aftermath of this development has resulted in releasing different kinds of substandard products, produced under manufacturer’s own standard to the dynamic market. hence in order to cope with this situation, the technical term “standard67 pathirana et al.,: validation of taguchi design ... 68 ruhuna journal of science 1, pp. 67–81, (2006) ization” has come into light along with regulatory requirement with the view to protect the consumer. since modern market is also dynamic and competitive the entrepreneurs strive hard to maintain competitive edge on their products while minimizing wastages occurring in the production process and boosting the productivity. on the other hand, sri lankan business establishments can be categorized into 3 key groups as traditional, semi modern and modern. the majority of them are coming under traditional and semi modern groups. however they do not posses adequate resources to face the challenges exerted by the conditions in the dynamic marketing environment. therefore switching on to advanced statistical designs is one of the options for any type of producer to improve the over all productivity of his business along with the quality while minimizing wastages. professor genichi taguchi, an engineer and an eminent scientist from japan who encountered with a similar problem was able to introduce a new statistical design, after deeply studying the “off line quality control” activities of production lines. the new statistical design called “taguchi design” is applicable to the modern industrial environment, in order to resolve most issues emerging from wilful errors and inadvertent errors in a production line. the other advantage of using this statistical design is the maximum utilization of tangible and intangible scarce natural resources in a production process. hence objective of the study was to use taguchi design in a production process of a local industry, where production process is carried out amid considerable number of variables. outcome of which is to be validated by two other designs to gauge the accuracy of taguchi design. 1.1. objectives� identification of error variables of cosmetic soap manufacturing process using control charts.� comparative study of factorial experiment, confounded blocks design and taguchi design.� manufacturing of cosmetic soap with respect to the 24 designs.� manufacturing of high quality soap for consumer market by identifying desirable limits for important variables of soap manufacturing process with respect to the advance taguchi design. 2. materials and methods 2.1. terminology we shall use the following terminology throughout the paper. total fatty matter (tfm), low saponification oil (a0), high saponification oil (a1), naoh low concentration (b0), naoh high concentration (b1), low temperature (c0), high temperature (c1), no stirring (d0), with stirring (d1), oil (a), naoh (b), temperature (c), stirring (d), number of replicate (n) and number of variable (f). pathirana et al.,: validation of taguchi design ... ruhuna journal of science 1, pp. 67–81, (2006) 69 table 1 algebraic signs for the taguchi design � � � � � � � � � variable test treatment a b c d combination (1) a0b0c0d0 ad a 1 b 0 c 0 d 1 + + bd a 0 b 1 c 0 d 1 + + ab a 1 b 1 c 0 d 0 + + cd a 0 b 0 c 1 d 1 + + ac a 1 b 0 c 1 d 0 + + bc a 0 b 1 c 1 d 0 + + abcd a 1 b 1 c 1 d 1 + + + + 2.2. parameters four key variables responsible for major quality characteristics of soap were identified after scrutinizing the soap manufacturing process. they are,� saponification value of oils,� concentration of sodium hydroxide (naoh),� saponification temperature, and� period of stirring. quality parameter of the samples,� total fatty matter (tfm) 2.3. taguchi method taguchi constructed a special set of general designs for factorial experiments that cover many applications. the special set of designs consists of orthogonal arrays (oa). the use of these arrays helps to determine the least number of experiments needed for a given set factors. (dayananda 1992) the latest statistical model for design of an experiment, “taguchi’s advance design”, was adapted for this study, which has the advantage of less treatment combinations in compared with other statistical designs. (see table 1) above design was used to analyze the data obtained for tfm of soap samples in order to determine optimum values for critical variables. 2.3.1. statistical analysis of data: the data obtained for tfm of soap samples of different treatment combinations were statistically analyzed in order to determine yopt using following formula. yopt = t n + 4∑ i=1 (k̄i − t n ) . (1) where, t = (ā0 + ā1) + (b̄0 + b̄1) + (c̄0 + c̄1) + (d̄0 + d̄1), k̄i = mean value of optimum level of each variable, and n = number of observations. (nawarathne 2003) pathirana et al.,: validation of taguchi design ... 70 ruhuna journal of science 1, pp. 67–81, (2006) table 2 algebraic signs for the 24 experiment design x x x x x x x x x fact effect run i a b a b c a c b c a b c d a d b d a b d c d a c d b c d a b c d (i) a0b0c0d0 + + + + + + + + a a 1 b 0 c 0 d 0 + + + + + + + + b a 0 b 1 c 0 d 0 + + + + + + + + ab a 1 b 1 c 0 d 0 + + + + + + + + c a 0 b 0 c 1 d 0 + + + + + + + + ac a 1 b 0 c 1 d 0 + + + + + + + + bc a 0 b 1 c 1 d 0 + + + + + + + + abc a 1 b 1 c 1 d 0 + + + + + + + + d a 0 b 0 c 0 d 1 + + + + + + + + ad a 1 b 0 c 0 d 1 + + + + + + + + bd a 0 b 1 c 0 d 1 + + + + + + + + abd a 1 b 1 c 0 d 1 + + + + + + + + cd a 0 b 0 c 1 d 1 + + + + + + + + acd a 1 b 0 c 1 d 1 + + + + + + + + bcd a 0 b 1 c 1 d 1 + + + + + + + + abcd a 1 b 1 c 1 d 1 + + + + + + + + + + + + + + + + divisor 16 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 2.4. validation test four-factor experiment design the treatment combination in this design is given in table 2 (george, william, hunter 1978). 2.4.1. statistical analysis of data of 24 design: the data obtained from the four-factor experiment design were analyzed using table 2, and the contrast and the effect with respect to each variable were calculated. the effect for each variable was calculated using the following formula (thomas 1998). effect = contrast 2f−1n . (2) the sum of square for any effect of each variables were calculated using, ss = (contrast)2 2f n . (3) the high percent contribution values were selected from anova table of which significant levels of each variable were calculated. a basic model was developed using significantly different values from the anova table. the function for the basic model is, y =β0 + β1x1 + β2x2 + β3x3 + β4x4 + β12x1x2 + β13x1x3 + β23x2x3 + β14x1x4 +β24x2x4 + β34x3x4 + β123x1x2x3 + β134x1x3x4 + β234x2x3x4 +β124x1x2x4 + β234x2x3x4 + β124x1x2x4 + β1234x1x2x3x4 + ǫ (4) pathirana et al.,: validation of taguchi design ... ruhuna journal of science 1, pp. 67–81, (2006) 71 figure 1 assignment of the 16 runs to two blocks -without replicate (abcd confounded). figure 2 partial confounding in the 24 design with replicates. the developed model for quality parameter tfm can be used to detect relationship between significantly different variables. this relationship can be depicted by either response surface or contour plot. (douglas 2000) 2.5. validation test confounded block design same data obtained for four-factor experiment designs were used for block design in order to validate taguchi design again. (douglas 2000) (see figure 1, 2) 2.5.1. statistical analysis of confounded block design. the effect and sum of square for abcd (block) was calculate using, block effect = ȳblock1 − ȳblock1 . (5) ssblocks = (y1) 2 8 + (y2) 2 8 − (y1 + y2) 2 16 . (6) where, ȳblocki = average value of each block i and yi = sum of block i. pathirana et al.,: validation of taguchi design ... 72 ruhuna journal of science 1, pp. 67–81, (2006) table 3 comparison of factorial design and taguchi design factors level total number of experiments taguchi design (factorial design) total number of experiments 2 2 4 4 3 2 8 4 4 2 16 8 7 2 128 8 15 2 32,768 16 4 3 81 9 however, the values of ssabc and ssabcd should be calculated using only the data in replicate i and replicate ii, respectively. the effect and sum of square for the replicate were calculated using, ssrep = n∑ i=1 r2i 2k − y2... 2kn (7) where ri is the total observation in the i th replicate. n is the number of replicates, and k is the number of variable. y... is the sum of total observations.(richard 1996) 2.6. comparison of factorial design and taguchi design the total number of experiments possible for different number of factors at 2 or 3 levels and the corresponding suggested taguchi number of experiments is shown in table 3. (www.stasoft.com 2004, pathirana 2005) taguchi has established oas that describes a large number of experimental situations. 3. results and discussion results pertaining to the tfm of wet soap manufactured with respect to taguchi design were validated by two statistical designs, fourfactor experiment design and block design. 3.1. tfm value for samples prepared from each design tfm value was obtained for soap samples prepared with respect to each design, taguchi design, four-factor experiment design and block design. 3.1.1. tfm value for taguchi design without and with replicate: the data obtained for tfm value of soap with respect to taguchi design without and with replicates (three replicates) are given in table 4 and table 5 respectively. tfm values obtained for taguchi design with respect to each variable are given in linear graphs in figure 3 and figure 4. these graphs clearly indicate that there is a positive relationship for tfm value with the increase of magnitude of each variable. hence, the best treatment combination in order to enhance tfm value of soap manufacture is a1b1c1d1. furthermore, the value of tfm is increased simultaneously with the increase of each variable. calculation from the model (using equation 1) for optimum value of tfm (yopt) is equal to 75.98% and 76.66% with and without replicates, respectively. (see figure 3, 4) pathirana et al.,: validation of taguchi design ... ruhuna journal of science 1, pp. 67–81, (2006) 73 table 4 tfm value for taguchi design without replicate variable a b c d treatment value test combination (1) a0b0c0d0 63.00 ad + + a1b0c0d1 68.25 bd + + a0b1c0d1 60.42 ab + + a1b1c0d0 66.30 cd + + a0b0c1d1 73.90 ac + + a1b0c1d0 65.32 bc + + a0b1c1d0 72.10 abcd + + + + a1b1c1d1 79.20 table 5 tfm value for taguchi design with replicates � � � � � � � � � variable test treatment replicate replicate replicate total combination i ii iii value (1) a0b0c0d0 62.50 63.00 62.50 188.00 ad a 1 b 0 c 0 d 1 68.20 67.30 68.00 204.00 bd a 0 b 1 c 0 d 1 60.52 60.88 60.86 182.26 ab a 1 b 1 c 0 d 0 66.50 66.20 66.30 199.00 cd a 0 b 0 c 1 d 1 72.70 73.30 74.00 220.00 ac a 1 b 0 c 1 d 0 64.30 64.50 64.20 193.00 bc a 0 b 1 c 1 d 0 70.90 71.40 71.70 214.00 abcd a 1 b 1 c 1 d 1 77.80 78.90 78.30 235.00 69.25 69.19 a v e r a e a v e r a e a a0 1 b1 67.02 g b g 0 67.08 (a) (b) 71.83 64.44 66.17 a v e r a g e a v e r a g e 70.10 (c) (d) d d0c c 01 1 figure 3 relationship of tfm with respect to each variablewith replicates pathirana et al.,: validation of taguchi design ... 74 ruhuna journal of science 1, pp. 67–81, (2006) a v e r a g e a v e r a g e (a) a a b0 01 b1 69.77 67.36 67.62 69.50 (b) a v e r a g e a v e r a g e (c) (d) d d0c c 01 1 64.49 72.63 66.68 70.44 figure 4 relationship of tfm value with respect to each variable without replicates 3.1.2. four factor experiment design with and without replicate for tfm. the data obtained for tfm value for without replicate design are given in table 6. a normal probability graph of cumulative value of probability vs effect is drawn in figure 5. the data obtained for tfm value for with replicates design are given in table 7 soap samples were prepared with respect to the 24 design and each treatment combination was replicated 3 times. a normal probability graph of cumulative value of probability vs effect is drawn in figure 6. these graphs (figure 5 and figure 6) clearly indicate that the variables c and d are more significant in compared with a and b. an anova table was developed for further studies of the significance of variables c and d and the results are given in table 8 and table 9. the anova table clearly indicates that calculated values for variables c (temperature) and d (stirring) are higher than the f distribution table values (f1,5(0.10) = 4.06 and f1,32(0.01) = 7.56 ). from table.8, c = 4.16 > 4.06 d = 5.26 > 4.06 and from table.9, c = 24.68 > 7.56 d = 70.25 > 7.56. pathirana et al.,: validation of taguchi design ... ruhuna journal of science 1, pp. 67–81, (2006) 75 table 6 tfm value for 24 factorial design (without replicate) run run treatment values number label combination 1 (i) a0b0c0d0 66.20 2 a a1b0c0d0 66.00 3 b a0b1c0d0 57.00 4 ab a1b1c0d0 58.00 5 c a0b0c1d0 55.67 6 ac a1b0c1d0 65.32 7 bc a0b1c1d0 71.90 8 abc a1b1c1d0 57.69 9 d a0b0c0d1 69.31 10 ad a1b0c0d1 68.25 11 bd a0b1c0d1 57.42 12 abd a1b1c0d1 56.90 13 cd a0b0c1d1 73.50 14 acd a1b0c1d1 48.20 15 bcd a0b1c1d1 69.39 16 abcd a1b1c1d1 78.20 figure 5 normal probability plots for the effects for tfm (without replicate) therefore, relatively high temperature and stirring time have a big effect on tfm of soap manufactured. since four-factor experiment design also shows that most important variables for tfm of soap manufacture are c and d and least effect variables are a and b, taguchi method can be validated by four-factor factorial design as taguchi method was also indicated that c and d variables at higher levels are critical for soap manufacturing process, in addition to a and b variables. pathirana et al.,: validation of taguchi design ... 76 ruhuna journal of science 1, pp. 67–81, (2006) table 7 tfm value for 24 experiment designwith replicates run treatment 1 replicate replicate replicate total label combination i ii iii (1) a0b0c0d0 69.63 66.50 65.33 201.46 a a 1 b 0 c 0 d 0 67.16 65.55 65.23 197.94 b a 0 b 1 c 0 d 0 58.69 56.72 55.63 171.04 ab a 1 b 1 c 0 d 0 56.72 56.80 57.40 170.92 c a 0 b 0 c 1 d 0 45.56 45.60 45.82 136.98 ac a 1 b 0 c 1 d 0 64.82 64.50 64.43 193.75 bc a 0 b 1 c 1 d 0 70.61 71.50 72.22 214.33 abc a 1 b 1 c 1 d 0 62.33 65.43 66.55 194.31 d a 0 b 0 c 0 d 1 70.20 69.22 68.50 207.92 ad a 1 b 0 c 0 d 1 68.11 67.45 68.22 203.78 bd a 0 b 1 c 0 d 1 58.20 58.40 57.70 174.30 abd a 1 b 1 c 0 d 1 54.42 57.27 58.98 170.67 cd a 0 b 0 c 1 d 1 72.98 73.30 74.40 220.68 acd a 1 b 0 c 1 d 1 45.39 50.34 48.87 144.60 bcd a 0 b 1 c 1 d 1 70.42 69.50 68.24 208.16 abcd a 1 b 1 c 1 d 1 78.30 78.42 78.85 235.57 figure 6 normal probability plots for the effect for tfm with replicates similarly out come of the anova tables with and without replicate do not show any significant difference. therefore, four-factor experiment design can also be used to validate taguchi method. a regression model can be developed using calculated f values, which should be more than the f table value. with replicates, the model is as follows. y = 63.47 + 0.46x3 + 1.77x4 + 0.64x1x2 − 1.86x1x4 + 5.84x2x3 + 1.13x3x4 + 3.70x1x2x4 + 1.69x1x3x4 + 3.82x1x2x3x4. (8) pathirana et al.,: validation of taguchi design ... ruhuna journal of science 1, pp. 67–81, (2006) 77 table 8 tfm (without replicate)-analysis of the variance source of sum of degree of mean f0 variation square freedom squares value c 27.01 1 27.01 4.16 d 34.19 1 34.19 5.26 ab 15.98 1 15.98 2.46 bc 350.91 1 350.91 53.99 cd 12.31 1 12.31 1.89 ac 25.68 1 25.68 3.95 ad 12.80 1 12.80 1.97 abd 215.89 1 219.85 33.82 bcd 34.02 1 34.02 5.23 abcd 214.84 1 214.84 33.05 error 32.52 5 6.50 total 976.15 15 table 9 tfm (with replicates)-analysis of the variance source of sum of degree of mean f0 variation square freedom squares value a 11.63 1 11.63 5.43 b 10.26 1 10.26 4.79 c 52.82 1 52.82 24.68 d 150.34 1 150.34 70.25 ab 19.52 1 19.52 9.12 ac < 0.01 1 < 0.01 < 0.01 ad 167.07 1 167.07 78.07 bc 1639.50 1 1639.50 766.12 bd 1.60 1 1.60 0.75 cd 61.49 1 61.49 28.73 abc 10.82 1 10.82 5.06 abd 655.57 1 655.57 306.34 acd 137.66 1 137.66 64.33 bcd 2.01 1 2.01 0.94 abcd 698.98 1 698.98 326.63 error 68.45 32 2.14 total 2390.96 47 significant f values are highlighted. 3.1.3. confounded block design with and without replicate for tfm. the data obtained for tfm value of soap base, prepared in accordance with the block design are given in figure 7. (abcd confounded without replicate). thus, each replicate of 24 designs must be run in two blocks. two replicates are run, with abcd confounded in replicate i and abc confounded in replicate ii. the data are given in figure 8. an anova table was made with respect to the variables and interactions that are given in table.10 and table.11. pathirana et al.,: validation of taguchi design ... 78 ruhuna journal of science 1, pp. 67–81, (2006) figure 7 tfm value (without replicate) 16 runs in two blocks figure 8 tfm partial confounding in the 24 design table 10 tfm (without replicate) analysis of variance for block design source of sum of degree of mean f − variation squares freedom squares value block(abcd) 642.25 1 642.25 98.81 c 27.01 1 27.01 4.16 d 34.19 1 34.19 5.26 ab 15.98 1 15.98 2.46 ad 12.80 1 12.80 1.97 bc 350.91 1 350.91 53.99 cd 12.30 1 12.30 1.89 ac 25.68 1 25.68 3.95 abd 215.89 1 213.89 32.91 bcd 34.02 1 34.02 5.23 error 32.52 5 6.50 total 1403.55 15 significant f values are highlighted. since the anova table.10 and table 11 revealed that calculated f values for c and d variables are higher than the table values, f1,5 (0.01) = (4.06) and f1,13(0.01) = 9.07, these variables are more important than others. pathirana et al.,: validation of taguchi design ... ruhuna journal of science 1, pp. 67–81, (2006) 79 table 11 anova for block design tfm (with replicates) source of sum of degree of mean f0 variation squares freedom squares value replicates 0.27 1 0.27 0.13 block within replicates 258.41 2 129.20 61.82 a 18.03 1 18.03 8.63 b 9.48 1 9.48 4.54 c 24.43 1 24.43 11.68 d 90.45 1 90.45 43.28 ab 7.33 1 7.33 3.51 ac 0.54 1 0.54 0.26 ad 116.36 1 116.36 55.67 bc 1135.50 1 1135.50 543.30 bd 0.07 1 0.07 0.03 cd 57.24 1 57.24 27.39 abc (replicate i only) 5.20 1 5.20 2.49 abd 439.83 1 429.83 205.67 acd 91.46 1 91.46 43.76 bcd 4.52 1 4.52 2.16 abc (replicate ii only) 204.78 1 204.78 97.98 error 27.18 13 2.09 total 2491.08 31 significant f values are highlighted. table 10: c = 4.16 > 4.06 d = 5.26 > 4.06. table 11: c = 11.68 > 9.07, d = 43.28 > 9.07. as block design indicates that c and d variables at higher levels are significantly contributed for the increase of tfm value, this can be used to validate taguchi design, because taguchi design is also showing similar results. same results were obtained for variables with and without replicates. 4. conclusion based on the results obtained under taguchi method, we can conclude the following: 1. the optimal combination of four variables for highest tfm is high saponification value of oil (coconut oil), high concentration of naoh (18%), high saponification temperatures (80�) and with stirring. this was the case for both with and without replicates. the highest tfm obtained under this combination with and without replicates were 75.98% and 76.66% respectively. this is almost similar to the sri lanka standards value of 76.5% (ceylon standard 27:1968). pathirana et al.,: validation of taguchi design ... 80 ruhuna journal of science 1, pp. 67–81, (2006) 2. when soap is manufactured with respect to advanced taguchi design after the identification of critical variables, the quality parameters are easily falling in line with regulatory requirements (tfm). 3. since the results for all quality parameters of soap have shown only a slightest difference with and without replicates, the experiment could be conducted even without replicates and achieve very precise results. therefore, we can recommend that the advanced taguchi design could be applicable for some industrial environments at low cost with other high quality comparative designs as they need more treatment combinations to evaluate the effect of each variable or their combinations. 4. taguchi method is an advance statistical design. it is most sensitive in terms of finding the effect of variation of variables. taguchi method was able to find a little improvement when the variables a and b are increased from low level to high level (a0 to a1 and b0 to b1). however, the other two designs do not strong enough to find the variation of these two variables. references box g, bisgard s, fung c 1987 : an explanation and critique of taguchi’s contributions to quality engineering, supplements to the journal of the university of wisconsion, u.s.a, 126-128. dayananda r.a 1992 :fifty year of quality technology, university of sri jayewardenepura. douglas c.m 2000. design and analysis of experiment, fifth edition, arizona. george e.p.box, william g.hunter, hunter j.s 1978 : statistics for experiments, john wiley & sons. http://kernow.curtin.edu.av/www/taguchi/cae204.html, 20.03.2004. methods of analysis of soap, ceylon standard 27:1968, bureau of ceylon standards. nawarathne sb 2003. m.phil thesis. extraction of adible grade sesame oil by controlling acid value and free fatty acid level in seeds and oil. pathirana c.p.s 2005. m.sc thesis validation of taguchi design using quality parameters of toilet soap manufactured in a local industry, university of sri jayewardenepura. richard r. johnson 1996: miller & freund’s probability and statistics for engineers. fifth edition, prentice hall of india private limited, new delhi. ronald e.w, and raymond h.m.: probability and statistics for engineers and scientists, collier macmillan publisher, london. thomas p.r 1998: statistical methods for quality improvement. wiley series acknowledgments c. p. s. pathirana would like to thank dr. l. a. l w. jayasekara, senior lecturer, department of mathematics at university of ruhuna for introducing me to the subject and providing the guidance for the success of this project. pathirana et al.,: validation of taguchi design ... ruhuna journal of science 1, pp. 67–81, (2006) 81 c. p. s. pathirana, also gratefully recalls the sincerity of mr. senaka samarasesinghe, managing director of harishchandra mills (pvt.) ltd., for kindly permitting to carry out the research work in his establishment. special words of gratitude go to mrs. sujeewa rathnayake, miss devika preshanthi, miss dammika wijesekera and the other staff members of the research lab of harishchandra mills (pvt.) ltd., for their kind cooperation. the authors, also, grateful to msc course coordinator, dr. s. k. boralugoda, for his guidance, advice and the encouragement to proceed with this research project. finally, we thank the department of mathematics, university of ruhuna matara, for the facilities provided for the success of this project. ruhuna journal of science vol 10 (2): 96-107, december 2019 eissn: 2536-8400 faculty of science doi: http://doi.org/10.4038/rjs.v10i2.76 university of ruhuna  faculty of science, university of ruhuna sri lanka 96 characterization of sea lettuce (ulva lactuca) from matara, sri lanka and development of nutribars as a functional food c. udayangani1, i. wijesekara1, 2*and i. wickramasinghe1 1department of food science & technology, faculty of applied sciences, university of sri jayewardenepura, gangodawila, nugegoda, sri lanka 2 centre for marine science & technology, faculty of applied sciences, university of sri jayewardenepura, gangodawila, nugegoda, sri lanka *correspondence: isuruw@sci.sjp.ac.lk; https://orcid.org/0000-0003-1688-8801 received: 29th april 2019, revised: 25th september 2019, accepted: 28th october 2019 abstract edible seaweed ulva lactuca is a rich source of dietary fiber, protein, and minerals, but currently underutilized in sri lanka. in the present study, nutribars (composed of cereals and golden syrups) were developed incorporating dried u. lactuca powder (moisture content; 15.29 ± 0.03%, dry basis) at 5 and 10% (w/w) ratios. seaweeds were manually collected in july, 2017 from matara, sri lanka, cleaned, and oven-dried at 60 ̊c for 8 h. the proximate composition, crude ulvan content, swelling capacity, water holding capacity (whc), and oil holding capacity (ohc) of powdered seaweed were evaluated. further, crude protein content was estimated in 0 (control), 5 and 10% of seaweed incorporated nutribars. the crude protein content in dried u. lactuca was 20.16 ± 0.16%. the whc of pulverized u. lactuca was 4.39 ± 0.07 g of water per g of seaweed powder, and ohc was 2.22 ± 0.27 g/g at room temperature (25 ̊c). significantly highest (p<0.05) protein content (8.55 ± 0.38%) was found for 10% u. lactuca added nutribar while it was 7.54% (± 0.15) and 7.89% (± 0.03) respectively for 0% and 5% seaweed added nutribars. moreover, the sensory evaluation results revealed that the nutribars incorporated with 5% u. lactuca (w/w) was shown almost similar sensory profile as the control except colour. however, 10% u. lactuca (w/w) added nutribars contained higher protein content than the control but rejected in overall acceptability. collectively, these results suggested that the under-utilized green seaweed u. lactuca can be incorporated at 5% (w/w) in nutribars. keywords: functional foods, nutraceuticals, nutribars, seaweeds, ulva lactuca. 1 introduction seaweeds are rich sources of bioactive functional food ingredients with different health beneficial biological activities (plaza et al. 2008, wijesekara https://orcid.org/0000-0003-1688-8801 c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 97 et al. 2011). sri lanka is an island with approximately 1700 km long sea-belt, and more than 500 species of seaweeds have been reported. seaweeds can be categorized in to three major groups such as green (chlorophyceae), brown (phaeophyceae), and red (rhodophyceae) seaweeds according to their pigment distribution. amongst them, edible green seaweed; sea lettuce (ulva sp.) has been consumed widely as a sea vegetable in china, japan, vietnam, and korea. furthermore, seaweeds have been consumed by these asian people dating back to 100 centuries but they have never found their rightful place in the daily diets of sri lankans. seaweed consumption is sporadically seen near the coastal areas but is still under-exploited. seaweed-derived ingredients and products are indirectly used commercially. as a natural food, the goodness of seaweeds has to be introduced and utilized. sri lanka contains a diverse species of green seaweeds including ulva lactuca (sea lettuce), u. faciata, u. intestinalis, chaetomorpha antenina, and caulerpa spp. (sea grapes) along the sea-belt of the country (durairatnam 1961). unfortunately, relatively less data is available for commercial exploitation of seaweeds of sri lanka (thadhani et al. 2019). therefore, it is a national need to explore the commercial potential of this valued underutilized marine resource of sri lanka. the objectives of the present study were to explore the nutritional composition of underutilized green seaweed u. lactuca and develop consumer acceptable nutribars incorporating u. lactuca as an alternative protein source. 2 material and methods 2.1 seaweed collection and processing the green seaweed (ulva lactuca) was manually collected during early july, 2017 from thalaramba beach, matara, sri lanka. the fresh seaweed was thoroughly washed with running tap water and dried at 60 ˚c in an electric oven (leader, taiwan) for 8 h. dried seaweed was then grinded and pulverized by using a grinder (panasonic mx -337n, japan), sieved through 355 micron sieve, and stored under the refrigerator condition (4 ̊c) until further use. 2.2 proximate analysis the moisture content of u. lactuca seaweed powder was determined according to the oven-dry method (aoac 925.10, 2005). the protein content of seaweed powder and nutribars were analyzed by kjeldhal method (aoac 978.04, 2005) and the n factor used in the calculation was 6.25 as generally c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 98 apply in food analysis. the lipid content of seaweed powder was determined according to a modified method (sanchez-machado et al. 2004). briefly, lipids were extracted from 2 g of the seaweed powder with 14 ml of a mixture of solvents (chloroform: methanol; 2:1, v/v) by subjected to vortex for 2 min. the residue was re-extracted two times and filtered through whatman no. 41 filter paper. the filtrates were pooled and concentrated to dryness in a water bath followed by measuring the final weight as lipids. in addition, the total ash content of seaweed powder was obtained gravimetrically (aoac 923.03, 2005). 2.3 extraction and fourier transform – infra red (ft-ir) analysis of crude ulvan the polysaccharide fraction crude ulvan was extracted from u. lactuca according to a previously published procedure (jiao et al. 2012) with some modifications. briefly, 4 g of dried seaweed powder was treated with methanol at room temperature (28 ̊c) for 2 h with magnetic stirring to remove pigments and lipids. the air-dried seaweed residue was then extracted with 100 ml of distilled water at 80 ˚c for 3 h. this step was repeated for another two times, followed by filtering through a muslin cloth, and the filtrates were pooled. the pooled filtrates were mixed with two volumes of absolute ethanol and allowed to precipitate the crude polysaccharide at room temperature for overnight. the precipitated crude ulvan was separated and dried in an oven (leader, taiwan) at 60 ˚c and subjected for ft-ir analysis. 2.4 swelling capacity (swc), water holding capacity (whc), and oil holding capacity (ohc) the swc of seaweed sample was analyzed by the bed volume technique after equilibrating in excess water (yaich et al. 2011). briefly, 1.0 g of seaweed powder was kept in a 10 ml measuring cylinder and 10 ml of distilled water were added. then, the mixture was vigorously stirred using a magnetic stirrer and allowed to settle for 20 h at room temperature (28 ˚c). the swelling capacity or swelling volume was measured and expressed as volume in ml of swollen sample per 1 g of sample. the whc of the u. lactuca seaweed powder was calculated by the modified centrifugation method reported previously (yaich et al. 2011). briefly, the seaweed powder (3 g) was dispersed in 25 ml of distilled water and placed in pre-weighed centrifuge tubes. then the centrifuge tubes were stirred for 5 min at the maximum speed and kept at the room temperature for 1 h followed by centrifugation (3000 g for 25 min). the supernatant was discarded and the water holding capacity of the seaweed was expressed as the g of water bound per 1 g of the residual sample in dry form. for the determination of ohc, 1 g of u. lactuca powder c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 99 was mixed with 12 ml of coconut oil in pre-weighed centrifuge tubes and followed the same procedure for whc. 2.5 manufacture of nutribars nutribars were made with ulva lactuca seaweed powder (0% as control, 5% and 10%, w/w) according to a common formulation (table 1). briefly, cowpea, green gram, and soya beans were roasted at 150 ˚c and finely grounded and mixed with rice flakes, corn flour, peanuts, and popcorns. the initial mixture (cowpea : green gram : popcorn : corn flour : soy flour : rice flakes : peanut was 12:10:10:8:5:5:5, respectively) was mixed thoroughly with golden syrup and with or without u. lactuca powder yielded the initial nutribar base. then, the sugar was mixed with water and heated under low flame till all the sugar crystals were dissolved. afterwards, the glucose syrup was added to this sugar syrup, boiled up to 112 ˚c, and removed from the flame followed by the mixing of previously prepared initial nutribar base. after the thorough mixing, the mixture was poured into moulds (length × width × thickness was 8.5 × 1.5 × 1.0 cm) and allowed to set. finally, the nutribars were removed from the moulds and packed in triple laminated pouches until further analysis. table 1: formulation of ulva lactuca incorporated nutribars (for 100 g of nutribar). ingredient (g) control (0% ulva) nutribar 5% ulva added nutribar 10% ulva added nutribar nutribar base 55 50 45 sugar 20 20 20 glucose syrup 10 10 10 butter 05 05 05 golden syrup 05 05 05 u. lactuca 00 05 10 water 05 05 05 2.6 analysis of product characteristics of nutribars the final quality attributes such as hardness and adhesiveness of the developed nutribars were measured using a laboratory texture analyzer (brookfield, usa). moreover, the hardness and adhesiveness were compared with a commercially available nutribar sample. in addition, the color values such as l*, a*, and b* values of developed nutribars were determined by using a chromometer (lovibond® lc1000, germany). c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 100 2.7 sensory evaluation the effect of sensory properties such as colour, intensities of aroma and taste, texture, and overall acceptability were evaluated among three types of nutribars by participants using a five-point scale (5 for “like extremely” down to 1 for “dislike extremely”) to score each attribute. thirty untrained panelists participated from the department of food science & technology, university of sri jayewardenepura, sri lanka. the three samples were labeled as 874 (control, 0% u. lactuca added), 759 (5% u. lactuca added), and 936 (10% u. lactuca added). nonparametric data obtained from this sensory evaluation were statistically analyzed by using kruskal-wallis test at 95% confidence level. the means separations were done by using mann-whitney test at 95% confidence level. 2.8 statistical analysis all experiments were carried out in triplicate and results were presented as mean ± standard deviation. the statistical analysis of data was carried out using one-way anova to test the significant difference of each variable (p<0.05), and followed by performed using the turkey test by the statistical software minitab 17 (pennsylvania, usa). 3 results and discussion 3.1 proximate composition of ulva lactuca the results of proximate analysis of dried and powdered u. lactuca seaweed samples are presented in the table 2. the dried and grinded seaweed powder (final moisture content %; 15.29 ± 0.03) contained considerable amount of protein (% in dry basis; 20.16 ± 0.16). therefore, underutilized sea lettuce u. lactuca in sri lanka has a potential to be introduced as alternative plantderived protein source in the local food industry. the protein content of seaweeds varies and depends on the species, season and environmental growth conditions (harnedy and fitzgerald 2011; peinado et al. 2014). for example, the protein content of palmaria palmata (rhodophyta) is higher in the winter-spring period than the summer-early autumn period (gallandirmouli et al. 1999). generally, brown seaweeds have lesser protein content (7 – 16%, dry weight) than red (21 – 47%, dw) and green seaweeds (10 – 26%, dw) (dawczynski et al. 2007; fleurence, 1999). however, seaweeds may contain non-protein nitrogen (n, such as free nitrates, pigments, nucleic c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 101 acids), resulting in an over estimation of their protein content (which estimated by general n-to-protein conversion factor of 6.25). therefore, the changes in protein content of u. lactuca in sri lankan sea waters need to be investigated though sri lanka lacks four seasons like sub-tropical countries. the studied u. lactuca in the present study contained 1.37 ± 0.05% (dry basis) of crude lipids. generally, seaweeds are low in lipids (1 – 5%, dry basis) according to the previously published reports (tabarsa et al. 2012). seaweeds are rich resources of essential minerals for human nutrition and health. the ash content (%) of u. lactuca was 17.17 ± 0.62 and this was in accordance with previously published data for this seaweed species. moreover, it has been reported that u. lactuca surprisingly contains the highest content of iron than some of terrestrial crops such as lettuce, cabbage, carrot, broccoli, and spinach (tabarsa et al. 2012). table 2: proximate composition of sea lettuce u. lactuca from matara coast, sri lanka. parameter composition (g/100 g of dry weight) moisture content 15.29 ± 0.03 crude fiber 36.78 ± 1.75 crude protein 20.16 ± 0.16 crude lipid 01.37 ± 0.05 ash content 17.17 ± 0.62 3.2 crude ulvan content ulvan is the major water-soluble sulfated polysaccharide (soluble dietaryfibers) found in green seaweeds of the order: ulvales and the common members are ulva spp. and enteromorpha spp. (jiao et al. 2011). the crude ulvan polysaccharide content (%) obtained in the present study was 17.21 ± 1.70 and this value is notably higher than the cold water and hot water extracted ulvan in previous reports (jiao et al. 2012; thanh et al. 2016). the yield and the chemical structure of seaweed polysaccharides differ from species to species and even in the same species the polysaccharide shows some structural differences according to stage of the seaweed life-cycle, anatomy, and geographical location (black et al. 1965). the ulvan from u. lactuca was mainly composed of rhamnose with variable contents of xylose and glucose and trace amounts of galactose and mannose (thanh et al. 2016). c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 102 fig. 1. ft-ir spectrum of the extracted crude ulvan from ulva lactuca current findings demonstrated that ulvan is health beneficial polysaccharide with different biological activities such as anticancer, antioxidant, antihyperlimidemic, and anticoagulant functions makes it suitable for a wide range of applications in the food industry (kim et al. 2011, wijesekara et al. 2011). based on ft-ir analysis (figure 1), the crude ulvan samples revealed polysaccharides features at about 3339.55 cm-1 (νo-h of -oh group), around 2950 cm-1 (νc-h of pyranose), 1601 cm -1 (νc=o of -coo) and 1030 cm -1 (νc-oh of glycoside). the strong absorption band at about 1214 cm-1 (νo=s=o of sulfate) and sharp band at 845.23 cm-1 (νc-o-s) suggested the presence of sulfate groups and substitution positions. in the current study, due to lack of purified ulvan as the standard sample, a comparison analysis was not carried out. however, the spectrum is similar to spectrums obtained for ulvan in previously published reports (jiao et al. 2012, tian et al. 2015, trivedi et al. 2016). 3.3 swc, whc, and ohc of seaweed powder to study the hydration properties of u. lactuca dried seaweed powder, swelling capacity (swc) and water holding capacity (whc) were determined at room temperature (28 ˚c). in the present study, the whc was 4.39 ± 0.07 g of water/g of dry seaweed powder and this value is lower than previously reported (6.66 g of water/g of dry seaweed) for the same seaweed species in another study (yaich et al. 2011). it has been reported that the whc of u. lactuca powder slightly increases with temperature probably related to the increment of solubility of fibers and proteins (fleury and lahaye 1991). the swc of the u. lactuca powder analyzed in the present study was 1.00 ± 0.10 ml/g of dry seaweed powder and this was higher than previously reported c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 103 (yaich et al. 2011). water exists in fibers in three forms; it is bound to the hydrophilic polysaccharides, it is held within the fiber matrix or it is trapped within the cell wall lumen. whc, determined by the centrifugation method in this study represented all three types of water associated with fibers. apart from different whc of fibers, the differences in whc and swc among the seaweed samples might be attributed to the different protein conformations and the variations in the number and nature of the water binding sites on protein molecules. these properties can influence the successful incorporation of fiber-enriched ingredients into foods. the ohc of u. lactuca powder at room temperature was 2.22 ± 0.27 g of oil/g of seaweed powder and this was notably higher than previously reported (1.68 g of oil/g of seaweed). however, coconut oil was used in this study to determine ohc since coconut oil is widely applied in cooking purposes of sri lanka. moreover, the ohc of u. lactuca is higher than some terrestrial sources such as orange (0.86-1.28 g/g dry weight) and peach (1.02-1.11 g/g dry weight) dietary fiber concentrates. the importance of this ohc property is related to the stabilization of foods with high fat content and emulsions, and to the physical entrapment of oil which acts as flavor retainer and increases the mouth-feel of foods. therefore, u. lactuca powder could be a potential functional ingredient to be incorporated in lipid-based foods; for example chocolates. fig. 2. sea lettuce ulva lactuca thalus and incorporated nutribars c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 104 3.4 properties of seaweed u. lactuca incorporated nutribars in the present study, nutribars were developed by incorporating u. lactuca seaweed powder at two different ratios (5 and 10%, w/w) and compared with the control (0% of seaweed powder) (figure 2). according to kjeldhal analysis, the crude protein contents of control, 5% u. lactuca added, and 10% u. lactuca added nutribars were 7.54 ± 0.15, 7.89 ± 0.03, and 8.55 ± 0.38, respectively. the results have confirmed that the protein content of 10% u. lactuca added sample showed significantly the highest protein content (p ≤ 0.05) of three types of nutribars. hardness is the amount of force required to bite through the sample with incisors and adhesiveness is the work necessary to overcome the attractive forces between the surface of the food and the surface of other materials with which the food comes into contact (e.g. tongue, teeth, and palate). according to instrumental texture analysis (table 3) significant differences in hardness and adhesiveness of nutribar samples were obtained. there was no significant difference of hardness in between control and 5% u. lactuca added nutribar samples. however, hardness of 10% u. lactuca added nutribar samples and commercially available nutribars were significantly different from each other. when adding more pulverized seaweed powder, hardness of the nutribar was increased. hardness of commercially available nutribar was greater than control and 5% u. lactuca added sample. decreasing order of adhesiveness of nutribars was 10% u. lactuca added > control > 5% u. lactuca added > commercially available sample. therefore, force required to remove 10% u. lactuca added nutribar from mouth is the highest. moreover, commercially available nutribar was the lowest. therefore, required force to remove from mouth is very low. reason for this is commercially available nutribar was crispy in texture. table 3: texture profile analysis results of control, u. lactuca added, and commercial nutribars. parameter control 5% u. lactuca 10% u. lactuca commercial hardness (g) 1035.0 ± 62.2c 1356.0 ± 281c 2146.3 ± 116.9a 1741.3 ± 78.9b adhesiveness (mj) 5.09 ± 1.28b 2.82 ± 3.17b,c 9.08 ± 0.51a 1.11 ± 0.89c values for texture profile analysis are means ± sd; n=3, and means in the same row followed by different letters are significantly different (p < 0.05). the colour analysis results of nutribars were presented in table 4. lightness of 5% u. lactuca added nutribars had significantly the highest lightness compared to other two bars. lightness of 10% u. lactuca added sample had c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 105 the significantly lowest lightness. powdered seaweed u. lactuca contains very low lightness value and dark colour appears when drying the raw seaweed samples due to the degradation of chlorophyll pigments (ali et al. 2014). table 4: mean chroma meter colour values for control, 5%, and 10% u. lactuca added nutribars. type of nutribar l* a* b* control 45.30 ± 1.32a 11.33 ± 0.67a 19.23 ± 0.91a 5% ulva added 57.43 ± 1.08b 4.87 ± 1.25b 20.80 ± 0.89a 10% ulva added 41.37 ± 1.30c 2.70 ± 0.10c 15.53 ± 0.59b values for colour analysis are means ± sd; n=3, and means in the same column followed by different letters are significantly different (p < 0.05). the negative a * values indicate the closeness to green and positive values indicate the closeness to red. the results showed nutribars are significantly different in a * value and the control was closeness to red colour followed by 5% and 10% u. lactuca added nutribars. negative b * values indicate closeness to blue colour and positive values indicate the closeness to yellowness. according to results, all three types of nutribars were shown to be as close to yellow than blue. the yellowness of control nutribars and 5% u. lactuca added samples were not significantly different and significantly the lowest b * value was shown by 10% u. lactuca added nutribars. fig. 3. web diagram of the average rank of three nutribars from sensory analysis c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 106 the effect of sensory properties such as colour, intensity of u. lactuca aroma, intensity of u. lactuca taste, texture and overall acceptability was evaluated among three different nutribars (figure 3). appearance of each and every sample is significantly different (p ≤ 0.05) from each other. there was no significant different (p ≤ 0.05) between aroma of 874 sample and 759 sample. however, there was a significant difference in between 874 and 936 samples as well as in between 759 and 936 samples were obtained. as observed in the present study, except for the appearance, all other sensory attributes were similar for 874 (control) and 759 (5% u. lactuca added) nutribars. the green colour and the aroma from the seaweed were the reason for the less consumer acceptance for 10% u. lactuca added nutribars. 4 conclusions the results obtained in the present study clearly demonstrated that the edible green seaweed u. lactuca (sea lettuce) from matara, sri lanka is rich in soluble dietary fiber, proteins, and minerals. this seaweed has a potential to be introduced to the food industry to develop novel functional food products since it has acceptable functional properties such as swelling capacity, water holding, and oil holding capacities. moreover, seaweed-incorporated food products have a leading consumer trend in the western countries. hence, promoting these seaweed-added foods will be a new resource to generate an additional income to the coastal community as well as foreign income to the country. collectively, it can be concluded that the incorporation of underutilized u. lactuca in nutribars is possible and acceptable at 5% (w/w) ratio. however, addition of aroma masks can be recommended when incorporating this seaweed in higher ratios to develop new food products. acknowledgements this research was supported by the university of sri jayewardenepura, sri lanka. research assistance in ft-ir analysis by dr. asitha cooray and team, instrument centre at faculty of applied sciences, university of sri jayewardenepura is gratefully acknowledged. technical assistance in proximate analysis from ms. hasanthika sandarenu, and comments of two anonymous reviewers are acknowledged. references ali ma, yusof ya, chin nl, ibrahim mn, basra sma. 2014. drying kinetics and colour analysis of moringa oleifera leaves. agriculture & agricultural science procedia 2: 394400. black wap, blakemore wr, colquhoun ja, dewar et. 1965. the evaluation of some red marine algae as a source of carrageenan and of its κand λ-components. journal of the c. udayangani et al. ulva lactuca incorporated nutribars ruhuna journal of science vol 10 (2): 96-107, december 2019 107 science of food & agriculture 16(10): 573-585. dawczynski c, schubert r, jahreis g. 2007. amino acids, fatty acids, and dietary fibre in edible seaweed products. food chemistry 103: 891–899. durairatnam m. 1961. contribution to the study of marine algae of ceylon. bulletin of fisheries resources station, ceylon 10:5-117. fleurence j. 1999. seaweed proteins: biochemical, nutritional aspects and potential uses. trends in food science & technology 10(1): 25-28. fleury n, lahaye m. 1991. chemical and physico-chemical characterisation of fibres from laminaria digitata (kombu breton): a physiological approach. journal of the science of food & agriculture 55(3): 389-400. galland-irmouli av, fleurence j, lamghari r, lucon, m, rouxel c, barbaroux o, bronowicki jp, villaume c, gueant jl. 1999. nutritional value of proteins from edible seaweed palmaria palmata (dulse). the journal of nutritional biochemistry 10(6): 353-359. harnedy pa, fitzgerald r. 2011. bioactive proteins, peptides, and amino acids from macroalgae. journal of phycology 47(2): 218-232. jiao g, yu g, wang w, zhao x, zhang j, ewart sh. 2012. properties of polysaccharides in several seaweeds from atlantic canada and their potential anti-influenza viral activities. journal of ocean university of china 11(2): 205-212. jiao g, yu g, zhang j, ewart hs. 2011. chemical structures and bioactivities of sulfated polysaccharides from marine algae. marine drugs 9(2): 196-223. kim sk, pangestuti r, rahmadi p. 2011. sea lettuces: culinary uses and nutritional value. advances in food & nutrition research 64: 57-70. peinado i, girón j, koutsidis g, ames jm. 2014. chemical composition, antioxidant activity and sensory evaluation of five different species of brown edible seaweeds. food research international 66: 36–44. plaza m, cifuentes a, ibanez e. 2008. in the search of new functional food ingredients from algae. trends in food science & technology 19(1): 31-39. sanchez-machado di, lopez-cervantes j, lopez-hernandez j, paseiro-losada p. fatty acids, total lipid, protein and ash contents of processed edible seaweeds. food chemistry 85(3): 439-444. tabarsa m, rezaei m, ramezanpour z, waaland jr. 2012. chemical compositions of the marine algae gracilaria salicornia (rhodophyta) and ulva lactuca (chlorophyta) as a potential food source. journal of the science of food & agriculture 92(12): 2500-2506. thadhani vm, lobeer a, zhang w, irfath m, su p, edirisinghe n, amaratunge g. 2019. comparative analysis of sugar and mineral content of sargassum spp. collected from different coasts of sri lanka. journal of applied phycology, accepted manuscript, doi.org/10.1007/s10811-019-01770-4 thanh ttt, quach tmt, nguyen tn, luong dv, bui ml, tran ttv. 2016. structure and cytotoxic activity of ulvan extracted from green seaweed ulva lactuca. international journal of biological macromolecules 93(a): 695-702. tian h, yin x, zheng q, zhu l, chen j. 2015. isolation, structure, and surfactant properties of polysaccharides from ulva lactuca l. from south china sea. international journal of biological macromolecules 79: 577-582. trivedi n, baghel rs, bothwell j, gupta v, reddy crk, lali am, jha b. 2016. an ingrated process for the extraction of fuel and chemicals from marine macroalgal biomass. scientific reports 6: 30728, doi: 10.1038/srep30728 wijesekara i, pangestuti r, kim sk. 2011. biological activities and potential health benefits of sulfated polysaccharides derived from marine algae. carbohydrate polymers 84(1): 1421. yaich h, garna h, besbes s, paquot m, blecker c, attia h. 2011. chemical composition and functional properties of ulva lactuca seaweed collected in tunisia. food chemistry 128(4): 895-901. ruhuna journal of science vol 10(1): 65-76, june 2019 eissn: 2536-8400 faculty of science doi: http://doi.org/10.4038/rjs.v10i1.51 university of ruhuna  faculty of science, university of ruhuna sri lanka 65 assessment of tree species resistance to air pollution around a metal-scrap recycling factory using air pollution tolerance index and anticipated performance index s. oyedeji 1 *, o.o. agboola 2 , j.k. oyekunle 1 , d.a. animasaun 1 and p.o. fatoba 1 1 department of plant biology, university of ilorin, 240003, ilorin, nigeria 2 department of botany, university of lagos, akoka, lagos, nigeria. *correspondence: oyedeji.s@unilorin.edu.ng; https://orcid.org/0000-0001-7357-1152 received: 23 rd march 2018, revised: 31 th december 2018, accepted: 05 th march 2019 abstract air pollution is one of the major global tribulations in many developing cities around the world. this study evaluates pollution resistance of terminalia catappa, anacardium occidentale and tectona grandis growing around a metal scrap-recycling factory in osun state, nigeria using air pollution tolerance index (apti) and anticipated performance index (api) with the view of recommending the species for greenbelt development in urban spaces. biochemical parameters such as ascorbic acid, total chlorophyll content, foliar extract ph and relative water content were analyzed in fresh leaves harvested from the tree species growing around the metal-scrap recycling factory (es) and a relatively unpolluted control site (cs). apti and api were obtained from the results of biochemical variables. the results showed that biochemical parameters (ascorbic acid, total chlorophyll, foliar extract ph and relative water content) and apti varied significantly (p<0.05) among tree species and between the sites. the order of apti varied as t. catappa (15.66) > a. occidentale (13.53) > t. grandis (13.17) with plants having a higher index at cs than at es. all three-tree species showed intermediate tolerance but t. catappa and a. occidentale were assessed as good performers while t. grandis performed moderately in polluted sites. the study recommended t. catappa and a. occidentale over t. grandis for use in developing greenbelts in urban centres especially in highly polluted areas, such as the vicinity of factories. keywords: atmospheric pollution, green belt development, pollution monitoring, tolerance. 1. introduction mailto:oyedeji.s@unilorin.edu.ng https://orcid.org/0000-0001-7357-1152 s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science 66 vol 10(1): 65-76, june 2019 pollution, especially atmospheric pollution, is one of the major problems arising from human population explosion and industrialization (odilora et al. 2006). it is true that industrialization improves the wealth of nations, but the degradation of environmental quality via pollution is a cause for concern. most developing countries, especially those in sub-saharan africa, depend mainly on revenues from industries (omoju 2014), hence their proliferation is considered a huge gain. in nigeria for example, metal scrap-recycling is fast becoming a major activity due to increasing per capita consumption of steel and shortage of iron ore in the country. this has led to the proliferation of metal recycling factories in urban centres and suburbs, where waste and scavenged metals are rolled to steel (ohamain 2013). these factories are becoming a cause for concern as their operations release toxic gases (such as sulphur and nitrogen oxides, carbon monoxide) and particles (such as sooth, metal particles and organic molecules) into the atmosphere. air pollutants represent a complex mixture of organic and inorganic substances of varying states and sizes that can enter tissues of living organisms in a number of ways (thakar and mishra 2010). many plants are very sensitive to air pollutants which can damage their leaves, impair plant growth and limit primary productivity (agrawal 1985, 2005, thakar and mishra 2010). much experimental work has been conducted on the analysis of air pollutants effects on crops and vegetation at various levels ranging from biochemical to ecosystems levels (hill 1971, cavanagh et al. 2009, gupta et al. 2011). often times, plants exposed to pollutants experience changes in their biochemical functioning (such as those of ascorbic acid, chlorophyll and relative water contents) even before visible damage to leaves can be noticed (liu and ding 2008). these changes in plants serve as the basis for pollution monitoring (prusty et al. 2005). bio-monitoring of air pollution has been found to be extremely useful in detecting the kind and level of pollutants in the air with or without measurements (prusty et al. 2005). studies on air pollution impacts on vegetation have shown that it alters ascorbic acid content (hoque et al. 2007) chlorophyll content (flowers et al. 2007), leaf extract ph (klumpp et al. 2000) and relative water content (rao 1977). these parameters separately produced conflicting results for same plant species (han et al. 1995). however, the air pollution tolerance index (apti) based on all four parameters has been used for identifying tolerance levels of plant species (singh et al. 1991, singh and rao 1993, liu and ding 2008, ogunkunle et al. 2015). air pollution tolerance index (apti) is an inherent quality of plants to encounter air pollution stress which at present is a prime concern, particularly, in urban and industrial areas. the non-mobile (stationary) nature of plant species exposes them continuously to point source pollution. this makes air pollution impact on plants to be proportional to the intensity of pollution in the environment. apti is a species-dependent attribute and expresses the inherent ability of a plant species to withstand stress emanating from pollution (lohe et al. 2015). it also informs the public s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science vol 10(1): 65-76, june 2019 67 of the local ambient air pollution status and the potential risk it would impose, particularly on vulnerable groups such as children, old people and those with existing cardiovascular and respiratory diseases (rajamanickam and nagan 2018). categorization of plants as sensitive or tolerant is determined by the level of biochemical parameters used to compute the apti for a species. tolerant plant species are sinks and living filters minimizing air pollution impact by absorption, adsorption, detoxification, accumulation, and/ or metabolization without sustaining serious foliar damage or decline in growth in the face of pollution (mondal et al. 2011). apti determination thus provides a reliable method for screening a large number of plants with respect to their resistance and susceptibility to air pollutants. this study assesses air pollution resistance of indian almond (terminalia catappa), cashew (anacardium occidentale) and teak (tectona grandis) growing around a metal scrap-recycling factory in osun state, south-west nigeria using their apti and api with the view of recommending the species for greenbelt development in urban spaces. 2 material and methods 2.1 study area the study was carried out in march, 2017 on tree stands growing around prism steel rolling mill, a metal scrap-recycling factory located on latitude n07°52'42.6" and longitude e004°39'12.0" (figure 1). the factory is situated along ikirun-osogbo expressway, ikirun, osun state. the factory became fully operational in 2012, and has since been recycling metal scraps into iron rods of 8, 12, 16, 20 and 32 mm in diameter of standard lengths. the scrap metals are conveyed to the factory from different parts of the country by heavy duty trailers. the pollution from metal smelting activities and those from exhaust fumes of trailers moving in and out of the factory make the study important. the prevailing climate in the site is distinctively tropical with annual rainfall of about 1400 mm and daily air temperature ranging from 23°c to 33°c. the relative humidity is about 55% during the dry season and about 90% during the rainy season. the study locations were classified into two zones: (a) the experimental site (es) and (b) the control site (cs) which is an agricultural land about 30 km away from the experimental site (figure 1). 2.2 sample collection three tree species (terminalia catappa l., tectona grandis l.f. and anacardium occidentale l.) established around the perimeter fencing of the s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science 68 vol 10(1): 65-76, june 2019 experimental site and also found in the control site were used for the study. tree stand selected for the study were less than 5 years old, and possess diameter of greater than 10 cm at breast height. fig. 1. maps of nigeria and osun state showing the location of the experimental and control sites. matured fresh leaf samples (~ 1 kg each) were collected in the morning from five selected stands of each tree species in the experimental site and control site. the leaf samples were immediately transferred to the laboratory for determination of the ascorbic acid, total chlorophyll content, leaf extract ph and relative water content (rwc). ascorbic acid (aa) content of leaves was determined using the spectrophotometric method (begum and harikrishna 2010). total chlorophyll content was determined by the spectrophotometric method described by arnon (1949). leaf extract ph was determined using 2.5 g fresh leaf sample homogenized in 10 ml distilled water using warring blender and read using a pre-calibrated digital ph meter (orion star benchtop, thermoscientific). relative water content (% rwc) of leaves was determined using the method of liu and ding (2008). apti was determined according to the method proposed by singh and rao (1983). tree species were classified as s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science vol 10(1): 65-76, june 2019 69 sensitive, intermediate or tolerant based on apti class proposed by padmavathi et al. (2013). 2.3 data analysis two-way analysis of variance (anova) was used to compare biochemical parameters and apti for three tree species in the experimental site (es) and control site (cs). significant means were separated using fisher’s protected lsd at α level of 0.05 using sas software 9.1.3 for windows. 2.4 determination of anticipated performance index average apti values with relevant biological characters (including plant type and habit, canopy structure, laminar size, texture and hardiness) and socioeconomic character (economic value) grading (using + or -) were combined to obtain the anticipated performance index (api) for the tree species. allotted points were scaled for grading based using the method of prajapati and tripathi (2008) as adopted by ogunkunle et al. (2015). 3. results & discussion 3.1 effect of air pollution on ascorbic acid ascorbic acid is an antioxidant found in all growing plants. it positively influences resistance to environmental stresses including air pollution (lima and fernandez 2000, mondal et al. 2011). the concentration of foliar ascorbic acid in the leaves varied significantly with tree species (f2, 24= 285.74; p< 0.01) and site (f1, 24= 5.79; p= 0.02). terminalia catappa showed the highest concentration at es (2.47 mg/g) and cs (2.44 mg/g), while tectona grandis showed the lowest concentration of 1.86 mg/g and 1.76 mg/g respectively (fig. 2a). the variation in foliar ascorbic acid concentrations among the trees species reflects species specificity to air pollution. there was no significant interaction between tree species and site (f2, 24= 1.02; p= 0.38) for foliar ascorbic acid concentration (table 1a). there was significant variation in the concentrations of ascorbic acid in tectona grandis at es and cs (fig. 2a). this result is consistent with the findings of ogunkunle et al. (2015) who also reported increases in aa in polluted areas. increases in ascorbic acid in plants in polluted areas (such as around the metal recycling factory in this study) has been speculated to be favoured by increasing production of reactive oxygen species (ros) (mondal et al. 2011). researchers are of the opinion that higher ascorbic acid content signals plant species tolerance against air s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science 70 vol 10(1): 65-76, june 2019 pollution, especially exposure to sulphur dioxide (varshney and varshney 1984, singare and talpade 2012). fig 2. ascorbic acid (a), total chlorophyll (b), leaf extract ph (c), and (d) relative water content of tree species around metal recycling factory and control site. bars represent mean ± standard deviations for n = 15. 3.2 effect of air pollution on total chlorophyll total chlorophyll concentration was significantly different between tree species (f2, 24 = 101.98; p< 0.01) and site (f1, 24 = 70.48; p< 0.01). there was also significant interaction between tree species and site (f2, 24 = 11.98; p< 0.01) (table 1a) indicating that site condition affected the tree species differently. tree species in cs had higher total chlorophyll than those at es, except for anacardium occidentale. tectona grandis in cs had the highest concentration of total chlorophyll (42.51 mg/g) while t. catappa at es had the lowest (19.30 mg/g) (figure 2b). the higher chlorophyll content is the control site is consistent with the result of gholami, et al. (2016) for tree species around polluted areas in iran. reduction in total chlorophyll concentrations in t. catappa a. occidentale t. grandis 0.0 0.5 1.0 1.5 2.0 2.5 3.0 a sc o rb ic a c id ( m g /g ) cs es (a) t. catappa a. occidentale t. grandis 0 10 20 30 40 50 t o ta l c h lo ro p h y ll ( m g /g f w ) cs es (b) t. catappa a. occidentale t. grandis 0 1 2 3 4 5 6 l e a f e x tr a c t p h cs es (c) t. catappa a. occidentale t. grandis 0 20 40 60 80 100 r e la ti v e w a te r c o n te n t (% ) cs es(d) s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science vol 10(1): 65-76, june 2019 71 tree species growing around the factory supports the argument that chloroplast is the primary site of attack by air pollutants leading to the increase in chlorophyllase activity and resultant degradation of chlorophyll (ninave, et al. 2001, tripathi and gautam 2007). s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science 72 vol 10(1): 65-76, june 2019 table 1a: f-statistics for biochemical parameters and atpi of tree species in experimental site (around metal scrap recycling factory) and control site. source of variation ascorbic acid total chlorophyll ph rwc apti df f p-value f p-value f p-value f p-value f p-value species 2 285.74 <0.0001 101.98 <0.0001 16677.90 0.0002 536.01 <0.0001 97.88 <0.0001 site 1 5.79 0.0242 70.48 <0.0001 18.87 <0.0001 17.82 0.0003 45.69 <0.0001 site× species 2 1.02 0.3774 11.91 0.0003 1.18 0.3232 2.02 0.1552 14.34 <0.0001 error 24 df = degrees of freedom. f = f-ratio (fisher’s ratio); rwc = relative water content; apti = air pollution tolerance index. table 1b: means of biochemical parameters and atpi of tree species in experimental site (around metal scrap recycling factory) and cont rol site. factor ascorbic acid total chlorophyll leaf extract ph relative water content atpi species terminalia catappa 2.46 a 24.51 b 4.855 b 84.56 a 15.66 a anacardium occidentale 2.34 b 23.58 b 3.637 c 71.78 b 13.53 b tectona grandis 1.81 c 37.31 a 5.456 a 54.79 c 13.17 c lsd0.05 0.0486 2.2189 0.0209 1.8829 0.3958 site cs 2.17 b 32.15 a 4.67 a 68.81 b 14.65 a es 2.23 a 24.78 b 4.63 b 71.95 a 13.59 b lsd0.05 0.0595 1.8117 0.0171 1.5374 0.3231 means with different superscripted letter are significantly different at p<0.05. s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science vol 10(1): 65-76, june 2019 73 table 2: air pollution tolerance index (apti) and class for tree species growing around a metal recycling factory (es) and control site (cs). tree species cs es average apti class terminalia catappa 16.72±0.45 a† 14.59±0.37 a 15.66 intermediate anacardium occidentale 13.58±0.41 b† 13.49±0.32 b 13.53 intermediate tectona grandis 13.65±0.65 b† 12.70±0.27 c 13.17 intermediate mean ± sd with the same letter down a column are not significantly different at p>0.05. † indicate significantly higher apti value between the sites (cs and es). table 3: anticipated performance index (api) grade and assessment for tree species around metal and steel recycling factory. tree species apti tree habit canopy structure type of plant laminar structure economic value total + % score api size texture hardiness grade assessment terminalia catappa +++ ++ ++ + + + + 11 68.75 4 good anacardium occidentale ++ + ++ + + + + ++ 11 68.75 4 good tectona grandis ++ ++ + ++ + + 9 56.25 3 moderate apti = air pollution tolerance index. s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science 74 vol 10(1): 65-76, june 2019 3.3 effect of air pollution on leaf extract ph leaf extract ph varied significantly among the tree species (f2, 24= 16677.90; p< 0.01) and sites (f1, 24= 18.87; p< 0.0001). there was no significant interaction between tree species and site (f2, 24= 1.18; p= 0.32) for leaf extract ph (table 1a). plants at es showed significantly lower leaf ph than in cs (table 1b). tectona grandis at cs showed the highest ph (5.47) while a. occidentale at es showed the least (3.62). terminalia catappa showed the highest reduction in ph (4.88 to 4.83) (fig. 2c). lower leaf extract ph in plants at es may be due to the presence of acidic pollutants such as so2 and nox in air vicinity of the factory. scholz and reck (1977) and escobedo, et al. (2008) also reported that acidic pollutants have the potentials to lower leaf extract ph. pollutants such as so2 and nox diffuse through leaf openings (stomata) and react with cellular water to form acid radicals in leaf matrix (mondal et al. 2011). 3.4 effect of air pollution on relative water content relative water content (rwc) was significantly different among the tree species (f2, 24 = 52.64; p< 0.01) and site (f1, 24 = 17.82; p< 0.01) (table 1a). the decreasing order of rwc was t. catappa> a. occidentale> t. grandis and was consistent for es and cs (fig. 2d). rwc ranges were 50.40% 84.05% at cs and 56.06% 87.90% at es. generally, plants at es showed higher rwc than those at cs (fig. 2d). rwc is associated with protoplasmic permeability in cells which causes loss of water and dissolved nutrients, resulting in early senescence of leaves (masuch et al. 1988). more water in a leaf will help to maintain its physiological balance under stress condition of air pollution (dedio 1975). therefore the plants with high relative water content such as t. catappa under polluted conditions may be tolerant to pollutants. 3.5 air pollution tolerance index of the tree species air pollution tolerance index (apti) varied significantly with tree species (f2, 24 = 97.88; p< 0.01) and site (f1, 24 = 45.69; p< 0.01). there was also significant interaction of site with species for apti (f2, 24 = 14.34; p< 0.01) (table 1a) indicating the environmental condition at the site had varying influence on the tree species. decreasing order of apti was t. catappa (15.66) > a. occidentale (13.53) > t. grandis (13.17). species at cs had higher apti values than those at es (table 1b and table 2). according to the apti classification, all tree species studied showed intermediate tolerance to pollution (table 2). the apti values in this study exceeded those reported by s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science vol 10(1): 65-76, june 2019 75 ogunrotimi, et al. (2017); t. catappa (12.5), a. occidentale (10.7) t. grandis (12.1) in polluted urban areas of ile-ife, osun state. tak and kakde, (2017) also reported lower apti values for similar species in urban centres of india. contrarily, thakar and mishra, (2010) reported higher apti values than in the current study for a. occidentale (22.17) and t. grandis (20.97) growing around an aluminium factory in jharsuguda, india. nayak, et al. (2015) also confirmed that tectona grandis and terminalia catappa showed intermediate sensitivity in polluted sites. since species tolerance to pollution is best explained by their apti value, it can be deduced that terminalia catappa is most tolerant while tectona grandis is least tolerant. 4 conclusions this study showed that assessment of tree species tolerance against air pollution using individual biochemical parameter (such as ascorbic acid, total chlorophyll, leaf extract ph or relative water content) alone may be misleading. for instance, tectona grandis which showed the highest foliar ph and total chlorophyll concentration showed the least tolerance to air pollution (apti). based on the api assessment, t. catappa and anarcadium occidentale are recommended to be used for greenbelt in polluted environments, such as vicinity of factories, as they proved to be good performers as opposed to tectona grandis that performed only moderately. acknowledgement the authors appreciate mr. d. o. jemirin, the laboratory technologist in the department of industrial chemistry for assisting with the determination of biochemical parameters in the leaf samples. comments from two anonymous rjs reviewers are acknowledged. references agrawal m. 1985. plant factors as indicators of so2 and o3 pollutants. in biological monitoring of the state of the environment (bioindicators). indian national science academy, new delhi, pp. 225-231. agrawal m. 2005. effects of air pollution on agriculture: an issue of national concern. national academy of science letters 23(3&4): 93-106. arnon di. 1949. copper enzymes in isolated chloroplasts polyphenol oxidase in beta vulgaris. plant physiology 2(1):1–15. begum a, harikrishna s. 2010. evaluation of some tree species to absorb air pollutants in three industrial locations of south bengaluru, india. european journal of chemistry 7: 151–156. cavanagh j-ae, zawar-reza p, wilson jg. 2009. spatial attenuation of ambient particulate matter air pollution within an urbanised native forest patch. urban forestry & urban greening, 8:21-30 doi:http://dx.doi.org/10.1016/j.ufug.2008.10.002 s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science 76 vol 10(1): 65-76, june 2019 dedio w. 1975. water relations in wheat leaves as screening test for drought resistance. canadian journal of plant science 55(2), 369-378. escobedo f, nowak dj, wagner j, de la maza cl, rodríguez m, crane de. 2008. analyzing the cost-effectiveness of santiago chile’s policy of using urban forests to improve air quality. journal of environmental management 86: 148–157. flowers md, fiscus el, burkey s. 2007. photosynthesis, chlorophyll fluorescence and yield of snap bean (phaseolus vulgaris l.) genotypes differing in sensitivity to ozone. environmental and experimental botany, 61: 190-198. gholami a, mojiri a, amini h. 2016. investigation of the air pollution tolerance index (apti) using some plant species in ahvaz region. the journal of animal & plant sciences 26(2): 475-480. gupta s, nayek s, bhattacharya p. 2011. effect of air-borne heavy metals on the biochemical signature of tree species in an industrial region, with an emphasis on anticipated performance index. chemistry and ecology 27: 381-392. han y, wang qy, han gh. 1995. the analysis about sod activities in leaves of plants and resistance classification of them. journal of liaoning university (natural sciences edition) 22: 71-74. hill ac. 1971. vegetation: a sink for atmospheric pollutants. journal of the air pollution control association 21:341-346 hoque ma, banu mna, okuma e. 2007. exogenous proline and glycine betaine increase nacl induced ascorbate–glutathione cycle enzymes activities, and proline improves salt tolerance more than glycinebetaine in tobacco bright yellow-2 suspension cultured cells. journal of plant physiology 164: 1457-1468. klumpp g, furlan cm, domingos m. 2000. response of stress indicators and growth parameters of tibouchina pulchra cogn, exposed to air and soil pollution near the industrial complex of culoutao, brazil. science of the total environment 246:79-91. lima js, fernandez eb. 2000. mangifera indica and phaseolus vulgaris in the bioindicator of air pollution in bahia, brazil. ecotoxicology and environmental safety 46(3): 275-278. liu y-j, ding h. 2008. variation in air pollution tolerance index of plant near a steel factory; implications for landscape-plant species selection for industrial areas. wseas transactions on environment and development 4(1): 24-32. lohe rn, tyagi b, singh v, tyagi pk, khanna dr, bhutian r. 2015. a comparative study for air pollution tolerance index of some terrestrial plant species. global journal of environmental science and management 1(4): 315-324. masuch g, kicinski hc, kettrup a, boss ks. 1988. single and combined effects of continuous and discontinuous o3 and so2 emission on norway spruce needless. 1. historical and cytological changes. international journal of environmental analytical chemistry 32: 213241. mondal d, gupta s, datta jk. 2011. anticipated performance index of some tree species considered for green belt development in urban area. international research journal of plant science 2(4): 99-106 nayak d, patel dp, thakare hs, satashiya k, shrivastava pk. 2015. evaluation of air pollution tolerance index of trees. research in environment and life sciences 8(1): 7-10. ninave sy, chaudhaun pr, gajghate dg, tarar jt. 2001. foliar biochemical features of plants as indicators of air pollution. bulletin of environmental contamination and toxicology 67: 133-140. odilara ca, egwaikhide pa, esekheigbe a, emua sa. 2006. air pollution tolerance indices (apti) of some plant species around llupeju industrial area, lagos. journal of engineering science and applications 4(2): 97-101. ogunkunle co, suleiman lb, oyedeji s, awotoye oo, fatoba po. 2015. assessing the air pollution tolerance index and anticipated performance index of some tree species for biomonitoring environmental health. agroforestry systems 89(3): 447–454. ogunrotimi dg, adereti fk, eludoyin ao, awotoye, oo. 2017. urban air pollution control: selection of trees for ecological monitoring using anticipated performance indices in a s. oyedeji et al. assessment of tree species resistance to air pollution ruhuna journal of science vol 10(1): 65-76, june 2019 77 medium-size urban area in southwest nigeria. interdisciplinary environmental review 18(1): 40–54. ohamian ei. 2013. scrap iron and steel recycling in nigeria. greener journal of environmental sciences and public safety 2:1-9. omoju o. 2014. environmental pollution is inevitable in developing countries. in: breaking energy, climate change infrastructure international. accessed at https://breakingenergy. com/2014/09/23/environmental-pollution-is-inevitable-in-developing-countries/ padmavathi p, cherukuri j, reddy ma. 2013. impact of air pollution on crops in the vicinity of a power plant: a case study. international journal of engineering research and technology 2(12): 3641–3651. prajapati sk, tripathi bd. 2008. anticipated performance index of some tree species considered for green belt development in and around an urban area: a case study of varaasi city, india. journal of environment management 88:1343–1349. prusty bak, mishra pc, azeez pa. 2005. dust accumulation and leaf pigment content in vegetation near the national highway at sbp, orissa, india. ecotoxicology and environmental safety 60: 228-235. rajamanickam r, nagan s. 2018. assessment of air quality index for cities and major towns in tamil nadu, india. journal of civil and environmental engineering 8: 304. doi: 10.4172/2165-784x.1000304 rao dn. 1977. use of plants as indicators and monitors of so2 pollution. chemical age of india 28: 665–672 scholz f, reck s. 1977. effects of acids on forest trees as measured by titration invitro inheritance of buffering capacity in piceaabies. water, air soil pollution 8(1): 41-45. singh sk, rao dn. 1983. evaluation of plants for their tolerance to air pollution. in: proceedings of symposium on air pollution control, indian association for air pollution control, new delhi, india, pp. 218–224. singh sk, rao dn, agrawal m, pandey j, narayan d. 1991. air pollution tolerance index of plants. journal of environmental management 32: 45-55. singare pu, talpade ms.2012. physiological responses of some plant species as a bio-indicator of roadside automobile pollution stress using the air pollution tolerance index approach. international journal of plant research 3(2): 9-16. tak aa, kakde ub. 2017. assessment of air pollution tolerance index of plants: a comparative study. international journal of pharmacy and pharmaceutical sciences 9(7): 83-89. thakar bk, mishra pc. 2010. dust collection potential and air pollution tolerance index of tree vegetation around vedanta aluminium limited, jharsuguda. the bioscan 3: 603-612. tiwari s, bansal s, rai s. 1993. air pollution tolerance index of some plants in urban areas of bhopal. advanced ecologia 16(1): 1-8. tripathi ak, gautam m. 2007. biochemical parameters of plants as indicators of air pollution. journal of environmental biology 28(1): 127-132. varshney srk, varshney ck. 1984. effects of sulphur dioxide on ascorbic acid in crop plants. environmental pollution 35(4): 285-290. ruhuna journal of science vol 13 (2): 205-216, december 2022 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v13i2.126 university of ruhuna faculty of science, university of ruhuna sri lanka 205 guard hair micro-morphology of four non-human primates in shasha forest reserve, osun state, nigeria fatsuma olaleru*1, 2, temitope r. olugbebi1 and michelle i. fasona1 1 department of zoology, faculty of science, university of lagos, nigeria 2 tetfund centre of excellence for biodiversity conservation and ecosystem management, university of lagos, nigeria *correspondence: folaleru@unilag.edu.ng orcid: https://orcid.org/0000-0002-3362-8668 received: 28th april 2022, revised: 11th november 2022, accepted: 29th december 2022 abstract mammalian guard hairs have been used for their identification and have been proved useful in wildlife population surveys and trafficking. the qualitative and quantitative features of the dorsal guard hairs of four non-human primates (nhps) from shasha forest reserve were studied using standard procedures to determine their differences. the nhps were mona (mm), putty-nosed (pnm), and white-throated (wtm) monkeys, and red-capped mangabey (rcm). the qualitative features determined were medulla pattern, and structure, scale margin distance and type, and scale pattern. the quantitative values studied were scale length and width, shaft diameter (μm), medulla diameter, and medulla index and fraction. continuous medulla pattern was the only qualitative feature common to all the species. the mm hair had the highest recorded values for all morphological characteristics except shaft diameter and scale width in which pnm had the highest value of 323.00 ± 58.37 and 630.55 ± 213.95μm respectively. the medullary diameter, index and fraction, and shaft diameter were highly significant (p<0.001) among all the species. post-hoc comparison showed that the medullary diameter, index, and fraction of hairs of the mm was significantly different (p≤0.001) from the other three nphs. the shaft diameter of the mm was also significantly different from that of the pnm (p≤0.001), and wtm (p≤0.004). the scale width of rcm differed from pnm (p≤0.01), and wtm (p≤0.05). these empirically established morphological differences in guard hairs of the four nhps in sfr would be useful in verifying their habitat occupancy and forensic evidence in case of illegal trafficking. keywords. dorsal hair, medulla hair pattern, mona monkey, non-invasive studies, qualitative hair features. 1 introduction the proper management and conservation of wild animals require population monitoring through regular surveys. such surveys provide a greater understanding of the distribution, extent, and status of populations, thereby facilitating the protection of threatened or rare species (rylands et al. 2008, plumptre et al. 2013). population surveys of terrestrial mammals have been conducted by direct or indirect techniques. https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v13i2.126 https://creativecommons.org/licenses/by-nc/4.0/ https://orcid.org/0000-0002-3362-8668 fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 206 direct techniques could be through live trapping or observation and counting of the target species. live trapping has been used for enumerating small mammals, but it requires adequate knowledge of the species, training, and in many instances a license (bertolino et al. 2009). even though live trapping could provide information on abundance, species, and sex of individuals captured, it could lead to capture of nontarget species, implication of stress to the trapped animals, and interferes with their activities (chiron et al. 2018). large mammals have been enumerated through observation and sighting of target species. population surveys of wild mammals especially arboreal species could be demanding in terms of the efforts in sighting, and the time to enumerate them correctly. nocturnal arboreals have been enumerated through the use of camera traps (bowler 2017). direct observation through line transects, and total/sweep count, have been used for the enumeration of non-human primates (nhps) in forest habitats (plumptre et al. 2013). direct counts of individuals in a troop or numbers of troops per distance covered, their dispersion and time used for such surveys are some of the parameters considered. in rainforests, such exercises could be quite difficult given the poor visibility of individuals, especially elusive species such as the putty-nosed monkey. indirect signs such as nest counts, scats/dung, and group calls have been used to survey fleeing or hideous species (plumptre 2000, plumptre et al. 2013). motion sensor cameras and hair tubes/lures have been used to study dietary composition, population genetics, and habitat occupancy of mammals (menike et al. 2012, cornally and lawton 2016, bowler 2017). these methods are non-invasive, and require no direct contact with the target animals. they are labour-efficient, less costly, reduce or have no interference with the animals’ activities, and do not expose researchers and the animals to zoonotic diseases (de bondi et al. 2010). some of the indirect methods could be limiting especially if the study species have uneven distribution or a low population density (ruell et al. 2009). the use of hair has proved to be quite reliable in non-invasive studies of the presence of mammals, especially those that have low population density in a location or cryptic ones (paez et al. 2021). it has been used to study the diet of carnivores and omnivores, and identify road kills, habitat occupancy and population genetics (menike et al. 2012). the presence and morphological analyses of chimpanzee hair found in the faeces of an adult chimpanzee in gombe national park, tanzania were used to confirm cannibalism/infanticide within this taxon (walker et al. 2017). silver hair found in night nests has been used to indicate the presence of adult male gorillas (gray et al. 2009). hair tubes/ lures of different designs have been used for hair sample collection from different terrestrial mammals (cornally and lawton 2016). the hair tubes/ lures must be located in different parts of the study location in order to obtain representative hair samples. hair samples are obtained when they get stuck to the strong adhesive glue applied to a rubber band stretched across the entrance to the tube (mcaney 2011). for arboreal mammals, hair samples could be obtained at nest sites (gray et al. 2009), water sites, or where they are known to feed and/or groom. hair samples could also be obtained from dead animals or museum specimens. fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 207 hair is strong and does not decompose fast, thereby making it easy to preserve. its durability and resilience are due to the presence of keratin, a sulphur-containing protein in its cortex. keratin makes hair to be less soluble and resistant to many chemical agents that could degrade it (velasco et al. 2009). the long chain nature of keratin makes the hair have a regular structure that is flexible (de sá dias 2004). this feature makes the hair retain its elasticity for a long time, thereby enhancing its durability. each mammalian species has a unique hair morphology that has been used to identify them. this is because the hairs vary in colour and have different morphological arrangements (farag et al. 2015). each mammalian species has a different hair configuration in the way the concentric layers of the cuticle, cortex and medulla are arranged. scale patterns and cross-sectional shapes are some of the basic information required in documenting the hair morphology of mammals (taru and backwell 2014). analysing these configurations can be used to identify a species (bhat et al. 2014). the micro-morphological characteristics of mammalian hair have been studied in forensic medicine (meyer et al. 2000), wildlife biology (sahajpal et al. 2009) and other disciplines. olaleru et al. (2020) provided a comparison of the mona monkey hair morphologies from omo, shasha, and oluwa forest reserves in southwestern nigeria. olugbebi et al. (2021) compared the morphological features of guard hair of some non-human primates in omo forest reserve and reported significant differences among the species. there is limited information on the morphological features of hairs of some members of the family cercopithecidae (cheech-pouched monkeys) that are found in shasha forest reserve (sfr). in this study, we compared the morphological characteristics of guard hair of four species of the cercopithecidae family: mona monkey (cercopithecus mona), putty-nosed monkey (cercopithecus nictitans), whitethroated monkey (cercopithecus erythrogaster), and red-capped mangabey (cercocebus torquatus) in sfr in southwestern nigeria. the findings would serve as a database in contrasting ecological studies, such as their habitat occupancy and predators. it could be useful in identifying members of this family in the event it is the only evidence found on poachers or wildlife traffickers. 2 material and methods 2.1 study area the shasha forest reserve is located in ife south local government area of osun state (figure 1). it lies between latitudes 6o 50’ and 7o 50’ n and longitudes 4o 15’ and 4o58’ e. it is part of the contiguous omo-oluwa-shasha forest reserves complex in southwestern nigeria (oludare and clement 2014). it is the smallest of the three reserves, covering about 310km2 (alo et al. 2020). it has a mean annual rainfall of about 2050 mm and mean monthly temperature of about 27˚c. shasha forest reserve is a lowland tropical rainforest. it has been reduced, due to anthropogenic activities, to fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 208 a secondary forest, thickets and farmlands of annual and perennial crops (adedeji and adeofun 2014, chenge and osho 2016). fig 1: map of shasha forest reserve in osun state, nigeria 2.2 hair sample collection and preservation guard hair strands from the dorsal regions of the studied non-human primates (nhps), namely, mona monkey (mm), putty-nosed monkey (pnm), white-throated monkey (wtm), and red capped mangabey (rcm) were collected from shasha forest reserve with the help of hunters and field guides. these were stored in labelled acid free paper envelopes and placed in zip lock bags that had silica gel, and kept at room temperature until they were used for the studies (garcia-alaniz et al. 2010). guard hairs were used because they produce pelage (coat) colour, are useful in differentiating species (tridico 2005, knecht 2012), and possess the features for which microscopic identification could be made (tridico 2015). 2.3 preparation of hair samples for morphological examination for each species, five strands of hair were randomly picked and prepared for viewing under the microscope as described by deedrick and koch (2004a). to remove dirt, the hair strands were immersed for five minutes in 70% ethanol, removed and air dried at room temperature (verma and joshi 2012). each hair strand for each species was placed on glycerin smeared slide, and then covered with a cover slip. these were used fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 209 for medullae pattern studies under light microscope that was attached to a digital camera for photomicrograph production at a magnification of x10. for scale cast, the same hair was cleaned in ethanol, removed and dried. this was used for scale impressions and cuticle pattern determination. clear nail polish was used to replace gelatin used by yasser et al. (2018) to produce cuticular scale pattern imprint. the polish was placed on another slide, allowed to set for two minutes, cleaned hair strand picked with fine-tipped forceps was placed on it and allowed to dry for five minutes. gentle removal of the hair leaves the scale pattern which was viewed under light microscope at a magnification of x40. 2.4 hair morphological evaluation and determination the qualitative parameters studied were medulla pattern and structure, cuticle scale distance, margin type, and distance. the quantitative characteristics studied were medulla diameter (μm), hair diameter (μm), medullary index, and fraction. hair and medullary diameters were measured at random points using a calibrated micrometer in the eyepiece. medullary index was calculated as medulla diameter/hair diameter. medullary fraction was calculated as (medulla diameter/hair diameter) x 100 (kitpipit and thanakiatkrai 2013). using the scale cast, the entire scale length, and width were determined in μm by using a calibrated micrometer in the eyepiece. these were determined from randomly selected cuticle scales (kitpipit and thanakiatkrai 2013). all morphological examinations were conducted in the instrument room, department of zoology, university of lagos. to ensure non mix-up of samples, the hair samples from each species were prepared and observed separately one after the other. cuticle scale patterns and medullae characteristics were determined by using available animal hair keys in literature as a guide. these included deedrick and koch (2004b), knecht (2012), cornally and lawton (2016), and yasser et al. (2018). 2.5 data analyses the data were analysed descriptively using microsoft excel and inferentially using spss (version 25). analysis of variance was used for the quantitative characteristics to determine differences in guard hair within and between species. a post-hoc test using least significant difference was used to separate means that were significant at p ≤ 0.05. 3 results 3.1 qualitative characteristics of selected monkey species in shasha forest reserve table 1 showed the qualitative characteristics of guard hairs from four non-human primate species in sfr. none of them had the same features, even though all had fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 210 ‘continuous’ medulla pattern. all species except mm had ‘uniserial’ medulla structure, and ‘distant’ scale margin. plate 1: photomicrographs of medulla and scale pattern of the guard hair of four studied nhps in shasha forest reserve (a and b = respective medulla and scale pattern of the hair of mona monkey; c and d = respective medulla and scale pattern of the hair of putty-nosed monkey; e and f = respective medulla and scale pattern of the hair of white-throated monkey; g and h = respective medulla and scale pattern of the hair of red-capped mangabey) a b c d e f g h fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 211 plate 1 shows the photomicrographs of guard hairs of the four monkey species studied. the scale patterns for mona and putty-nosed monkeys were coronal (a and c), while those of white throated monkey and red-capped mangabey were imbricate (e and g). the mm and rcm had ‘smooth’ scale margin type, while mm and pnm had ‘coronal’ scale pattern. apart from scale margin type which were ‘smooth’ and ‘crenate’ respectively in rcm and wtm all the other qualitative features were the same for these two species. the hairs of pnm and wtm differed only in the scale pattern which were ‘coronal’ and ‘imbricate’ respectively. table 1: qualitative characteristics of guard hair from mona, putty-nosed, red-capped mangabey and white-throated monkeys in shasha forest reserve. hair features medulla features scale features sample pattern structure margin distance margin type pattern mona monkey’s hair, continuous amorphous intermediate smooth coronal putty-nosed monkey’s hair continuous uniserial distant crenate coronal white throated monkey’s hair continuous uniserial distant crenate imbricate red-capped mangabey’s hair continuous uniserial distant smooth imbricate 3.2 quantitative morphological characteristics of guard hairs of four species of non-human primates in shasha forest reserve table 2 shows the mean and standard deviation in the quantitative characteristics of the guard hair of four non-human primates in sfr. the mona monkey hair had the highest recorded values for all morphological characteristics except shaft diameter and scale width in which putty-nosed monkey had the highest value of 323.00±58.37 and 630.55 ±213.95μm respectively. putty-nosed monkey was recorded to have the lowest value of medullary index which was 0.10±0.03μm. table 2: mean and standard deviation of morphological characteristics of hair from four nhps in shasha forest reserve. features sample n medullary diameter (μm) medullary index medullary fraction shaft diameter (μm) scale length (μm) scale width (μm) mmh 5 198.00±24.90 0.66±0.11 66.20±11.03 311.00±97.04 180.15±46.01 468.78±114.18 pnmh 5 32.00±8.37 0.10±0.03 10.20±3.03 323.00±58.37 175.00±19.95 630.55±213.95 wtmh 5 28.00±5.70 0.20±0.06 19.60±6.31 149.00±34.89 179.19±47.88 394.10±101.65 rcmh 5 32.00±6.71 0.18±0.04 17.80±4.09 183.00±26.36 175.49±44.78 448.15±54.02 mmh= mona monkey’s hair, pnmh= putty-nosed monkey’s hair, wtmh= white-throated monkey’s hair, rcmh= red-capped mangabey’s hair 3.3 differences in the morphological characteristics of guard hairs between the four studied non-human primate species in shasha forest reserve table 3 shows the anova of the quantitative features of guard hairs of four studied nhps in sfr. there were no significant differences (p≤0.05) between species for scale fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 212 length and scale width, while other characteristics such as medullary diameter, index and fraction, and shaft diameter were highly significant (p<0.001). table 3: analysis of variance in the morphological characteristics of guard hairs of the four non-human primate species in shasha forest reserve. morphological characteristics samples df mean square f significance medullary diameter between groups 3 35018.333 182.506 <0.001* within groups 16 191.875 total 19 shaft diameter between groups 3 39085.000 10.608 <0.001* within groups 16 3684.375 total 19 medullary index between groups 3 0.325 69.366 <0.001* within groups 16 0.005 total 19 medullary fraction between groups 3 3249.783 69.366 <0.001* within groups 16 46.850 total 19 scale length between groups 3 33.568 0.020 0.996 within groups 16 1703.124 total 19 scale width between groups 3 51777.954 2.874 0.069 within groups 16 18016.031 total 19 note: * = p ≤ 0.001 3.4 significant difference in the morphological characteristics between the guard hairs of the four studied non-human primates in shasha forest reserve table 4 showed the post hoc separation of means. the medullary diameter, index, and fraction of guard hairs of the mm were significantly different at p ≤ 0.001 from the guard hairs of the other three nphs. the shaft diameter of the mm was also significantly different from that of the pnm (p ≤ 0.001), and wtm (p ≤ 0.004). the shaft diameters of the wtm and pnm were both different from that in rcm. the pnm had medullary index and fraction that both differed from that of rcm. the scale width of rcm differed from that of pnm (p ≤ 0.01), and with that of wtm (p ≤ 0.05). fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 213 table 4: the post hoc on morphological hair characteristics between the four studied nonhuman primate species in shasha forest reserve. dependent variable (i) species (j) species mean difference (i-j) std. error sig. medullary diam. mm pnm 170.000 8.761 <0.001*** wtm 166.000 8.761 <0.001*** rcm 166.000 8.761 <0.001*** shaft diameter mm pnm 162.000 38.389 0.001*** wtm 128.000 38.389 0.004** wtm rcm -140.000 38.389 0.002** pnm rcm -174.000 38.389 <0.001*** medullary index mm pnm 0.466 0.043 <0.001*** wtm 0.484 0.043 <0.001*** rcm 0.560 0.043 <0.001*** pnm rcm 0.094 0.043 0.045* medullary fraction mm pnm 46.600 4.329 <0.001*** wtm 48.400 4.329 <0.001*** rcm 56.000 4.329 <0.001*** pnm rcm 9.400 4.329 0.045* scale width rcm pnm 236.448 84.891 0.013** wtm 182.392 84.891 0.047* mm = mona monkey, pnm = putty-nosed monkey, wtm = white-throated monkey, rcm = red-capped mangabey; *** = p ≤ 0.001, ** = p ≤ 0.01, *=p ≤ 0.05 4 discussion the objective of this study was to compare the morphological characteristics of the guard hair of the four members of the cercopithecidae family in shasha forest reserve. our findings showed that none of the nhps had the same qualitative hair features. these confirmed their unique specificity. they had some similarities in some features that could make it difficult not to mistake one for the other, if other features are not used. for instance, they all had the same medullary pattern that was ‘continuous’, a pattern that was reported brown howler monkey (alouatta guariba) hair from brazil (tremori et al. 2018), and in domesticated animals such as camel, cow, horse, sheep, cat and dog in egypt (yasser et al. 2018). using this feature alone is not advisable. all the quantitative parameters were different too, but with some being similar between two species. the medullary diameters of pnm and rcm in this study site were similar. these were also similar to that for blue nile monkey (cercopithecus mitis) reported farag et al. (2015). the medulla diameter of wtm though different from those of the rest in the same location, was similar to that obtained by tremori et al. (2018) for brown howler monkeys. the similarity in cuticular scale pattern in hairs of mm and pnm, and that of wtm and rcm could infer relatedness. cornally and lawton (2016) explained that many closely related species share similar cuticular scale patterns. the ‘distant’ scale margin fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 214 of hairs recorded for all the species apart from mm implies that the scales were less dense or relatively sparse. scale density is shown in how they are packed and could be described as ‘close’ when dense (cornally and lawton 2016). the similarities between species in some qualitative features at the microscopy level has limited inference (tridico 2005). similarities of species in qualitative hair features did not translate to similarities in quantitative characteristics. for instance, there was a unique similarity in the qualitative hair features between red-capped mangabey and white throated monkey, and between putty-nosed and white throated monkeys. quantitatively, the medullary and shaft diameters, and the medullary index and fraction of the species were significantly different. these differences in the morphology of the hair confirms the samples were from different species. the results obtained could be used at least at the preliminary stages of evidence of habitat occupancy, dietary component of predators, or exhibits on poachers. although cheaper when compared to standard dna profiling, hair morphology screening can be time consuming. with more detailed study techniques, especially at the molecular level, more differences could be established in the hairs of these species. molecular studies could perhaps show if these similarities could be due to genetic relatedness or adaptation to the environment. 5 conclusions the use of mammalian hair for micro-morphological studies is relatively cheap and easy to assess, the identity of these four species of nhps have been buttressed further by the empirical data provided through the micro-morphological study of their guard hairs. the differences in the hair features verified the presence of these species in sfr. with the present results as a database, hair samples collected at feeding, sleeping or watering sites could be analysed and the outcome compared to verify which species of nhps the hair belongs to. this result could also be useful database for comparing other poached members in the future. for ecological and conservation uses, environmental dna studies will be needed to map out the spatial distribution of these species in the reserve. acknowledgments we appreciate the assistance of the staff of shasha forest reserve for guidance during sample collection from anonymous hunters (that we are also grateful to) who permitted us to collect hair from their kills. we are grateful to mrs. oluwatumininu a. oluwarotimi of the department of cell biology and genetics, university of lagos for the technical assistance she rendered. we appreciate our anonymous reviewers for their constructive comments. references adedeji oh, adeofun co. 2014. spatial pattern of land cover change using remotely sensed imagery and gis: a case study of omo-shasha-oluwa forest reserve, sw nigeria (19862002). journal of geographic information system 6: 375-385. fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 215 alo aa, adetola aa, agbor cf. 2020. modelling forest cover dynamics in shasha forest reserve, osun state, nigeria. journal of agriculture and environment 16(1): 129-142. bertolino s, wauters l, pizzul a, molinari a, lurz p, tosi g. 2009. a general approach of using hair-tubes to monitor the european red squirrel: a method applicable at regional and national scales. mammalian biology 74: 210-219. bhat ma, awadh bs, aijaz ad, syed wb. 2014. studies on hair of some wild animals for species identification as an aid to wildlife forensics. internationaljournal for agro veterinary and medical sciences 8(1): 2-5. bowler mt, tobler mw, endress ba, gilmore mp, anderson mj. 2017. estimating mammalian species richness and occupancy in tropical forest canopies with arboreal camera traps. remote sensing in ecology and conservation 3(3): 146-157. chenge ib, osho jsa. 2016. matrix growth model for a natural forest in shasha forest reserve, nigeria. journal of natural science research 6(14): 99-106. chiron f, hein s, chargé r, julliard r, martin l, roguet a, jacob j. 2018. validation of hair tubes for small mammal population studies. journal of mammalogy 99(2): 478-485. cornally a, lawton c. 2016. a guide to the identification of irish mammal hair. irish wildlife manuals, no. 92. national parks and wildlife service, department of the arts, heritage, regional, rural and gaeltacht affairs, ireland. 1-40 pp. de bondi n, white jg, stevens m, cooke r. 2010. a comparison of the effectiveness of camera trapping and live trapping for sampling terrestrial small-mammal communities. wildlife research 37:456-465. doi.org/10.1071/wr10046. deedrick dw, koch sl. 2004a. microscopy of hair part i: a practical guide and manual for animal hairs. forensic science community 6(1): 1-45. deedrick dw, koch sl. 2004b. microscopy of hair part ii: a practical guide and manual for animal hairs. forensic science community 6(3): 1-32. de sá dias tc. 2004. analysis of the conditioning action of cosmetic substances added in ammonium thioglycolate-based hair straightener. masters dissertation. faculty of pharmaceutical sciences, university of são paulo, 1-106 pp. farag mf, ghoniem mh, abou-hadeed ah, dhama k. 2015. forensic identification of some wild animal hair using light and scanning electron microscopy. advances in animal and veterinary science 3(10): 559-568. garcía-alaníz n, naranjo ej, mallory ff. 2010. hair-snares: a noninvasive method for monitoring felid populations in the selva lacandona, mexico. tropical conservation science 3(4): 403411. gray m, mcneilage a, fawcett k, robbins mm, ssebide b, mbula d, uwingeli p. 2009. virunga volcano range mountain gorilla census, 2003. censusing the mountain gorillas in the virunga volcanoes: complete sweep method versus monitoring. african journal of ecology 48:588-599. kitpipit t, thanakiatkrai p. 2013. tiger hair morphology and its variations for wildlife forensic investigation. maejo international journal of science and technology 7(3): 433-443. knecht l. 2012. the use of hair morphology in the identification of mammals. in wildlife forensics: methods and applications, eds: huffman je, wallace jr. wiley and blackwell chichester, united kingdom. 129-143 pp. mcaney k. 2011. a pilot study to test the use of hair tubes to detect the irish stoar along hedgerows in county galway. vincent wildlife trust. 14 pp. menike u, manawadu d, meegaskumbura s. 2012. identification of sri lankan shrew using hair anatomy. ceylon journal of science (biological science) 41(1): 45-66. meyer w, seger h, hulmann g, neurand k. 2000. studies on the identification of mammals based on the hair cuticle pattern a comparison of domestic mammals and their wild forms from a forensic point of view. archives for criminology 200: 45-55. https://www.google.com/search?sxsrf=alekk03a0dlfeoo9gsphoydniqcpm1bbrw:1582753187405&q=untersuchungen+zur+artbestimmung+von+s%c3%a4ugetieren+anhand+des+haarkutikula+musters-ein+vergleich+von+haussaugetieren+und+ihren+wilden+stammformen+unter+forensischer+sicht.+archive+fur+kriminologie&spell=1&sa=x&ved=2ahukewjpgykml_dnahxq4iukhxarbbqqbsgaegqibhal https://www.google.com/search?sxsrf=alekk03a0dlfeoo9gsphoydniqcpm1bbrw:1582753187405&q=untersuchungen+zur+artbestimmung+von+s%c3%a4ugetieren+anhand+des+haarkutikula+musters-ein+vergleich+von+haussaugetieren+und+ihren+wilden+stammformen+unter+forensischer+sicht.+archive+fur+kriminologie&spell=1&sa=x&ved=2ahukewjpgykml_dnahxq4iukhxarbbqqbsgaegqibhal https://www.google.com/search?sxsrf=alekk03a0dlfeoo9gsphoydniqcpm1bbrw:1582753187405&q=untersuchungen+zur+artbestimmung+von+s%c3%a4ugetieren+anhand+des+haarkutikula+musters-ein+vergleich+von+haussaugetieren+und+ihren+wilden+stammformen+unter+forensischer+sicht.+archive+fur+kriminologie&spell=1&sa=x&ved=2ahukewjpgykml_dnahxq4iukhxarbbqqbsgaegqibhal fatsuma olaleru et al. guard hair micro-morphology of non-human primates in nigeria ruhuna journal of science vol 13 (2): 205-216, december 2022 216 olaleru f, olugbebi tr, fasona mi. 2020. morphological studies on the guard hair of the mona monkey (cercopithecus mona) in omo-shasha-oluwa forest reserves of southwest nigeria. egyptian academic journal of biological sciences b. zoology 12(2): 15-23. oludare ha, clement oa. 2014. spatial patterns of land cover change using remotely sensed imagery and gis: a case study of omo-shaha-oluwa forest reserve, southwestern nigeria (1986-2002). journal of geographic information system 6: 375-385. olugbebi tr, olaleru f, fasona mi, omoregie qo. 2021. guard hair morphology of cercopicidae family in omo forest reserve, ogun state, nigeria. egyptian academic journal of biological sciences b. zoology 13(2): 197-206. paez am, sundearam m, willoughby jr. 2021. comparison of minimally invasive monitoring methods and live trapping in mammals. genes 12(12): 1949. https://doi.org/10.3390/genes12121949. plumptre aj. 2000. monitoring mammal populations with line transect techniques in african forest. journal of applied ecology 37(2): 356-368. plumptre aj, sterling ej, buckland st. 2013. primate census and survey techniques. in primate ecology and conservation: a handbook of techniques, eds: sterling, ej, bynum n, blair mf. oxford university press, oxford. 10-26 pp. ruell ew, riley spd, douglas mr, pollinger pp, crooks kr 2009. estimating bobcat population sizes and densities in a fragmented urban landscape using noninvasive capture-recapture sampling. journal of mammalogy 90(1): 129-135. rylands ab, williamson ea, hoffmann m, mittermeier ra 2008. primate surveys and conservation assessments. oryx 42: 313-314. sahajpal v, goyal s, singh k, thakur v 2009. dealing wildlife offences in india: role of the hair as physical evidence. international journal of trichology 1(1): 18-26. taru p, backwell lr. 2014. hair morphology of some artiodactyls from southern africa. sabinet african journals 4(1): 26-32. tremori tm, garcia fmm, flórez lmm, gonçalves bp, de camargo bwdf, gwinnett c, teixeira cr, rocha ns. 2018. hair analysis of mammals of brazilian wildlife for forensic purposes. open journal of animal sciences 8: 335-345. doi.org/10.4236/ojas.2018.83025. tridico s. 2005. examination, analysis, and application of hair in forensic scienceanimal hair. forensic science review 17(1): 18-28. tridico sr. 2015. morphological and molecular approaches to characterise modifications relating to mammalian hairs in archaeological, paleontological and forensic contexts. phd thesis submitted to school of veterinary and life sciences, murdoch university. 176 pp. velasco mvr, de sá dias tc, de freitas az, vieira júnior nd, pinto cas de o, kaneko tm, baby ar 2009. hair fiber characteristics and methods to evaluate hair physical and mechanical properties. brazilian journal of pharmaceutical sciences 45(1): 154-162. verma k, joshi b. 2012. different animal species hairs as biological tool for the forensic assessment of individual identification characteristics from animals of zoological park, pragti maidan, new delhi, india. journal of forensic research 3(160): 2157-7145. walker cs, walker kk, paulo g, pusey ae. 2017. morphological identification of hair recovered from faeces for detection of cannibalism in eastern chimpanzees. folia primatologica 89: 240-250. doi: 10.1159/000488509. yasser aa, ali s, ghallab a. 2018. hair histology as a tool for forensic identification of some domestic species. excli journal 17: 663-670. javascript:; javascript:; javascript:; javascript:; microsoft word rjs-vol-ii-gunawickrama2-new-1.2.doc © 2007 faculty of science university of ruhuna ruhuna journal of science vol. 2, september 2007, pp. 70-81 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x 70 abstract. morphological variation of the euryhaline cichlid fish etroplus suratensis (bloch) from six geographically apart estuarine localities along the southern and western coasts of sri lanka was studied. significant heterogeneity in morphology of the cichlid were found with respect to nine morphometric characters (n=218). fish of nilwala estuary and garanduwa lagoon were not significantly different in morphology, yet they show discernible differences from the other four samples (kahanda lagoon, chilaw lagoon, walawe estuary and koggala lagoon) with respect to the most of the studied characters. concordant results were found by multivariate analysis of the sizecorrected morphological data as well. three functions were significant in discriminating the populations of which the first two functions accounted for 95% of the covariance (cv1 85.4 % and cv2 9.7%). the function with the greatest discriminatory power (cv1) can clearly separate samples of nilwala estuary (l5) and garanduwa lagoon (l6) from the rest of the samples, while the magnitude of the discrimination between the latter samples is much smaller. classification functions could correctly classify an average of 65.7% of the individuals into their respective a priori population units. no evidence was found for isolation-by-distance model. the results suggest that e. suratensis populations in some of the studied estuarine localities maintain significant morphological heterogeneity, and the morphological variation can be used to differentiate some of these populations. keywords: cichlid, morphometry, phenotypic variation, partial isolation. 1 introduction the green chromide, etroplus suratensis (bloch) is an indigenous cichlid fish restricted in distribution to sri lanka and india (ward & wyman, 1977). it is euryhaline, commonly occurring in riverine estuaries and coastal lagoons, and in natural and man-made freshwater habitats in sri lanka (pethiyagoda, 1991). some aspects of its biology including the length-weight relationship, diet and feeding, and aspects of reproductive biology have been studied (ward & wyman, 1977; costa, 1983; de silva et al., 1984). it is an important food fish in sri lanka, being caught mainly as a part of subsistence fishery (de silva, 1988). sri lankan fishermen have noted markedly decreasing catches of the green chromide in some estuarine habitats, and attribute this to a decline in population density (pers. comm.). as human populations tend to concentrate close to estuaries, anthropogenic threats to that environment are unavoidable. moreover, estuaries naturally receive large amounts of allochthonous material including domestic wastes morphological heterogeneity and population differentiation in the green chromid etroplus suratensis (pisces: cichlidae) in sri lanka k. b. suneetha gunawickrama department of zoology, faculty of science, university of ruhuna, matara, sri lanka correspondence: suneetha@zoo.ruh.ac.lk gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . 7 1 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) and other pollutants. thus, the estuarine ichthyofauna seems to be threatened by factors like habitat degradation and pollution. reportedly declining stocks of e. suratensis indicate that it is becoming increasingly important to study the existing levels of genetic and phenotypic variation, with regard to conservation planning. any management and conservation practice should be based upon prior knowledge of the existing natural variation, in order to preserve any adaptive variation that might be present. although estuaries are not closed hydrological systems like lakes, it is hypothesized that fish in different estuaries may have adapted to maintain their stocks largely within each estuary, resulting in some degree of isolation and an identifiable phenotypic differentiation. morphological characters for many fishes, including highly dispersible marine and estuarine species have revealed spatial separation of populations (schaefer, 1991; elliott et al., 1995; uiblein, 1995; hurlbut & clay, 1998; jerry & cairns, 1998). intra-specific phenotypic or genetic variation of the green chromide has not been studied previously, yet intra-specific variation has been indicated in the other extant species of the genus in sri lanka, e. maculatus (bloch) (pethiyagoda, 1991). the present work aims to test the hypothesis of intraspecific phenotypic variation for e. suratensis by studying the morphology of the fish among six geographically discrete estuarine localities in sri lanka. the null hypothesis of no population heterogeneity in morphology is statistically tested, and the possibility of differentiating the estuarine populations using morphological variation is assessed. 2 materials and methods 2.1 samples six estuarine localities including two riverine estuaries and four coastal lagoons in sri lanka were selected for sampling (fig. 1). adult fish were collected using various sized gillnets (mesh size range 4.0-6.0 cm) while a few samples were collected from fishermen. as the sex of individual fish was indistinguishable externally, a random sample of fish (n=104) from all six locations was sexdetermined internally, and preliminary testing was carried out to see whether there was any sexual dimorphism. no significant sexual dimorphism with respect to the selected morphometrics was observed, therefore the data analysis was performed without taking the sex of the individuals into consideration. 72 gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) figure. 1. map of the collection localities of etroplus suratensis. 2.2 morphometrics and meristic measures standard length (ls) was used as the measure of overall fish size. twelve reliably measurable morphometric characters were selected for the study (table 1). figure. 2. schematic diagram of the fish showing the positions of the superficial points used to collect data in etroplus suratensis (see table i for description). 20k gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . 7 3 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) all body measurements were taken from the lateral side of the fish to the nearest 0.1 mm using a vernier caliper, and were the distance between the verticals or horizontal lines made across the identified superficial landmark points (fig. 2). these landmark points were utilized to maintain the character homology between specimens as much as possible. all length measurements (l) were taken parallel to the antero-posterior body axis. head depth (hd) was taken perpendicular to the body axis between dorsal and ventral margins of the head, starting from the point where the ventral edge of the operculum intersects the ventral body margin. maximum body depth (mbd) was measured between the two visually detectable widest points of the trunk, perpendicular to the body axis. mouth gape (mg) was measured vertically between the mid points of upper and lower jaws of the fully-opened mouth. the length of the fin bases of dorsal and anal fin was measured between the verticals made across the externally visible origins of the first spine and the last ray of the fin. table 1. morphometric characteristics of etroplus suratensis studied. character code description number code (fig. 2) standard length ls mouth tip to the mid point of caudal fin origin 1-3 caudal peduncle length lcp length from the anal fin insert to the midpoint of the caudal peduncle 2-3 pre-dorsal length lprd mouth tip to the origin of dorsal fin 1-4 head length lh mouth tip to the posterior edge of operculum 1-5 orbital length lo length (along axis) of the orbit 6-7 post-orbital length lpo posterior edge of orbit to posterior edge of operculum 5-7 pre-orbital length lpro mouth tip to anterior edge of orbit 1-6 mouth gape mg height between the mid points of the upper and lower jaws of the fully opened mouth not shown head depth hd depth from the ventral point of intersection of the outer operculum edge to the dorsal head margin (measured vertically) 8-9 maximum body depth mbd distance between points at deepest part of body (measured vertically) 10-11 dorsal fin base fbdo length between the visible origins of the first spine and the last ray of the dorsal fin 4-12 anal fin base fban length between the visible origins of the first spine and the last ray of the anal fin 2-14 pectoral fin length lpecf length of the fin from dorso-posterior part of fin base to the edge of the fin 13-15 2.3 data and statistical analysis as the fish exhibiting allometric growth show a strong correlation in shape measurements to body size (reist, 1985, 1986), raw data of each morphometric 74 gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) character were standardized to standard length (ls) to obtain data sets with means of zero and unit standard deviation using the following equation (elliott et al., 1995). ms = mo (l / ls)b the working formula for each specimen for a given morphometric character is: log10ms= log10mo + b log10l– b log10 ls where ms is the standardized character measurement, mo is the observed character measurement, l is the mean standard length for all fish from six collections, ls is the standard length of the specimen, and b is the slope of the regression of log10 mo on log10ls for all fish. correlation analysis of the standardized data against size showed that the allometric transformation successfully removed size dependence. the standardized data were analyzed by univariate and multivariate methods. differences among geographic samples were tested by one-way analysis of variance (anova) followed by newman-keuls multiple comparison test (zar, 1984). tests were considered significant at 0.05 probability level with the sequential bonferroni adjustment (rice 1989). discriminant function analysis (dfa) was performed to identify the characters that were important in distinguishing population groups, and to formulate classification functions for each location. based on these functions, dfa classifies individual fish into a group, which is then compared with the a priori group of that individual to get the percentage of classification success. pairwise squared mahalanobis distance (d2) among samples were calculated and tested for their significance, and the agreement to the isolation-by-distance model (slatkin, 1993) was tested by pearson correlation analysis using pair-wise d2 and pair-wise geographic distance (approx. distance derived from a scaled map). all data analysis and statistical analyses were carried out by using the software package statistica v 7.0 (statsoft, usa). 3 results table 2. collection localities, sample sizes (n) and size statistics (standard length ls) of adult etroplus suratensis samples. location estuary type n size range (ls) (mm) mean ls (mm) sd l1: kahanda lagoon coastal lagoon 32 66.5128.4 90.6 17.6 l2: chilaw lagoon coastal lagoon 32 75.6116.3 93.5 9.2 l3: walawe estuary riverine estuary 31 77.2154.2 110.0 21.7 l4: koggala lagoon coastal lagoon 35 110.6143.6 125.2 8.5 l5: nilwala estuary riverine estuary 38 85.5174.0 126.5 28.4 l6: garanduwa lagoon coastal lagoon 49 52.6130.0 89.8 12.1 the size statistics of the fish samples is given in table 2. significant differences in size were found among samples where the largest fish were recorded from koggala gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . 7 5 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) -6 -4 -2 0 2 4 6 cv1 -5 -4 -3 -2 -1 0 1 2 3 4 5 c v 2 l1 l2 l3 l4 l5 l6 lagoon and nilwala estuary (p<0.05, anova). the mean ls (± sd) for all fish analyzed is 105.3 ± 23.8 mm (n=218). the sex ratio was male biased (1.2:1). univariate anova revealed highly significant differences (p<0.05, with bonferroni adjustment) in nine of the studied morphometric characters leading to the rejection of the null hypothesis of ‘no heterogeneity in fish morphology among estuarine populations’. there were no significant differences in caudal peduncle length, pre-orbital length, and length of dorsal fin-base among samples (table 3). fish from nilwala estuary and garanduwa lagoon share several of the morphometric characters that are significantly different from those in the other four locations. in this respect, they have shorter pre-dorsal lengths, shorter post-orbital lengths, smaller mouth gape, deeper heads, and shorter anal fin bases. the samples from nilwala estuary and garanduwa lagoon are not differentiable from each other by any of the morphometrics. fish from koggala lagoon have the largest maximum body depth. slight, yet significant differences in most of the other morphometric characteristics were found among the studied populations (table 3). figure. 3. plot of the first and second canonical variate (cv) scores for the six collections of etroplus suratensis. cv1 and cv2 explain 85.4% and 9.7% of the variation among the individuals respectively. (○ l1: kahanda lagoon; □ l2: chilaw lagoon, ◊ l3: walawe estuary, + l4: koggala lagoon, • l5: nilwala estuary, ∆ l6: garanduwa lagoon). discriminant function analysis (dfa) for size-standardized data identified eleven characters that contribute significantly for the derived functions (wilks' lambda= 0.106; n= 218; approximate f(60, 921)= 9.43; p<0.001) (table 4). head depth of the fish is the character that contributed most to the discrimination among groups while length of the pectoral fin had the least contribution (p=0.103). among the five functions (canonical roots or covariates) derived, only three were statistically significant (chi square tests, p<0.05) (table 5). according to the standardized coefficients for roots (table 6), first function is weighted most heavily by mg, hd, lo, lpro, lpo and fban while the second function is weighted mostly 76 gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) by mbd, lcp, fbdo and lpecf. about 85.5% of all discriminatory power is explained by the first function, and therefore it is clearly the most important discriminant function. as detected by the respective eigen values, the first two canonical variables (cv) in the dfa collectively account for 95.1% of the variance in the data (table 6). according to the canonical means, first function can clearly discriminate nilwala (l5) and garanduwa (l6) populations from the rest of the groups (by having larger negative means), and chilaw (l2) & koggala (l4) samples from kahanda (l1) & walawe (l3) samples to a lesser degree (table 7). accordingly, the plot of the first two cv's separates the data into two discernible clusters (fig. 3) in which the cv1 has the most important power in discriminating l5 and l6 from the rest. although the magnitude of the discrimination is much smaller, cv2 seems to discriminate l2 from l4 in the latter cluster based on positive and negative canonical means respectively (table 7). the studied characters are not able to discriminate l5 from l6, or l1 from l3 adequately. the derived classification functions could correctly classify an average of 65.7% of the individuals into their respective a priori groups (table 8). the pair-wise squared mahalanobis distance (d2) for all pairs of populations were highly significant (table 9) yet there was no agreement to the isolation-by-distance model (pearson’s r = -0.05, p>0.05). table 3. summary of the morphometrics of etroplus suratensis presented as percentages of ls (mean ± sd). superscripts indicate test results of the anova followed by newman-keuls multiple comparison tests on size-adjusted characters (any measurements with shared superscript letters are not significantly different from each other at p<0.05 with sequential bonferroni adjustment). character length (mm) kahanda lagoon (l1) n= 32 chilaw lagoon (l2) n= 32 walawe estuary (l3) n= 31 koggala lagoon (l4) n= 35 nilwala estuary (l5) n= 40 garanduwa lagoon (l6) n= 50 lcp* 10.2± 1.4 a 11.4± 1.4 a 10.3± 1.4 a 10.2± 1.4 a 10.8± 1.7 a 10.3± 2.1 a lprd 33.6± 1.5 a 33.8± 2.2 a 33.2± 1.5 a 34.1± 1.7 a 30.5± 2.3 b 31.3± 2.5 b lh 30.0± 1.0 ab 29.9± 1.4 ab 30.6± 1.5 a 30.0± 1.4 a 28.0± 2.4 b 30.2± 2.4 ab lo 8.5± 0.8 ad 8.1± 0.6 bd 8.2± 0.9 acd 7.2± 0.6 b 8.2± 0.6 ac 9.2± 1.0 c lpo 12.6± 1.1 a 12.7± 0.9 a 12.9± 1.2 a 13.0± 0.8 a 10.7± 1.1 b 11.4± 1.4 b lpro * 9.2± 1.1 a 9.2± 1.3 a 9.5± 1.4 a 9.6± 1.0 a 9.9± 1.3 a 10.0± 1.4 a mg 10.8± 0.9 a 11.3± 0.9 ac 11.1± 0.8 ac 11.6± 0.8 c 9.6± 0.9 b 9.8± 1.1 b hd 45.3± 2.0 a 41.2± 3.1 c 44.5± 2.8 a 46.2± 2.3 a 48.8± 2.0 b 49.0± 5.0 b mbd 57.1± 3.4 a 55.9± 2.8 a 56.8± 2.3 a 58.8± 1.8 b 55.9± 2.1 a 56.6± 2.9 a fbdo* 59.9± 3.6 a 60.4± 3.2 a 59.7± 2.2 a 60.0± 1.9 a 58.7± 2.7 a 58.7± 3.3 a fban 43.1± 3.0 ac 43.8± 2.0 ac 42.5± 3.2 a 44.8± 2.0 c 39.7± 2.9 b 39.9± 3.3 b lpecf 26.7± 2.8 ab 25.5± 2.8 a 27.2± 2.0 b 27.0± 1.6 b 25.6± 2.0 ab 26.7± 1.8 ab * not significant gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . 7 7 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) table 4. summary results of the forward stepwise discriminant function analysis of standardized morphological data of etroplus suratensis. the characters are listed in the descending order of contribution to the discrimination between locations (wilk’s partial lambda). character wilk’s lambda partial wilk’s lambda f-to remove p-level hd 0.148 0.716 15.564 <0.001 mg 0.142 0.748 13.214 <0.001 lo 0.130 0.817 8.780 <0.001 lprd 0.119 0.896 4.563 0.001 fbdo 0.117 0.910 3.899 0.002 fban 0.116 0.913 3.750 0.003 lh 0.116 0.914 3.686 0.003 lpro 0.116 0.917 3.525 0.004 mbd 0.116 0.919 3.433 0.005 lcp 0.114 0.934 2.748 0.020 lpo 0.113 0.937 2.629 0.025 lpecf* 0.111 0.955 1.863 0.103* * not significant table 5. results of the chi square analysis to determine which roots are statistically significant (a stepdown table for all roots; asterisks indicate significant tests with p<0.05). roots remove d eigen value canonical r wilk’s lambda χ2 df p-level 0 4.137 0.897 0.106 455.0 60 <0.001* 1 0.469 0.565 0.546 122.8 44 <0.001* 2 0.173 0.384 0.802 44. 8 30 0.040* 3 0.045 0.207 0.941 12.1 18 0.825 4 0.018 0.132 0.983 3. 6 8 0.894 table 6. standardized coefficients for the three significant discriminant functions (roots) derived for morphology of etroplus suratensis. character root 1 root 2 root 3 mg 0.575 -0.132 -0.068 hd -0.544 -0.504 0.370 lpo 0.299 -0.131 -0.172 lo -0.480 0.303 -0.276 fban 0.327 0.104 0.183 lcp -0.038 0.420 0.047 lpecf -0.070 -0.326 -0.089 lh 0.180 -0.322 -1.008 lprd 0.345 0.111 0.580 lpro -0.426 0.049 0.212 fbdo 0.308 0.392 0.086 mbd -0.155 -0.499 0.279 eigen value 4.137 0.469 0.173 cumulative proportion 0.855 0.951 0.987 78 gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) table 7. means of canonical variables (roots) for each locality based on the discriminant functions derived for morphology of etroplus suratensis. location/ group root 1 root 2 root 3 root 4 root 5 l1 -0.681 0.042 0.259 -0.450 -0.090 l2 2.246 1.278 -0.018 0.178 -0.072 l3 1.098 -0.203 -0.757 -0.090 0.180 l4 2.171 -1.035 0.373 0.178 -0.0002 l5 -2.416 0.365 0.448 0.035 0.183 l6 -2.384 -0.240 -0.280 0.088 -0.147 table 8. proportion of e. suratensis correctly classified into groups by classification functions based on the a priori knowledge of their original location. group numbers refer to the respective location number (rows: observed classification, columns: predicted classification). groups group percent correct l1 l2 l3 l4 l5 l6 1: l1 59.4 19 3 3 5 1 1 2: l2 74.2 0 23 2 6 0 0 3: l3 54.8 4 4 17 5 0 1 4: l4 82.9 2 2 2 29 0 0 5: l5 58.3 0 0 0 0 21 15 6: l6 64.6 1 1 1 0 14 31 total 65.7 26 33 25 45 36 48 table 9. squared mahalanobis distance between pairs of populations (below diagonal) and the probability values (above diagonal) (asterisks indicate tests that are not significant at p<0.05 with sequential bonferroni adjustment). l1 l2 l3 l4 l5 l6 l1 <0.0001 0.046 * <0.0001 <0.0001 <0.0001 l2 4.45 <0.0001 <0.0001 <0.0001 <0.0001 l3 1.47 4.19 <0.0001 <0.0001 <0.0001 l4 3.80 5.51 3.23 <0.0001 <0.0001 l5 10.04 22.87 14.14 23.06 0.084 * l6 10.05 23.82 12.49 21.83 1.01 4 discussion results from both univariate and multivariate analyses provide congruent evidence for the existence of significant intraspecific morphological heterogeneity in e. suratensis among selected estuarine localities of sri lanka. derived classification functions provide moderate percentage of classification success (ranging from 54.8% to 82.9%), indicating that the morphological characters used in this analysis gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . 7 9 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) provide some discriminatory power for the e. suratensis populations studied. few population-specific characteristics are also evident including the largest body depth (mbd) of fish in koggala lagoon and the shortest head depth (hd) of fish in chilaw lagoon. among three significant functions in discriminating the populations, the first function (cv1) is clearly the most important one, and the magnitude of discrimination of the other two functions is comparatively much smaller. the close affinity of the fish populations between garanduwa lagoon and nilwala estuary is evident. of the characters identified as being most weighted for cv1, nilwala estuary and garanduwa lagoon shared four characters that are also significantly different from those of the other four samples (smaller mg, larger hd, and shorter lpo and fban). in addition, lprd which is weighted most heavily for the cv3 is also a shared character between nilwala estuary and garanduwa lagoon yet differs from all samples of the second cluster. overall, the close affinity between l5 and l6 suggests that the two samples represent similar geographical existence and origin, and the subsequent population intermixing is the likely explanation for the observed morphological homogeneity. in fact, the garanduwa lagoon receives freshwater influxes originating from ‘polwatta ganga’ that has a hydrological connection to the river nilwala before discharging to the sea at modara in weligama (annon. 1988). such hydrographical connectivity between the two water bodies is likely to facilitate the intermixing of pelagic fish populations, disrupting any potential mechanisms of population divergence. wide overlap exists in the cv values among the samples of the second cluster as well, although discrimination to a lesser degree is possible. the morphological heterogeneity observed has no significant correlation to the geographical distance between estuaries, as evidenced by the negative results from the isolation-bydistance model. chilaw lagoon represents the geographically most remote location among the studied estuarine populations, yet the pair-wise mahalanobis distances involving the chilaw lagoon sample are not systematically large. this is also displayed by the clustering of the chilaw population together with three other geographically remote populations in the second cluster. morphometric measurements have been widely used to discriminate populations of various fish species (elliott, et al., 1995; uiblein, 1995; hurlbut & clay, 1998). the conventional approach for such analysis is based on measurements along the antero-posterior body axis and the depth measurements. however, no previous investigation on morphology of e. suratensis populations has been reported so far. in this study, variation in various morphological characters of the fish is found between geographically separated estuarine localities. the reasons behind such variability might include geographic isolation, phenotypic plasticity and local adaptation. the waters off the south coast of sri lanka contribute to the overall current regime around the island having a direction closely connected with the indian ocean monsoons (schott et al., 1994, shankar et al., 2002). this coastal current regime is likely to impose tidal currents in and out of coastal lagoons and riverine estuaries, allowing some interchange between drift-vulnerable life stages of fishes. however, e. suratensis is a substrate spawner that has a specialized life 80 gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) history strategy to protect the young during their early stages (ward & wyman, 1977). such life strategy is likely to contribute to the retention of early stages that may be vulnerable to tidal flow, largely within the estuarine habitat, and consequently facilitating their populations to maintain to some extent separately from other neighboring populations. genetic polymorphism or environmental factors may induce morphological variability among spatially separated fish populations (carvalho, 1993), and phenotypic plasticity in fish morphology has been documented for various species, including cichlids (wimberger, 1991, 1992). fishes are considered to be phenotypically more variable than most other vertebrates, having relatively higher within-population coefficients of variation of phenotypic characters (carvalho, 1993). the differentiable variation in morphology among fish populations has been suggested as indicative of the presence of stock structuring and restricted movement among geographically isolated populations (uiblein, 1995; roby et al., 1991; palumbi, 1994; jerry & cairns, 1998). in e. suratensis, presence of such population subdivision or strong stock structure is not supported due to the low degree of differentiation detected. morphological divergence has been reported in estuarine fish populations that are not completely geographically separated, suggesting that partial isolation may play a role in population subdivision (roby et al., 1991, suneetha & naevdal, 2001). the results of the present study can also be explained by a similar postulate of partial isolation as the studied estuarine habitats are not completely isolated. population differentiation may also occur despite opportunity for extensive gene flow between populations when there are relatively strong differential selective pressures exerted on the different populations by local environmental factors, such as temperature (verspoor & jordan, 1989). as the present analysis does not include environmental data for the sample localities, it is not possible to confirm whether the observed variation is associated with local environmental conditions, and therefore, further environmental comparisons of these estuaries would be worthwhile. in addition, genetic investigations of the variation and population differentiation involving more estuarine samples of e. suratensis will be useful in substantiating the findings of the present study. knowledge of the genotypic and phenotypic variation of species is a pre-requisite in conserving them, particularly where populations with subtle adaptive variation are likely to exist. acknowledgements this study was funded by the national science foundation (sri lanka) grant number rg/2003/bt/03. the technical assistance provided by s. kapukotuwa and a.g. samanthi, and assistance in preparation of fig.2 by anuradha de silva is acknowledged. references anonymous (1988). national atlas of sri lanka, survey department, colombo, sri lanka. gunawickrama: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . 8 1 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 70-81 ( 2 0 0 7 ) carvalho, g. r. (1993). evolutionary aspects of fish distribution: genetic variability and adaptation. journal of fish biology 43 (supplement a), 53-73. costa, h. h. (1983). biological studies of the pearl spot etroplus suratensis (pisces, cichlidae) from three different habitats in sri lanka. internationale revue der gesamten hydrobiologie 68 (4), 565-580. de silva, s. s. (1988). reservoirs of sri lanka and their fisheries. fao fisheries technical paper 298, 1-128. de silva, s. s., maitipe, p. & cumaranatunge, r. t. (1984). aspects of the biology of the euryhaline asian cichlid, etroplus suratensis. environmental biology of fishes 10, 77-87. elliott, n. g., haskard, k. & koslow j. a. (1995). morphometric analysis of orange roughy (haplostethus atlanticus) off the continental slope of southern australia. journal of fish biology 46, 202-220. hurlbut, t. & clay, d. (1998). morphometric and meristic differences between shallow and deepwater populations of whitehake (urophycis tenuis) in the southern gulf of st. lawrence. canadian journal of fisheries and aquatic sciences 55, 2274-2282. jerry, d. r. & cairns, s. c. (1998). morphological variation in the catadromous australian bass, from seven geographically distinct riverine drainages. journal of fish biology 52, 829-843. palumbi, s. r. (1994). genetic divergence, reproductive isolation and marine speciation. annual review of ecology and systematics 25, 547-572. pethiyagoda, r. (1991). freshwater fishes of sri lanka. colombo: wildlife heritage trust of sri lanka. reist, j. d. (1985). an empirical evaluation of several univariate methods that adjust for size variation in morphometric data. canadian journal of zoology. 63, 14291439. reist, j. d. (1986). an empirical evaluation of coefficients used in residual and allometric adjustment of size covariation. canadian journal of zoology 64, 1363-1368. rice, w. r. (1989). analysing tables of statistical tests. evolution 43, 223-225. roby, d., lambert, j. d. & sevigny j. m. (1991). morphometric and electrophoretic approaches to discrimination of capelin (mallotus villosus) populations in the estuary and gulf of st. lawrence. canadian journal of fisheries and aquatic sciences 48, 2040-2050. schaefer, k. m. (1991). geographic variation in morphometric characters and gill raker counts of yellowfin tuna thunnus albacares from the pacific ocean. fishery bulletin u. s. 89, 289-297. schott, f., reppin, j., fischer, j. & quadfasel, d. (1994). currents and transports in the monsoon current south of sri lanka. journal of geophysical research 99 (c12), 25127-25142. shankar, d., vinayachandran, p. n. & unnikrishnan, a. s. (2002). the monsoon currents in the north indian ocean. progress in oceanography 52, 63-120. slatkin, m. (1993). isolation by distance in equilibrium and non-equilibrium populations. evolution 47, 264-279. suneetha, k. b. & naevdal, g. (2001). genetic and morphological stock structure of the pearlside, maurolicus muelleri (pisces, sternoptychidae) among norwegian fjords and offshore area. sarsia 86, 191-201. uiblein, f. (1995). morphological variability between populations of neobythites stefanovi (pisces: ophidiidae) from deep red sea and gulf of aden. marine ecology progress series 124, 23-29. verspoor, e. & jordan, w. c. (1989). genetic variation at the me-2 locus in the atlantic salmon within and between rivers: evidence for its selective maintenance. journal of fish biology 35 (supplement a), 205-213. ward, j. a. & wyman, r. l. (1977). ethology and ecology of cichlid fishes of the genus etroplus in sri lanka: preliminary findings. environmental biology of fishes 2(2), 137-145. wimberger, p. h. (1991). plasticity of jaw and skull morphology in the neotropical cichlids geophagus brasiliensis and g. steindachneri. evolution 45, 1545-1561. wimberger, p. h. (1992). plasticity of fish body shape. the effects of diet, development, family and age in two species of geophagus (pisces: cichlidae). biological journal of the linnaean society 45, 197-218. zar, j. h. (1984). biostatistical analysis (second edition). new jersey: prentice-hall. rjs-vol-1-sept-2006-1.dvi ruhuna journal of science vol. 1, september 2006, pp. 1–15 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. a study of salt secretion by mangroves of rekawa lagoon, sri lanka lp. jayatissa, wat. weerakkody, np. dissanayake department of botany, university of ruhuna, matara, sri lanka lpj@bot.ruh.ac.lk, lankadimuthu@yahoo.com senanayake g. department of agricultural biology, faculty of agriculture, university of ruhuna, mapalana, kamburupitiya, sri lanka sanjeewani s. department of molecular biology, box 590, swedish university of agricultural sciences, se-751 24, uppsala, sweden. a method to measure the salt secretion by mangroves which are open to the sea spray was developed and used to measure the salt secretion by mangrove species of rekawa lagoon, an ecosystem with the highest diversity of true mangroves in southern sri lanka. out of the twelve species of mangroves, only four species i.e. acanthus ilicifolius, aegiceras corniculatum, avicennia marina, and avicennia officinalis proved to be able to secrete salts. under the salinity regime (26.2 ± 4.41 ppt) existed during the experimental period, the salt secretion by leaves of these four species were 47.2 ± 18.3, 35.1 ± 16.0, 149.3 ± 45.9, and 81.6 ± 30.5 mg salt cm−2 day−1 respectively. this result corroborates the published records and, hence validates the technique used in this study to measure the salt secretion. these four species exhibited increases in salt secretion with increases in soil salinity, consistent with previous reports. results of this study also shows that the capacity to secrete salts at any given salinity was different between four species, following an order of av. marina > av. officinalis > ac. ilicifolius, and ae. corniculatum. the salt secretion by these species immediately after reducing the soil salinity was increased significantly implying an opportunistic removal of salt, which was accumulated in the plant body under high saline condition. capacity for salt secretion by the four species as well as the magnitude of the increase of salt secretion as a response to increasing soil salinity, vary in parallel to the variations in salt tolerance given in published reports for the same species. key words : mangroves, salt secretion, salt tolerance, acanthus, aegeceras, avecennia. 1. introduction the importance of the salt tolerant plants are increasing over the sea level rise due to global warming as lands unsuitable for normal terrestrial plants are increasing day by day and only salt tolerant plants can make them productive (dahdouhguabas et al, 2005). mangroves which are woody plants that grow in intertidal areas of lagoons, estuaries and sheltered bays in tropical and sub-tropical latitudes, are one of the major group of salt tolerant plants. although the salinity is often considered as a “stress”, nacl may also be a resource for halophytic species (ball 1 jayatissa et al.: a study of salt secretion... 2 ruhuna journal of science 1, pp. 1–15, (2006) 2002). the classic growth response of halophytes including mangroves to increasing salt concentration is similar to that of nutrients, with variation in the shape of the response curve reflecting concentrations which are deficient, saturating and toxic to growth (ball, 2002). however the level that becomes toxic is depends on the species. although the toxic level or tolerance limit is comparatively higher for all the mangroves, the degree of tolerance varies among different mangrove species (allen et al., 2003; tomlinson, 1986) the order of salinity tolerance of mangrove species are thought as one of the reason for zonation pattern of mangrove communities in the intertidal zone (kathiresan and bingham,2001). the physiological tolerance of mangroves for high salt levels depends on the mechanisms they have to extract fresh water from salty water in the soil (ball, 1996). mangroves avoid heavy salt loads by one or more of the following mechanisms; (a) salt exclusion most of the mangroves avoid salts using ultrafilters in root systems, which exclude salts while extracting water from the soil. they keep salt ions as filtered out during the water absorption through this ultrafiltration mechanism, eg. rhizophora, bruguiera, and ceriops. (b) salt secretion some mangroves take salty water up, and then secrete only the salts through specialized glands in the leaves called ‘salt glands’. eg. avicennia, aegiceras. (dschida et al., 1992; fitzgerald et al., 1992). (c) salt accumulation it is reported that the salt concentrations in the sap of mangrove plants can also be reduced by transferring the salts into senescent leaves and/or to the bark of stem and aerial roots or their wood (tomlinson, 1986). the transferred salts, particularly sodium and chloride ions, deposit and store in such parts. naskar and mandal (1999) reports that excoecarea and lumnitzera accumulate sodium and chloride ions in senescent leaves, but withdrawn potassium and phosphate ions prior to leaf senescence. mangroves avoid heavy salt loads through a combination of above processes (kathiresan and bingham, 2001). however, all the mangrove species do not possess these mechanisms in the same order. particularly the salt secretion takes place only in four, out of 20 true mangrove genera in the world, i.e. acanthus, aegialitis, aegicerous and avicennia. the salt ‘glands’ that secrete sodium chloride to control the salt balance in the plant body of these genera could be the most distinctive structure developed in mangrove leaves. salt glands occur on the surface of both side of leaves in these genera, but are not necessarily equally frequent on upper and lower surfaces (tomlinson, 1986). in most of the cases, it is more abundant in adaxial surface (hong and eong, 1984). the precise mechanism of salt secretion is not understood, but it is an active process, as evidenced by atpase activity in the plasmalemma of the secretary cells and the possibility of stopping the salt secretion by metabolic inhibitors (drennan et al., 1992, tomlinson, 1986). the salt secreting mangrove species allow more salt into the xylem than do the non-secreting species. therefore, the nacl concentration of salt secreting mangroves is about 10 times higher than that of non-secreting mangroves (tomlinson, 1986). however, they still exclude about 90% of the salts (scholander et al., 1962). although the salinity tolerances of some mangrove species have been studied, much information on salt jayatissa et al.: a study of salt secretion... ruhuna journal of science 1, pp. 1–15, (2006) 3 tolerance remains to be discovered. particularly, the studies on the salt secretion by mangroves are scarce. still the precise mechanism of salt secretion is not known. even the little knowledge on the salt balance of salt secreting mangroves may not be universal as the salt balance or water use characteristics of the same species are reported to vary depending on climatic and edaphic factors (youssef and saenger, 1988). therefore it is important to conduct studies to enrich the knowledge on salt secretion by mangroves. it is vital as the importance of salt tolerant plants is increasing day by day due to continuing climatic changes. 1.1. objectives this is a preliminary study on the salt secretion by mangroves in sri lanka. the study aims firstly to measure the salt secretion by mangroves in situ, and compare the rate of salt secretion by different species, which are in the same community under the same environmental condition. secondly, this project aims also to study the variation of salt secretion by mangroves with the variation of soil salinity under green house condition. effects of the previous exposure to higher or lower salinity on the salt secretion by mangroves under subsequent salinity regime were also studied as it gives an implication on effects of fluctuating salinity levels in the natural environment on the salt secretion by mangroves. 2. materials and methods 2.1. in situ measurement of salt secretion by mangroves the salt secreting mangrove species are listed in the literature. but, as a preliminary step, it was decided to test all the true mangrove species, which are co-existing in a same community, for salt secretion. if the expected results are obtained, then the methodology used to measure the salt secretion in this study gets validated. moreover, the capacity of salt secretion by different species can be copared as they were measured under the same environmental conditions. all the true mangroves in rekawa lagoon, in southern sri lanka, were selected for this study as it is located nearby and having highest species diversity of true mangroves in the southern sri lanka. moreover almost all the species which are reported as common mangroves in sri lanka, occur in rekawa lagoon also. followings are the mangrove species from rekawa lagoon, selected for this study (1) acanthus ilicifolius* (7) ceriops tagal (2) aegiceras corniculatum (8) excoecaria agallocha (3) avicennia marina (9) heritiera littoralis (4) avicennia officinalis (10) lumnitzera racemosa (5) bruguiera gymnorriza (11) rhizophora mucronata (6) bruguiera sexangula (12) sonneratia caseolaris (* according to tomlinson, 1986, this species is considered as a mangrove associate). in order to distinguish salt secreting mangroves and compare the level of secretion between species, the salt secretion by these mangroves were quantified in situ. quantification of the salt secretion in the field was rather difficult without a proper control, because salt brought by the sea spray could also be deposited on jayatissa et al.: a study of salt secretion... 4 ruhuna journal of science 1, pp. 1–15, (2006) mangrove leaves. it is unavoidable, as all the mangroves are located near to the sea. covering the plants or plant parts from the sea spray during the experiment, cannot be recommended as it may affect the normal physiological processes of plants and hence the salt secretion too. therefore, a proper control for the in situ measurement of the salt secretion was planned and included as follows to the methodology to measure the salt secretion in situ. three individuals from each species were selected for the test and two closely located twigs, each having at least 10 leaves, were selected from each of the three individuals. one twig from each individual was cut in water and left the cut end in double distilled water for about half an hour to continue transpiration replacing the salty xylem sap by distilled water. then the leaves of the cut twig were washed thoroughly by distilled water and the twig was tethered to the same place of the mother plant with the same orientation. leaves of the other twig were also washed thoroughly by double distilled water and blotted dry. both twigs from each individual were left for 24hours and then, 10 leaves from each twig was carefully removed and washed in 50ml of double distilled water to get all the salts on the surface of leaves into the water. then, area of all the leaves used to collect secreted salts was quantified manually using millimeter papers on which the exact size and shape of leaves are marked. the salinities of two 50ml solutions, one containing salts from leaves of the detached twig and the other from the intact twig, were quantified separately using a salinometer (hanna, conductivity/ salinity meter, modle hi 9835). the amount of salt collected from leaves of each twig was standardized as ‘mg salt cm−2 day−1’. the ‘detached twig’ in this experiment was considered as the ‘control’. if there is a value for the amount of salts corresponds to detached twig, it should be purely due to salts deposited on leaves by the sea spray and it should be common for the intact twig also. therefore the salt secretion by leaves of each individual was corrected as follows; salt secretion = (‘mg salt cm−2 day−1’ corresponds to the intact twig) (‘mg salt cm −2 day−1’ corresponds to the detached twig) the above whole procedure was repeated three times during a six month period, to measure the salt secretion by same mangrove species (but using different individuals), when the salinity level of the lagoon water is different. although the salinity level of lagoon water is not always same to that of soils in the intertidal zone, it is true that the soil salinity of the intertidal zone vary depending on the salinity of lagoon water. therefore, on each occasion of measurement of salt secretion by mangrove plants in rekawa lagoon, the salinity of lagoon water was also measured by a hand refratometer (atago s/mill-e, japan) 2.2. measurement of salt secretion under different soil salinities salt secreting mangrove species which were identified based on the results of in situ studies described in 2.1, were used for the ex situ measurement of salt secretion under different levels of the soil salinity. young individuals of each of salt secreting mangrove species with at least few twigs in each, were brought into the university green house and planted individually in jayatissa et al.: a study of salt secretion... ruhuna journal of science 1, pp. 1–15, (2006) 5 plastic pots (with 20cm diameter and 30cm height) having few holes at the bottom, and filled with the soil brought from the same mangrove. pots with young plants were placed individually on 7cm deep plastic trays to get extra water collected after watering. water with salinities of 5, 15, 25, and 35 ppt were prepared separately by mixing seawater and tap water, and left in separate tanks to be used in the experiment. initially all the pots in each species were irrigated once a day by the water with the salinity of 10 ppt. excess water accumulated in trays were returned to the relevant tank every other day and salinity of the water in tanks were measured by a hand refractometer (atago s/mill-e, japan) and adjusted by adding fresh water or sea water more, as necessary. ten pots with individual plants which were established and rooted up to the bottom of the pot, from each species were used in the experiment to measure the salt secretion under different salinity levels. pots were assigned to five salinity levels, 0, 5, 15, 25, and 35 ppt, as two pots per one salinity level. each plant was irrigated with excess water with the salinity assigned to it, twice a day and left for three days to get acclimatized to the particular salinity. meanwhile, at least 10 mature leaves from each individual plant were selected and their area was quantified manually without detaching leaves, by marking the exact shape and size of each leaf on millimeter papers. on the fourth day, the salt secretion during a 24 hour period, by leaves of which the area is known, were measured following the same method used for in situ measurements. (but a detached twig from each individual was not used in this ex situ experiment as it was verified that there was no sea-spray into the green house). during that period plants were irrigated three times with excess water having the same salinity assigned to each plant while keeping them in trays with excess water with the same salinity. the procedure was repeated four times, but each time, the assigning of plants of the same species into five salinity levels was done by re-randomization. before each measurement, plants were left three days in trays with water having the same salinity assigned to the plant and irrigated twice a day with the same water. the salt secretion was standardized as ‘mg salt cm−2 day−1’. the experiment was ended within 16 consecutive days in june 2006. 2.3. measurement of salt secretion under fluctuating soil salinities same mangrove species, which were identified as salt secreting mangroves, were used to measure the salt secretion under fluctuating salinity levels. same individual plants, which were used to measure the salt secretion under different salinities, were also used to measure the salt secretion under fluctuating salinities as follows. eight individual plants from each species were selected for this experiment and all of them were acclimatized to a moderate salinity by keeping them in trays with excess water of the salinity of 20 ppt for three days while irrigating plants twice a day with the same water. then salt secretion of all these plants, during a 24 hour period under the same salinity, was measured following the same procedure that is described in above 2.1 and used to measure salt secretion under different salinities. jayatissa et al.: a study of salt secretion... 6 ruhuna journal of science 1, pp. 1–15, (2006) at the end of the 24 hour period, one half of the individuals of each species (i.e. four individuals) were transferred to trays with low saline water (i.e. 05 ppt) and the other half to trays with high saline water (i.e. 35ppt). then the salt secretions of these plants during the two subsequent days (i.e. day two and day three) under the new salinity regime were measured separately. the salt secretion during the first day under moderate salinity, and during the second day and third day under low or high salinity was standardized as ‘mg salt cm−2 day−1’. 2.4. data analysis mean and standard deviation values of in situ salt secretion by mangrove plants in rekawa lagoon were calculated to distinguish salt secreting species. one-way anova with tukey-kramer hsd test (zar, 1984) was used to test significant differences of the rates of salt secretion among four of the salt secreting species as well as salt secretion of the same species under different salinity levels. correlation analysis was performed to find out the relationship between soil salinity and salt secretion in different species. regression equations were established to explain significant correlations. 3. results 3.1. interspecific variations of the salt secretion out of the twelve mangrove species tested, only four species, acanthus ilicifolius, aegiceras corniculatum, avicennia marina and avicennia officinalis showed salt secretion. mean values received for the amounts of salts collected from leaves of other eight species were zero or negative when corrected with respect to the control value (table 1). the mean salinity of the lagoon water that was measured from different sites adjacent to individual plants used to measure the salt secretion, during the experimental period was 26.2 ± 4.41ppt. the mean values of salt secretion by the four species are given in figure 1. according to that variations, av. marina showed highest salt secretion that is significantly different from all the others whilst the salt secretion by ac. ilicifolius and ae. corniculatum was the lowest and not significantly different each other. salt secretion by av. officinalis was at the moderate level. 3.2. variations of salt secretion with the variations of soil salinity the variations of the salt secretion by each species with the variations in soil salinity are given in figure 2 and the relationships between the two parameters relevant to each of the four species are given in table 2. it shows a positive relationship between the salt secretion by each species and the soil salinity. figure 2 clearly shows that all the four species secrete salts even under fresh water condition, although the magnitude is lowest. moreover it shows that the salt secretion by av. marina and av. officinalis increase with the increase of soil salinity up to 35 ppt, but the salt secretion by ac. ilicifolius and ae. corniculatum increase only up to 25 ppt and 15 ppt in soil salinity respectively. jayatissa et al.: a study of salt secretion... ruhuna journal of science 1, pp. 1–15, (2006) 7 table 1 results of the in situ measurement of salt secretion by true mangroves in rekawa lagoon. values were corrected with respect to the control experiment and given as mean ± standard deviation. mangrove species mean salt secretion (mg salt cm−2 day−1) acanthus ilicifolius 47.2 ±18.3 aegiceras corniculatum 35.1±16.0 avicennia marina 149.3± 45.9 avicennia officinalis 81.6± 30.5 bruguiera gymnorrhiza 0.0 bruguiera sexangula 0.0 ceriops tagal 0.0 excoecaria agallocha 0.0012 ± 0.0018 heritiera littoralis 0.0 lumnitzera racemosa 0.0 rhizophora mucronata -0.05 ± 0.07 table 2 relationship between salt secretion (‘y’ mg salt cm−2 day−1) by each species and the soil salinity (‘x’ ppt). species relationship r se of se of intercept slope acanthus ilicifolius y = 25.51 + 1.17x 0.86* 3.32 0.16 aeigiceras corniculatum y = 28.46 + 1.12x 0.80* 4.09 0.20 avicennia marina y = 47.00 + 4.39x 0.92* 9.08 0.44 avicennia officinalis y = 47.17 + 1.82x 0.84* 5.61 0.27 3.3. salt secretion under fluctuating salinities salt secretion by av. marina was increased significantly when the plants were transferred from moderate soil salinity to higher soil salinity. but the salt secretions by the other three species, ac. ilicifolius, ae. corniculatum and av. officinalis, were not significantly changed when the soil salinity was increased from moderate to higher level. salt secretions by all the four species were increased significantly when the soil salinity was lowered, but only for a short period. the increased salt secretion was again reduced after the first 24 hour period under the reduced soil salinity (figure 3). 4. discussion salt tolerance of mangroves basically depends on physiological mechanisms they have for the salt and water balance and the efficiency of those mechanisms (kathiresan and bingham,2001; tomlinson, 1988; ye et. al. 2005). salt secretion is one of the remarkable mechanism take place in some mangroves as a way of balancing water and salt in the plant body. however, it is reported that all the mangrove species do not possess this mechanisms in the same order. this study also revealed that only four species, ac. ilicifolius, ae. coniculatum; av. marina, and av. officinalis, out of jayatissa et al.: a study of salt secretion... 8 ruhuna journal of science 1, pp. 1–15, (2006) figure 1 mean values of the salt secretion by different mangrove species.(ai, acanthus ilicifolius; ac, aegiceras coniculatum; am, avicennia marina: ao, avicennia officinalis). note. (longitudinal bars indicate standard deviations. bars with different superscripts, indicate salt secretion values which are significantly different at p < 0.05). soil salinity (ppt) figure 2 variation of salt secretion (mean ± standard deviation) by ac. illicifolius leaves, with the variation of soil salinity. note. longitudinal bars indicate standard deviations. the twelve common mangroves of sri lanka, can secrete salts. this corroborates the jayatissa et al.: a study of salt secretion... ruhuna journal of science 1, pp. 1–15, (2006) 9 soil salinity (ppt) figure 3 variation of salt secretion (mean ± standard deviation) by ae. corniculatum leaves, with the variation of soil salinity. note. longitudinal bars indicate standard deviations. soil salinity (ppt) figure 4 variation of salt secretion (mean ± standard deviation) by av. marina leaves, with the variation of soil salinity. note. longitudinal bars indicate standard deviations. published records and, hence validates the technique used in this study to measure salt secretion. jayatissa et al.: a study of salt secretion... 10 ruhuna journal of science 1, pp. 1–15, (2006) soil salinity (ppt) figure 5 variation of salt secretion (mean ± standard deviation) by av. officinalis leaves, with the variation of soil salinity. note. longitudinal bars indicate standard deviations. figure 6 variation of the salt secretion (mean ± standard deviation) by ac. illicifolius, under fluctuating soil salinity. note. longitudinal bars indicate standard deviations. different superscripts with bars, indicate salt secretion values which are significantly different at p < 0.05. jayatissa et al.: a study of salt secretion... ruhuna journal of science 1, pp. 1–15, (2006) 11 figure 7 variation of the salt secretion (mean ± standard deviation) by ae. corniculatum, under fluctuating soil salinity. note. longitudinal bars indicate standard deviations. different superscripts with bars, indicate salt secretion values which are significantly different at p < 0.05. figure 8 variation of the salt secretion (mean ± standard deviation) by av. marina, under fluctuating soil salinity. note. longitudinal bars indicate standard deviations. different superscripts with bars, indicate salt secretion values which are significantly different at p < 0.05. jayatissa et al.: a study of salt secretion... 12 ruhuna journal of science 1, pp. 1–15, (2006) figure 9 variation of the salt secretion (mean ± standard deviation) by av. officinalis, under fluctuating soil salinity. note. longitudinal bars indicate standard deviations. different superscripts with bars, indicate salt secretion values which are significantly different at p < 0.05. results of this study shows that the capacity to secrete salts at any given salinity was different between four species, following an order of av. marina > av. officinalis > ac. ilicifolius, and ae. corniculatum. av. marina proved to be the species with the highest tolerance to high saline conditions among the true mangrove species in rekawa lagoon (jayatissa and wickramasinghe 2006). many other studies have also shown that av. marina has a better salinity tolerance (clough, 1984; downton, 1982; ye et al, 2005). the order of the capacity to secrete salts by the other three species also parallel to the order of their salinity tolerance. (jayatissa and wickramasinghe 2006; ye et. al. 2005). this indicates that salt secretion by mangroves is related to their salt tolerance. thus, it can be predicted that av. marina could grow in high saline conditions as it proved to be the most efficient salt secreting species. similarly, ac. ilicifolius, ae. corniculatum and av. officinalis mostly grow in less saline habitats. in the present study, all the four species exhibited increase in salt secretion with increases in soil salinity, consistent with previous reports (ball, 1988; sorbrado, 2002; boon and allaway, 1982; ye et al., 2005). however, the two avicennia species, av. marina and av. officinalis, showed better correlation between the salt secretion and soil salinity as their salt secretions increased continuously up to the maximum level of soil salinity. conversely, the salt secretion of the other two species, which are known to have a lower salinity tolerance compared to avicennia species, was reached to a maximum before reaching to the maximum level of soil salinity. these variations also indicate that the salt secretion by mangroves is strongly related to their salt tolerance. jayatissa et al.: a study of salt secretion... ruhuna journal of science 1, pp. 1–15, (2006) 13 it is reported that the salt secretions by mangroves are stimulated by salts in soil (boon and allaway, 1986; drennan and pammenter, 1992; ball, 1988; sobrado, 2002). however, a salt secretion by all the four species even under fresh water condition was recorded in this study. this little secretion can be expected as plants were kept under moderate salinity before transferring them to the fresh water condition and hence at least a lower content of salts could be left in the plant body as well as in soil when the salt secretion was measured. increases of the salt secretion by mangroves with the increasing soil salinity can be expected as salinity is often considered as a “stress” and the salt secretion is a physiological adaptation developed to overcome the stress. but this study revealed that the salt secretion by all the four species immediate after reducing the soil salinity was increased significantly. this increased salt secretion gradually comes back to normal if the plants were remained under the same lower level of soil salinity. this could be an opportunistic removal of salts which was accumulated in the plant body under high saline condition. there are some reports for similar opportunistic actions by mangroves. kathiresan and bingham (2001) reports that mangrove species, particularly those that are less tolerant to high saline conditions, opportunistically absorb and store more water when they are exposed to low saline conditions. lin and sternberg (1994) also reports that the fine root biomass of mangroves increases in the wet season, as a response to decreased salinity of the surface waters, directly enhancing the uptake of low-salinity water. this study reveals that a mangrove species cannot tolerate higher salinities just becoming a salt secreting species. the salt tolerance appeared to be depends mainly on the capability of the species to increase its salt secretion with the increase of soil salinity. relevant to that capability, the four of the salt secreting species can be arranged in order as av. marina > av. officinalis > ac. ilicifolius, and ae. corniculatum. as an active physiological process, the salt secretion may be affected by many other physiological processes and environmental factors. therefore the effects of such factors on salt secretion of mangroves should be studied further under the green house condition as well as natural field condition, in order to understand the mechanism fully. references allen ja, krauss kw, hauf, rd. 2003. factors limiting intertidal distribu-tion of the mangrove species xylocarpus granatum. oecologia 135:110-121. ball mc. 1988. salinity tolerance of the mangroves, aegiceras corniculatum and avicennia marina, part 1. water use in relation to growth, carbon partitioning and salt balance. australian journal of plant physiology 13: 689-698. ball mc. 1996. comparative ecophysiology of mangrove forest and tropical lowland moist forest. in “tropical forest plant ecophysiology” (ss mulkey, rl chazdon and ao smith, eds), pp.461-469. chapman and hall, new york. ball mc. 2002. interactive effects of salinity and irradiance on growth: implications for mangrove forest structure along salinity gradient. trees 16:126-139. jayatissa et al.: a study of salt secretion... 14 ruhuna journal of science 1, pp. 1–15, (2006) ball mc, cochrane mj and rawson hm. 1997. growth and water use of the mangroves rhizophora apiculata and r. stylosa in response to salinity and humidity under ambient and elevated concentrations of atmospheric co2. plant cell and environment 20:1158-1166. boon pi, allaway wg. 1982. assesment of leaf-washing techniques for measuring salt secretion in avicennia marina, vireh, australian journal of plant physiology, 9, 728-734. boon pi, allaway wg. 1986. rates and ionic specificity of salt secretion from excised leaves of the mangrove, avicennia marina (forsk.) vierh. aquatic botany 26: 143-153. clough bf. 1984. growth and salt balance of the mangrove avicennia maraina (forsk.) vierh. and rhizophora stylosa griff. in relation to salinity australian journal of plant physiology 11: 419-430. dahdouh-guebas f, lp jayatissa d, di nitto jo, bosire d, lo seen and koedam n. 2005. how effective were mangroves as a defense against the recent tsunami? current biology 15(12): 443-447. downton, wjs. 1982. growth and osmotic relations of the mangrove, avicennia marina, as influenced by salinity. australian journal plant physiology 9: 519-528. drennan pm, berjak p and pammenter nw. 1992. ion gradients and adenosine triphosphatase localization in the salt glands of avicennia marina (forsk.) vierh. south african journal of botany 58(6): 486-490. dschida w, platt-aloia k. and thomson w. 1992. epidermal peels of avicennia germinans (l.) stern: a useful system to study the function of salt glands. annals of botany 70(6): 501-509. fitzgerald ma, orlovich da and allaway wg. 1992. evidence that abaxial leaf glands are the sites of salt secretion in leaves of the mangrove avicennia marina (forsk.) vierh. new phytologist 120: 1-7. hong wc and eong oj 1984. electron microscopy studies on the salt glands of acanthus ilicifolius l. in: soepadmo e, rao an and macintosh dj. (eds.) proceedings of the asian symposium on mangrove environment; research and management. unesco. pp 827. jayatissa lp and wickramasinghe waadl. 2006. guidance for mangrove replanting; 1. interspecific variations in responses of mangrove saplings to two contrasting salinities. ruhuna journal of science 1: 47-60 kathiresan k and bingham bl. 2001. biology of mangroves and mangrove eosystems. advances in marine biology 40: 81-251. lin, gh and sternberg ldsl 1994. utilization of surface water by red mangrove(rhizophora mangle l.): an isotopic study. bulletin of marine science 54 (1):94-102. naskar k and mandal r. 1999. ecology and biodiversity of indian mangroves, part i global status. daya publishing house, delhi, india. p359 jayatissa et al.: a study of salt secretion... ruhuna journal of science 1, pp. 1–15, (2006) 15 scholander p, hammel ht, hemmingsen ea and cray w. 1962. salt balance in angroves. plant physiology 37:722-729. sobrado m.2002. effect of drought on leaf gland secretion of the mangrove avicennia marina germinans l. trees 16:1-4. tomlinson pb 1986. the botany of mangroves. the cambridge university press. london, uk. ye y. tam, nfy. lu cy, wong sh. 2005. effects of salinity on germination, seedling growth and physiology of three salt secreting mangrove species. aquatic botany 83:193-205. youssef t and saenger p. 1998. photosynthetic gas exchange and accumulation of phytotoxins in mangrove seedlings in response to soil physico-chemical characteristics associated with water logging. tree physiology 18(5): 317-324. zar jh. 1984. biostatistical analysis (second edition). prentice-hall inc. a simon and schuster company, englewood cliffs. new jersey, pp718. ruhuna journal of science vol 11 (2): 143-156, december 2020 eissn: 2536-8400 © faculty of science http://doi.org/10.4038/rjs.v11i2.93 university of ruhuna © faculty of science, university of ruhuna sri lanka 143 survival modelling of teenage childbirth among nigerian women o. m. oladuti department of statistics, the federal university of technology, akure, nigeria correspondence: omoladuti@futa.edu.ng; orcid: https://orcid.org/0000-0001-9852-2066 received: 4th april, 2019, revised: 2nd may, 2020, accepted: 30th december, 2020 abstract childbearing is an essential event in woman’s life when she neglects her education and career for motherhood. the main cause of population growth in nigeria is teenage childbearing as reported by many researchers with no significant intervention to reduce this menace. this work is designed to compare the results of cox, gompertz and weibull models with a view to determine the model that best fits the data. to evaluate the effect of some risk factors on the hazard of teenage childbirth among nigerian women, data obtained from national demographic health survey 2013 was analysed. results of the analysis showed that having a child at teen age depends on geopolitical zone, location, educational level, circumcision, household wealth index, religion belief, use of contraceptive, whether had a terminated pregnancy, forced sexual acts, awareness of hiv/aids, sti and the age of first sexual acts. akaike information criterion (aic) was employed to evaluate performances of the three models. weibull regression model has the minimum aic value compared to both gompertz and cox regression models. this shows that weibull regression model provides best fit to the data. keywords: cox proportional hazard model, risk factors, survival time, teenage childbirth 1 introduction pubescence is a teenage period between the ages 13 and 19 years which marks the beginning of appearance of sexual characteristics and reproductive maturity. this comprises about twenty percent (20%) of the world population with about eighty five percent (85%) in developing countries (nwosu 2005). according to metro (2012), childbirth is the primary killer of teenagers. one of the societal hitches affecting many nations of the world in which nigeria is not an exception is the childbearing among youngsters. adolescent childbirth is an issue of enormous concern in societies experiencing high deficiency (berthoud et al. 2004). in a large scale, one-fifth of 15 http://doi.org/10.4038/rjs.v11i2.93 https://creativecommons.org/licenses/by-nc/4.0/ mailto:omoladuti@futa.edu.ng https://orcid.org/0000-0001-9852-2066 o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 144 million women under the age of 20 give birth and more than fifty percent of these teenage girls involve in sexual acts before their teen ages (guttmacher 2004, maliki 2012). as a result, the rate at which adolescents give birth seems to be on the increase in developing countries (mahfouz et al. 1995, adekanle et al. 2008). according to who (1991), about 16 million teenagers give birth in developing regions though the prevalence of childbearing among teenagers is declining in the developed nations (sarantaki and koutelekos 2002, hamilton et al. 2007, hoffman and maynard 2008). according to united nation (1997), about one-half of women become mothers before their twentieth birthday. helen and anam (2014) examined childbearing trends among 15-19 years old teenagers in nigeria between 1990 and 2008 in their study. the result showed that the proportions of teenagers who had started childbearing were high but decreased by 6%. according to yampolslaaya and greenbaum (2002), roughly 60% of mothers who were teenagers lived in poverty at the time of giving birth and their motherhoods were branded with indignity, humiliation, and eventually dropout from schools (guttmacher 2004, maliki 2012). teenage sexual acts in nigeria have been on increase. the evidences were discovered by the national demographic health survey (ndhs) in 2013, and prevalence around the world and the determinants and its penalties might vary from one region to another. in sub-saharan african countries, sexual acts are more prominent and harmful with only 30% women had their first sexual initiation at the age of 20 and above, 54% of them have the experience before their 18th birthday, an astounding 24% indicated that they had not even been 15 yet before experiencing their sexual acts for the first time while over 50% of young women under age 20 years giving birth to children (okafor 1997, aboyeji et al. 2001, nwosu 2005). this has contributed quite significantly to the fecundity rates observed in the countries of the region, particularly in nigeria. nigeria is one of the countries in africa with high prevalence of teenage pregnancy. different studies have documented results on teenage pregnancy. fagbamigbe et al. (2018) analyzed menarche onset timing among nigerian girls, cox proportional hazard model and accelerated failure time (aft) models were considered. the result showed that menarche age in nigeria stretched broadly between age 9 and 22 years and peaked at age 13 and 14 years. time series analysis techniques ware used by fagbamigbe et al. (2019) to illustrate variations and trends of teenage pregnancy in nigeria between 1961 to 2013. the result revealed that one out of every two girls leaving the teen age is likely to have the knowledge of at least one occurrence of pregnancy in nigeria. the effect of adolescent pregnancy on teenagers in amassoma community of southern ijaw local government area of bayelsa state in nigeria was considered by maliki (2012) using simple percentages. also, a descriptive crosssectional study of teenage pregnancy and incidence of abortion among teenagers in north central nigeria was carried out by aderibigbe et al. (2011). ayo et al. (2016) estimated the regional differences in adolescent childbearing in nigeria with chi-square test and cox model. many of these studies employed o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 145 descriptive statistics and logit model to estimate the impacts of covariates on teenager childbearing which restricted attention to the events that do not occur within the shortest time observed and do not account for the censoring observations. in this study, cox proportional hazard, gompertz, and weibull models were employed to examine the socio-economic and demographic factors influencing adolescent childbirth among nigerian teenagers and compared them using akaike information criterion (aic) to determine the most efficient one that fit the data appropriately. 2 material and methods 2.1 source of data and study design the data used for this study was obtained from national demographic and health survey (ndhs) conducted in 2013. the design for the collection of data was utilized the fact that there are 37 states including federal capital territory (fct) in nigeria. local government areas (lgas) in each state was subdivided into various localities that were further divided into enumeration areas (eas). clusters were used as the primary sampling units based on the enumeration areas. this comprised 372 urban and 532 rural areas across the six geo-political zones in nigeria. from the main data of the survey, female respondents aged 15-49 years data with their complete information were extracted on teenager childbirth, i.e., individuals who had their first child at 13-19 years of age. time to first childbirth of respondents who had their first child between 13-19 years of age inclusively was considered as the survival time in this study and individuals who had not given birth as at the time of survey were right censored. the analyses were based on only 36583 female respondents of age 15-49 years who have the information required for this study. 2.2 cox proportional hazard regression model cox models for survival analysis are usually built from baseline hazard function because there are cases that the exact characteristics of the survivor function are not known but some information is available on how the failure rate changes over time. the survival models measure the failure of an individual at time t and examine the relationship of the survival distribution to covariates under study. cox (1972) introduced proportional hazards model which is most frequently used and the simplest member out of a large family of models that focuses directly on the hazard function. cox proportional hazard model is a semi-parametric, most flexible continuous time model used to investigate covariates effects on hazard function. cox model makes no assumption about the shape of the baseline hazard function but o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 146 estimates the relationship between the hazard and covariates. for an individual 𝑖 with covariate vector 𝑥𝑖 , the cox model is given as 𝜉(𝑡/𝑥) = 𝜉𝑜 (𝑡)𝑒𝑥𝑝(𝛽 ′𝑥), 𝑡 > 0 where 𝜉(𝑡/𝑥) is the hazard of an individual with covariate vector 𝑥, 𝜉𝑜 (𝑡) the baseline hazard function, 𝛽′ the vector of covariates and 𝑥 = (𝑥1, 𝑥2, … , 𝑥𝑝) ′ the vector of explanatory variables. the estimated hazard ratio (hr) is simplified as 𝐻𝑅 = 𝜉(𝑡, 𝑥 ∗) 𝜉(𝑡, 𝑥) = 𝜉�̂� (𝑡)𝑒𝑥𝑝(∑ �̂�𝑖 𝑥𝑖 ∗) 𝑝 𝑖=1 𝜉�̂� (𝑡)exp (∑ �̂�𝑖 𝑥𝑖 ) 𝑝 𝑖=1 = 𝑒𝑥𝑝 {∑ �̂�𝑖 (𝑥𝑖 ∗ − 𝑥𝑖 ) 𝑝 𝑖=1 } 2.3 gompertz regression model gompertz is a parametric model in survival analysis. it is a valuable tool in demography and in other scientific disciplines with parameter 𝜆 > 0 𝑎𝑛𝑑 𝛾 > 0 which determine the scale and shape of the hazard function respectively. the survival and hazard functions are given as 𝑠(𝑡) = exp ( 𝜆 𝛾 (1 − 𝑒 𝛾𝑡 )) and ℎ(𝑡) = 𝜆exp (𝛾𝑡), 0 ≤ 𝑡 ≤ ∞, 𝜆 > 0 the hazard function of gompertz proportional hazard model is given as 𝜉(𝑡/𝑥) = 𝜆𝑒 𝛾𝑡 𝑒𝑥𝑝(𝛽′𝑥), 𝜆 > 0; 𝑡 > 0 2.4 weibull regression model the weibull distribution is another commonly used model for survival analysis that provides estimate of baseline hazard function as well as coefficients for covariates. weibull distribution is the inclusion of shape parameter to generalization of the exponential distribution which makes it more flexible. suppose that survival times have a weibull distribution with scale parameter λ and shape parameter θ, the survival and hazard functions are given as 𝑠(𝑡) = 𝑒 −𝜆𝑡 𝜃 and ℎ(𝑡) = 𝜃𝜆𝑡𝜃−1; λ>0, θ > 0 weibull proportional hazard model is given as 𝜉(𝑡/𝑥) = 𝜃𝜆𝑡 𝜃−1𝑒𝑥𝑝(𝛽′𝑥), 𝜆 > 0; 𝑡 > 0 o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 147 3 results and discussion from national demographic and health survey 2013, information on whether (or not) the respondents have their first birth at teen ages was extracted, age-at-first birth was recorded in years and those who have their first birth after their teen ages were right censored and those who have never given birth were considered right censored at their current ages. also, information on categorical covariates which were thought to be associated with response variable were included, namely, religion (christianity/ muslim/ others), location (urban/ rural area), level of educational attainment (no education/ primary/ secondary/ higher), household wealth index, hwi (poorest/ poorer/ middle/ richer/ richest), current marital status (never in union/ married/ widowed/ no longer married), use of contraceptives (yes/ no), ever had a terminated pregnancy (yes/ no), circumcision (yes/ no), current age group of the respondents (years 15-19/ 20-24/ 25-29/ 30-39/ 40-49, ever heard of hiv/aids (yes/ no), ever heard of sti (yes/ no), ever forced to perform unwanted sexual acts (yes/ no), and geo-political zone (north-central/ north-east/ northwest/ south-east/ south-south/ south-west). the continuous covariate age-at-first sex (in years) was included in the analysis. 3.1 descriptive statistics table 1 presents some descriptive information about the 36583 surveyed samples of ndhs data considered. based on geopolitical zones, the highest number of respondents is from the north-west while south-east has the least number. with respect to locality, 22001 out of 36583 are from rural area. considering the level of education and age range, the majority was with no education and between 30-39 years. household wealth index was richer for the largest number of respondents. out of 36583, 27343 are yet to be circumcised. with respect to religion, 52.35% practices christianity and 46.67% practices islam (muslim). the majority never had a terminated pregnancy nor forced to perform sexual acts. those who had awareness of hiv/aids and sexual transmitted infections (stis) were the highest number with 17 years as survival time. 3.2 models the estimated results for model 1 (table 2), model 2 (table 3) and model 3 (table 4) are presented. the estimated hazard ratios with its coefficients for cox regression model, gompertz regression model and weibull regression model and the p-values with confidence intervals were used to validate whether each covariate contributed significantly to the model or not. each factor contribution in the model is determined by its hazard ratio for the three models. o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 148 table 1: frequency distribution and survival time of the categorical covariates (total 36583) of surveyed samples. factors factors’ levels respondent (%) survival time geo-political zones north-west 8489 (23.20) 18 north-east 6421 (17.55) 16 north-central 6075 (16.61) 16 southwest 5429 (14.84) 16 south-east 4309 (11.78) 18 south-south 5860 (16.02) 19 location urban 14582 (39.86) 19 rural 22001 (60.14) 16 level of education no education 12542 (34.28) 16 primary 6833 (18.68) 17 secondary 13722 (37.51) 19 higher 3486 (9.53) 19 circumcision yes 9240 (25.26) 18 no 27343 (74.74) 17 current age-group 15-19 years 7384 (20.18) 17 20-24 years 6278 (17.16) 17 25-29 years 6631 (18.13) 17 30-39 years 9495 (25.95) 17 40-49 years 6795 (18.57) 16 household wealth index (hwi) poorest 5961 (16.29) 16 poorer 7019 (19.19) 16 middle 7670 (20.97) 17 richer 8057 (22.02) 18 richest 7876 (21.53) 18 religion christianity 19153 (52.35) 19 muslim 17073 (46.67) 16 others 357 (0.98) 16 contraceptive used yes 5930 (16.21) 19 no 30653 (83.79) 17 ever had a terminated pregnancy yes 3869 (10.58) 17 no 32714 (89.42) 17 ever forced to perform sexual acts yes 1118 (3.06) 18 no 35465 (96.94) 17 ever heard of hiv/aids yes 33694 (92.10) 17 no 2889 (7.90) 16 ever heard of sti yes 33946 (92.79) 17 no 2637 (7.21) 16 o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 149 table 2: results of cox proportional hazard model covariates hazard ratio (hr) coefficient s. e p-value 95% conf. int. of hr geo-political zones north-west 1.067 0.065 0.049 0.155 (0.976, 1.167) north-east 1.144 0.135 0.052 0.003 (1.046, 1.252) north-central 1.124 0.117 0.048 0.007 (1.033,1.224) southwest 0.952 -0.049 0.042 0.267 (0.873, 1.038) south-south 1.128 0.121 0.046 0.003 (1.041, 1.223) location rural 0.964 -0.037 0.022 0.113 (0.921, 1.009) level of education no education 1.754 0.562 0.106 <0.001 (1.558, 1.974) primary 1.911 0.647 0.111 <0.001 (1.705, 2.141) secondary 1.470 0.385 0.083 <0.001 (1.316, 1.641) circumcision 1.050 0.048 0.023 0.026 (1.006, 1.095) current marital status single 0.169 -1.776 0.063 0.012 <0.001 (0.147, 0.195) married 1.065 0.052 0.198 (0.967, 1.173) widowed 1.254 0.226 0.083 0.001 (1.102, 1.427) current agegroups 15-19 years 0.638 -0.449 0.024 <0.001 (0.592, 0.688) 25-29 years 0.860 -0.151 0.023 <0.001 (0.817, 0.905) 30-39 years 0.825 -0.193 0.020 <0.001 (0.786, 0.865) 40-49 years 0.810 -0.211 0.022 <0.001 (0.769, 0.853) household wealth index poorer 1.074 0.071 0.025 0.003 (1.025, 1.125) middle 1.146 0.136 0.032 <0.001 (1.086, 1.210) richer 1.060 0.058 0.035 0.079 (0.993, 1.130) richest 0.876 -0.132 0.038 0.002 (0.805, 0.953) religion christianity 0.937 -0.065 0.074 0.408 (0.803, 1.093) muslim 0.963 -0.038 0.075 0.625 (0.827, 1.121) contraceptive used 0.886 -0.121 0.024 <0.001 (0.840, 0.935) ever had a terminated pregnancy 1.066 0.064 0.027 0.010 (1.015, 1.120) forced sexual acts 1.230 0.207 0.067 <0.001 (1.106, 1.368) ever heard of hiv/aids 0.884 -0.124 0.077 0.155 (0.745, 1.048) ever heard of sti 1.092 0.088 0.992 0.334 (0.914, 1.304) age at first sex 0.711 -0.340 0.003 <0.001 (0.707, 0.716) o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 150 table 3: results of weibull regression model covariates hazard ratio (hr) coefficient s. e p-value 95% conf. int. of hr geo-political zones north-west 1.238 0.213 0.057 <0.001 (1.131, 1.355) north-east 1.189 0.173 0.055 <0.001 (1.087, 1.302) north-central 1.303 0.264 0.056 <0.001 (1.197, 1.418) southwest 1.107 0.102 0.049 0.022 (1.015, 1.207) south-south 1.243 0.218 0.051 <0.001 (1.147, 1.347) location rural 0.967 -0.034 0.023 0.145 (0.923, 1.012) level of education no education 2.522 0.925 0.155 <0.001 (2.237, 2.845) primary 2.998 1.098 0.176 <0.001 (2.672, 3.364) secondary 2.041 0.713 0.115 <0.001 (1.827, 2.280) circumcision 1.015 0.015 0.021 0.487 (0.973, 1.059) current marital status single 0.157 -1.849 0.011 <0.001 (0.137, 0.181) married 1.236 0.212 0.061 <0.001 (1.122, 1.362) widowed 1.344 0.295 0.089 <0.001 (1.181, 1.530) current agegroups 15-19 years 0.636 -0.453 0.024 <0.001 (0.590, 0.685) 25-29 years 0.719 -0.330 0.019 <0.001 (0.683, 0.757) 30-39 years 0.569 -0.563 0.014 <0.001 (0.542, 0.598) 40-49 years 0.416 -0.878 0.012 <0.001 (0.393, 0.439) household wealth index poorer 1.060 0.058 0.025 0.015 (1.011, 1.110) middle 1.157 0.146 0.032 <0.001 (1.096, 1.222) richer 1.125 0.118 0.037 <0.001 (1.054, 1.201) richest 0.787 -0.240 0.034 <0.001 (0.722, 0.857) religion christianity 0.925 -0.078 0.073 0.326 (0.792, 1.081) muslim 0.932 -0.070 0.073 0.369 (0.800, 1.086) contraceptive used 0.764 -0.269 0.021 <0.001 (0.724, 0.807) ever had a terminated pregnancy 1.044 0.044 0.026 0.082 (0.994, 1.097) forced sexual acts 1.157 0.146 0.063 0.007 (1.041, 1.287) ever heard of hiv/aids 0.733 -0.310 0.064 <0.001 (0.618, 0.869) ever heard of sti 1.305 0.266 0.118 0.003 (1.093, 1.559) age at first sex 0.734 -0.309 0.002 <0.001 (0.729, 0.739) o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 151 table 4: results of gompertz regression model covariates hazard ratio (hr) coefficient s. e p-value 95% conf. int. of hr geo-political zones north-west 1.220 0.199 0.056 <0.001 (1.114, 1.335) north-east 1.141 0.132 0.053 0.004 (1.043, 1.249) north-central 1.283 0.250 0.056 <0.001 (1.179, 1.397) southwest 1.112 0.106 0.049 0.016 (1.020, 1.213) south-south 1.247 0.221 0.051 <0.001 (1.151, 1.352) location rural 0.974 -0.026 0.023 0.262 (0.931, 1.020) level of education no education 2.401 0.876 0.147 <0.001 (2.129, 2.707) primary 2.884 1.059 0.169 <0.001 (2.570, 3.236) secondary 2.034 0.710 0.115 <0.001 (1.820, 2.272) circumcision 1.010 0.010 0.022 0.631 (0.969, 1.054) current marital status single 0.159 -1.836 0.011 <0.001 (0.138, 0.184) married 1.248 0.222 0.062 <0.001 (1.133, 1.375) widowed 1.272 0.241 0.084 <0.001 (1.118, 1.448) current age-groups 15-19 years 0.585 -0.536 0.022 <0.001 (0.543, 0.630) 25-29 years 0.751 -0.286 0.019 <0.001 (0.714, 0.791) 30-39 years 0.610 -0.494 0.015 <0.001 (0.582, 0.641) 40-49 years 0.402 -0.912 0.012 <0.001 (0.380, 0.425) household wealth index poorer 1.058 0.057 0.025 0.017 (1.010, 1.109) middle 1.154 0.143 0.032 <0.001 (1.093, 1.218) richer 1.121 0.115 0.037 0.001 (1.051, 1.197) richest 0.794 -0.231 0.035 <0.001 (0.729, 0.864) religion christianity 0.929 -0.732 0.073 0.354 (0.796, 1.085) muslim 0.931 -0.712 0.073 0.361 (0.799, 1.085) contraceptive used 0.767 -0.265 0.021 <0.001 (0.727, 0.810) ever had a terminated pregnancy 1.037 0.037 0.026 0.143 (0.988, 1.090) forced sexual acts 1.109 0.103 0.060 0.056 (0.997, 1.233) ever heard of hiv/aids 0.733 -0.311 0.064 <0.001 (1.118, 1.448) ever heard of sti 1.297 0.260 0.118 0.004 (1.086, 1.550) age at first sex 0.753 -0.284 0.002 <0.001 (0.748, 0.758) o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 152 table 5: summary of the results (hazard ratio) of three models. covariates hazard ratio cox weibull gompertz geo-political zones north-west 1.067 1.238* 1.220* north-east 1.144* 1.189* 1.141* north-central 1.124* 1.303* 1.283* southwest 0.952 1.107* 1.112* south-south 1.128* 1.243* 1.247* location rural 0.964 0.967 0.974 level of education no education 1.754* 2.522* 2.401* primary 1.911* 2.998* 2.884* secondary 1.470* 2.041* 2.034* circumcision 1.050* 1.015 1.010 current marital status single 0.169* 0.157* 0.159* married 1.065 1.236* 1.248* widowed 1.254* 1.344* 1.272* current age-groups 15-19 years 0.638* 0.636* 0.585* 25-29 years 0.860* 0.719* 0.751* 30-39 years 0.825* 0.569* 0.610* 40-49 years 0.810* 0.416* 0.402* household wealth index poorer 1.074* 1.060* 1.058* middle 1.146* 1.157* 1.154* richer 1.060 1.125* 1.121* richest 0.876* 0.787* 0.794* religion christianity 0.937 0.925 0.929 muslim 0.963 0.932 0.931 contraceptive used 0.886* 0.764* 0.767* ever had a terminated pregnancy 1.066* 1.044 1.037 forced sexual acts 1.230* 1.157* 1.109* ever heard of hiv/aids 0.884 0.733* 0.733* ever heard of sti 1.092 1.305* 1.297* age at first sex 0.711* 0.734* 0.753* *significance p<0.05 o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 153 table 5 shows the summary of the results comprising the estimated hazard ratio (hr), coefficients, standard errors and confidence intervals for cox proportional hazard, weibull regression and gompertz regression models respectively. based on geopolitical zones, those who are in north-east, north-west, northcentral, south-west and south-south were more likely to experience adolescent childbirth while south-west were less likely with cox model compare to those who are in south-east. those who reside in rural area were less likely to have teenage childbirth compared to their counterpart. from the level of education, those that have no education, primary and secondary education were more likely to have teenage childbirth (p<0.001) compared to those who have higher education. nevertheless, with gompertz and weibull models, respondents with no education, primary and secondary education were significantly (p<0.001) twice more likely to give birth at teen age than those with higher education. those that were not circumcised were more likely to have their first child at their teen period than those that were circumcised. based on their current marital status, respondents who are presently single were less prone to teenage childbirth (p<0.001) while those who are currently married and widowed were more likely to have teen childbirth at their teen age compared to those divorced/ separated. those within age-groups 15-19, 25-29, 30-39 and 40-49 years had the hazard ratios of which are significantly (p<0.001) lower than those aged 2024 years which implies that there were less teenage childbirths in those age groups. based on the household economic status variable measure by wealth index, those who are poorer, middle and richer were more likely to give birth for the first time at their teen age while the richest were less prone to it compared to the poorest. according to the results based on the religion belief, the respondents who practice christianity and muslim were less likely to have teenage childbirth than others. respondents that are not using contraceptive were less prone to teenage childbirth compared to those that are using it. those that had never terminated pregnancy were more likely to have to teenage childbirth than their counterparts. for those who were never forced to perform unwanted sexual acts, they were more likely to have a baby at their teen age compared to their counterparts. those who heard of hiv/aids were less prone to teenage childbirth compared to those who did not, and also those who heard of sti were more likely to have their first child at teen age compared to their counterparts. the hazard of teenage childbirth decreases with a one-year increase in age at first sex by 0.711 (p<0.001) which means the earlier the age at first sexual initiation the greater the chance of having a child at teen age. akaike information criteria were used to compare the three models under proportional hazards framework of survival data. the relative performances of the models are shown in table 6. it was observed that weibull has the least value (aic=22488.98) which shows best performance while aic values of cox regression and gompertz models are 283793.1 and 31343.8, respectively. o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 154 table 6: comparison of relative performances of cox, gompertz and weibull models. models number of observations log-likelihood aic cox 36583 -141867.60 283793.00 weibull 36583 -11213.49 22488.98 gompertz 36583 15640.90 31343.80 fig 1. kaplan-meier (k-m) plots of survival and hazard functions of time. plots of the kaplan-meier (k-m) curves to the survival and hazard experience of time-to-teenage childbirth presented in figure 1 show that the event occur between ages 16 and 19. o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 155 4 conclusions we have successfully modelled the prevalence of teenage childbirth and the risk factor associated with it among nigerian women. data used to model the teenage childbearing were obtained from national demographic health survey 2013 and the cox regression model was applied to test the effect of each covariate (categorical and continuous) in the hazard rate. the results showed the rate of teen age childbearing was related in all geo-political areas though slightly higher with people from northeast, north-west, north-central, south-west and south-south. those living in urban area had higher risk of given birth at teen age compare to those in rural areas. a respondent with no education, primary and secondary has a tendency of given birth at teen age. the risk of teenage childbirth for those who were not circumcised was slightly higher. respondents whose household wealth index is richest were less prone to teenage childbirth. based on their religion belief, those who practice christianity and muslim have the lower risk of teenage childbirth than other. non usage of contraceptives decreases risk of teenage childbirth. respondents who had never terminated pregnancy nor forced to perform sexual acts have higher risk of teenage childbirth. those who are aware of sexual transmitted infections (stis) have higher risk of teenage childbirth than those who heard of hiv/aids. we also observed that the higher the age at first sexual acts, the lower the risk of teenage childbirth. it might be because of knowledge acquired based on level of education or awareness through social media on sexual transmitted infections (stis) and menace of having a child at teen age. based on the three models, the results revealed that weibull regression model is the best since it has the least aic value. this shows that the parametric weibull regression model could better determine the factors associated with teenage childbearing than both the parametric gompertz model and semi-parametric cox model, respectively. acknowledgments two anonymous reviewers are acknowledged for valuable comments on the initial draft of the manuscript. references aboyeji ap, fawole aa, ijaiya ma. 2001. knowledge and previous contraceptive use by pregnant teenagers in ilorin, nigeria. tropical journal of obstetrics and gynecology 18(2):73-76. aderibigbe s, araoye m, akande t, olumuyiwa mj, ibrahim mo, babatunde o. 2010. teenage pregnancy and prevalence of abortion among in-school adolescents in north-central, nigeria. asian social science 7(1): 122-123 adekanle da, adeyemi as, odu oo. 2008. teenage and nonteenage pregnant women in southwestern nigeria: a descriptive study. calicut medical journal 6 (3): 693, e5 o. m. oladuti teenage childbirth among nigerian women ruhuna journal of science vol 11 (2): 143-156, december 2020 156 adeniyi f, fagbamigbe rotimi f, afolabi oyindamola, b yusuf. 2019. trend analysis of teenage pregnancy in nigeria (1961-2013): how effective is the contraceptive use campaign? international journal of public health science 8 (2): 163-173. ayo s, adebowale, adeniyi f, fagbamigbe, ayodeji m adebayo. 2016. regional differences in adolescent childbearing in nigeria. journal of population and social studies 24(2): 101-116. doi 10.14456/jpss.2016. berthoud r, ermisch j, fransesconi m, liao t, pevalin d, robson k. 2004. long term consequences of teenage births for parents and their children. second edition, jessica kingsley publishers, london and philadelphia. cox dr. 1972. regression models and life tables. journal of the royal statistical society series b (methodological) 34: 187-220. fagbamigbe af, obiyan mo, fawole oi. 2018. parametric survival analysis of menarche onset timing among nigerian girls. heliyon 4 (2018) e01105. doi: 10.1016/j.heliyon.2018. e01105 guttmacher. 2004. sex and america’s teenagers. available at www.guttmacher.org. accessed on march 7, 2011. hamilton be, martin ja, ventura sj. 2007. births preliminary data for 2006. hyattsville, md: national center for health statistics. national vital statistics reports 56(7). helen na, anam b. 2014. teenage childbearing trend in nigerian 1990-2008. education and science journal of policy review and curriculum development 4(1): 17418771. hoffman sd, maynard ra. 2008. kids having kids: economic costs and social consequences of teen pregnancy. second edition, urban institute press, washington dc. jiezhi q. 2009. comparison of proportional hazards and accelerated failure time models. thesis submited to the college of graduate studies and research, university of saskatchewan, department of mathematics and statistics. maliki ae. 2012. the effect of teenage pregnancy on adolescents in amassoma community of southern ijaw local government area of bayelsa state. asian social science 8. 10.5539/ass. v8n1p62. mahfouz aa, el-said mm, al-erian ra, hamid am. 1995. teenage pregnancy: are teenagers a highrisk group? european journal of obstetrics & gynecology and reproductive biology 59 (1): 17-20. metro. 2012. childbirth is the number one killer of teenagers. metro news. published june 27, 2012. nwosu an. 2005. effects of health education on the reproductive health behavior of in-school adolescent girls in enugu. doctoral dissertation, university of nigeria. okafor a. 1997. sexual knowledge and sources of sexual information of secondary school students in anambra state, nigeria. health and movement education journal 1(1): 9-15. svetlana y, eric cb, paul eg. 2002. early pregnancy among adolescent females with serious emotional disturbances risk factors and outcomes. journal of emotional and behavioral disorders 10 (2):108115. sarantaki a, koutelekos i. 2002. teenage pregnancy. health science journal 1 (2): 1-6. united nation. 1997. the right to reproductive and sexual health. united nations department of public information. viegas oa, wiknsosatro g, sahagun gh, chaturachinda k, ratnam ss. 1992. safe childbirth needs more than medical services. world health forum 13: 59-65. who 1991. world health organisation, maternal and mortality ratios and rates: a tabulation of available information. 3rd edition. geneva unpublished report. who 2005. world health organisationfacts and figures from the world health report. who 2013. world health organization, nigeria: maternal and perinatal country profile 13(1), geneva. http://www.who.int/maternal_child_adolescent/epidemiology/ profiles/maternal/ nga.pdf , accessed on february 9, 2013. yampolskaya s, brown e, greenbaum p. 2002. early pregnancy among adolescent females with serious emotional disturbances: risk factors and outcomes. journal of emotional and behavioral disorders 10(2): 108-115. rjs-vol-1-sept-2006-6.dvi ruhuna journal of science vol. 1, september 2006, pp. 47–60 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. guidance for mangrove replanting: 1. interspecific variations in responses of mangrove saplings to two contrasting salinities l.p. jayatissa department of botany, university of ruhuna, matara, sri lanka. lpj@ruh.ac.lk w.a.a.d.l. wickramasinghe department of botany, university of ruhuna, matara, sri lanka. lankadimuthu@yahoo.com the early growth of seven species of true mangroves representing all the categories relevant to viviparity (i.e. true viviparous species, crypto viviparous species and non-viviparous species) and including pairs of species which are closely related as well as species commonly used in replanting, was studied in response to two contrasting salinity regimes, low saline (i. e. 3-5 ppt) and high saline (i.e. 25-27 ppt). growth performance of the seven species (i.e. avicennia marina, a. officinalis, bruguiera gymnorrhiza, b. sexangula, rhizophora apiculata, rhizophora mucronata, and sonneratia caseolaris) in terms of plant dry weight, relative growth rate (rgr), leaf area and shoot height was assessed. the percentage water content of plants under the two salinity levels was also assessed. performances of all the aspects of a. marina under the low saline and high saline conditions were not significantly different implying that this species has the highest salinity tolerance among the seven species. s. caseolaris did not survive under high saline conditions proving that it is the lowest in salinity tolerance. the performances of the other five species were in between these two ends, and showed considerable variation. the rgr of each of a. officinalis, b. gymnorrhiza, b. sexagula and r. apiculata was significantly lower under high saline conditions, with reductions in growth compared with low salinity conditions of 51%, 40%, 64% and 32% respectively. by considering variations in the performance of all the factors assessed, it was possible to arrange the seven species in descending order of salinity tolerance as a. marina > r. mucronata > b. gymnorriza and r. apiculata > a. officinalis, and b. sexangula > s. caseolaris, showing that even taxonomically similar species may be distant in salinity tolerance. the percentage water content of the least saline tolerant mangrove species, i.e. a. officinalis, b. sexangula, rhizophora apiculata, was higher when they were grown under low saline conditions, implying that species with less tolerance to salinity may opportunistically absorb and keep more water when the salinity is low. as salinity of the habitat appears to be a primary factor controlling the survival and growth of seedlings planted, these interspecific variations in salinity tolerance of species should be taken into consideration in mangrove replanting. key words : salinity tolerance, mangrove replanting, interspecific variation, growth 1. introduction mangroves are woody shrubs and trees that are salt and flood tolerant and hence dominate intertidal areas of lagoons, estuaries and sheltered bays along tropical and 47 jayatissa and wickramasinghe: guidance for mangrove replanting ... 48 ruhuna journal of science 1, pp. 47–60, (2006) subtropical coastlines (ball 2002; tomlinson 1986; tuffers et al. 2001). these tidal forests are of enormous ecological and economic importance (bandaranayake 1998; bandaranayake 2002). despite the importance of mangroves in providing ecosystem goods (food, medicines and timber) and services (such as fisheries nurseries and erosion control) to local communities living behind and within the forest, reportedly 50% of the world’s mangrove forests have been destroyed in the second half of the 20th century, and current loss rates vary from 1 to 20% of total forest area per year (alongi 2002). hence the conservation and restoration of mangrove ecosystems deserves higher priority. the increasing awareness of sea level rise due to global warming, which threatens to entirely inundate much land and render other low lying areas suitable only for salt tolerant plants (dahdouh-guabas et al. 2005), has raised the profile of mangroves as potential coastal protection belts. the massive tsunami that hit southeast asia on december 26th 2004, killing over 400,000 people and leaving millions homeless, was a dramatic and tragic reminder of this ecological function. in the aftermath of the killer tsunami, the common-sense view that mangroves can act as living dykes against ocean surges was taken seriously and received empirical support (clarke 2005; dahdouh-guebas et al. 2005; danielsen et al. 2005 ; liu et al. 2005; williams 2005). concurrently, governments across the indian ocean have announced a plethora of new schemes to protect and replant mangroves and thereby attempt to rectify the widespread losses of mangroves during the last decades. when a destroyed mangrove area is going to be replanted, ecological aspects of mangroves should be taken into consideration particularly in choosing the species suitable for the selected site. salinity and hydrology (i.e. period and frequency of flooding) in selected habitats are some of the primary factors which determine the survival and growth of replanted mangroves and, hence the success of the replanting projects, because different true mangrove species vary in tolerance to such ecological factors (allen et al. 2003; hwang and chen 2001; ye et al. 2005). although the salinity is often considered as a “stress”, nacl may also be a resource for halophytic species (ball 2002). the classic growth response of halophytes, including mangroves, to increasing salt concentration is similar to that shown for nutrients, with variation in the shape of the response curve reflecting concentrations which are deficient, saturating and toxic to growth (ball 2002). however, the concentration at which salt becomes toxic depends on the species. although mangroves are a group of salt tolerant plant species, the degree of tolerance varies depending on the species (tomlinson 1988; ye et al. 2005). this range of degrees of tolerance to salinity is one of the factors thought to generate zonation patterns in mangrove communities in the intertidal zone (macnae 1968). although the salinity tolerances of some mangrove species have been studied, much information on salt tolerance that would be of use to restoration efforts remains to be discovered. the salinity tolerance or water use characteristics of the same species may vary depending on climatic and edaphic factors (youssef and saenger 1998). hence field studies of mangrove distribution and growth correlated with salinity will usually be confounded by many other pertinent variables. growth jayatissa and wickramasinghe: guidance for mangrove replanting ... ruhuna journal of science 1, pp. 47–60, (2006) 49 and mortality in the field is likely to be controlled by the most severe conditions the tree experiences. therefore, in the present study, the performance of eight true mangrove species, including species commonly used in replanting, were tested under conditions representing extreme, but naturally occurring, salinity levels. viviparity may also affect the salinity tolerance of mangrove saplings (ye et al. 2005). the distribution and composition of mangrove species in sri lanka, a small island with one third of the worlds true mangrove species (jayatissa et al. 2002), show that the co occurrence of two species of the same genus in the same habitat is unlikely (jayatissa, unpublished data), implying different mangrove species in the same genus also may have different tolerance levels for edaphic factors. therefore, mangrove genera with different categories of vivipary were selected for this study, with replicate species within in genera selected where possible. 2. objectives the main objective of this project was to study the interspecific variations in the salinity tolerance of common mangroves in sri lanka, at their early growth. as a specific objective, particular attention was paid to differences between taxonomically more related species in the same genera in their responses to high saline conditions, because such variations are apparently neglected in many replanting programs, particularly in which experts are not involved, thus leading to failures. 3. materials and methods 3.1. selection of species out of the fourteen mangrove genera in which 20 species of true mangroves are reported to occur in mangrove communities along the coastline of sri lanka, four genera include more than one species in each genus (jayatissa et al. 2002). the responses of a mangrove species to harsh environmental conditions in its early growth may vary depending on whether the species is viviparous or not (ye et al. 2005). therefore vivipary was considered during the selection of species for this study. by considering these two facts, eight common species were selected for the study as four species from viviparous genera, i.e. bruguiera gymnorrhiza, b. sexanguila, r. epiculata, and rhizophora mucronata, two species from cryptoviviparous genera, i.e. avicennia marina, and a. officinalis, and two species from the nonviviparous genera, sonneratia caseolaris and s. alba. however, mature seeds of s. alba, a comparatively rare species, were not available during the study period, hence the study was restricted to the rest of the seven species. 3.2. culture and experimental design mature propagules or seeds of the selected species were collected from natural mangrove sites and used as planting materials. a sandy soil was prepared by mixing sieved loam soil with sand and organic matter (i.e. degraded mangrove litter) in 1:1:1 proportion. initially propagules and seeds of all the species were planted in plastic pots (with 5 cm diameter and 15 cm height) filled with the prepared soil jayatissa and wickramasinghe: guidance for mangrove replanting ... 50 ruhuna journal of science 1, pp. 47–60, (2006) mixture and kept in a nursery irrigated with fresh water until the establishment of seedlings. seedlings with the first two leaves unfurled were considered as established seedlings; eight seedlings from each species, all of similar size, were transferred individually to larger plastic pots (15 cm diameter and 40 cm height) filled with the same mixture of soil. pots with seedlings were placed individually on 7 cm deep plastic trays to get extra water collected after watering and four replicate pots from each species were assigned to each salinity treatment. the experiment was started with these established seedlings in order to minimize masking of salinity effects by other effects (i.e. effects of seed or propagule quality). two salinity regimes, ‘low saline’ (i.e. 3-5 ppt) and ‘high saline’ (i.e. 25-27 ppt), were selected for the experiment. low saline and high saline water was prepared by mixing seawater and tap water, left in tanks and used to irrigate seedlings in pots. each pot was irrigated twice a day by the water with the salinity assigned to each pot. excess water accumulated in trays was returned to tanks every other day and the salinity of the water in tanks was checked by a hand refractometer (atago s/mill-e, japan) and adjusted by adding tap water when necessary once every four days. commercially available fertilizer was also applied once a month by dissolving a recommended dose in high saline and low saline water before used to irrigate seedlings in pots. seedlings in pots were distributed and left in the green house according to a completely randomized design. 3.3. data collection the shoot height of each sapling (starting from the propagule end in the case of the viviparous species) was measured once a week. plants or saplings were harvested after three months of growth. the plastic pots were removed, and the soil carefully washed away by tap water to get the intact root system. cleaned plants were blotted dry and separated into roots, hypocotyls, stem and leaves. the fresh weights of these four parts of each plant were measured and leaf area was quantified manually using millimeter paper on which the exact size and shape of leaves were marked. then all parts were oven dried at 60� for dry weight. the difference between the dry weight and fresh weight of individual plants was taken as the water content. relative growth rate (rgr) for each plant was calculated for individual plants according to; rgr = w/t where w is the dry weight of each plant without the hypocotyl of viviparous species and t is the growth period, i.e. 13 weeks (ye et al. 2005). it was observed that hypocotyl part of planted propagules of rhizophora and bruguiera were enlarged slightly during the 13-week growth period implying that hypocotyls also were grown. it can be assumed that their dry weight have also been increased during that growth period. if it is true, the initial dry weight of the propagule should be subtracted from the dry weights of each of the three months old saplings to get the growth of the sapling to calculate rgr of each sapling. in order to calculate the initial dry weight of propagules of three months old saplings, jayatissa and wickramasinghe: guidance for mangrove replanting ... ruhuna journal of science 1, pp. 47–60, (2006) 51 a regression of dry weight over fresh weight of mature propagules at the initial stage is needed. but when the mean dry weight of propagule parts of three months old saplings, were compared with the mean dry weight of mature propagules at the initial stage, it was revealed that the dry weight of propagule parts of saplings was lower, although the propagule parts of saplings have been grown. it is possible as a large amount of foods are stored in propagules at the initial stage and those food reserves are used for the early growth of the plants converting storage tissues of the propagule into structural tissues of the sapling. therefore, in the calculation of rgr, the propagule parts of three months old saplings were omitted and the dry weight of other parts, i.e. roots, newly grown stem, and leaves, were considered as the net growth of the total sapling during the experimental period. 3.4. data analysis mean and standard deviation values of total dry weight, mean leaf size, percentage water content and rgr were calculated separately for quadruplicate pots of each species grown under each salinity level. two sample t-tests (zar 1984) were carried out for individual species to learn whether the plants grown under two different salinities are different in each of the above parameters. one-way anova with tukey-kramer hsd test (zar 1984) was used to test significant interspecific differences in dry weight, mean leaf size, water content and relative growth rate among the seven mangroves. for this purpose, data under high saline condition and low saline condition were considered and used as one set of data, neglecting the salinity levels. 4. results 4.1. interspecific variations irrespective of salinity effects the comparison of the seven mangrove species for interspecific variations in dry weight, percentage water content, rgr and mean leaf size revealed that r. mucronata was significantly different from all the other species showing the highest values in three factors, dry weight, rgr, and mean leaf size, and the lowest value in percentage water content (table 1). for all the species studied, the higher values of percentage water contents were recorded when they are grown under low saline conditions and the highest water content was recorded from s. caseolaris under low saline conditions. when differences between the two species in each of the same genus are considered, r. apiculata and r. mucronata show significant differences in rgr, dry matter accumulation and percentage water content whilst the two species in the genus avicennia, i.e. a. marina and a. officilanis, show a significant difference only in percentage water content (table 1). b. gymnorrhiza and b. sexangula were not significantly different in rgr and percentage water content, but in mean dry weight and mean leaf size. the highest diversity among species was recorded in percentage water content. the growth rate in terms of the increment of shoot height also shows a remarkable interspecific variation. the lowest and highest growth rates were recorded from b. sexangula and both species of avicennia respectively. the growth rates of the rest of the species were at intermediate levels (figure 1). jayatissa and wickramasinghe: guidance for mangrove replanting ... 52 ruhuna journal of science 1, pp. 47–60, (2006) 4.2. interspecific variations in response to salinity the negative effect of salinity on the growth of saplings was most pronounced for s. caseolaris as all the individuals under high saline conditions died within the first two weeks period. however the growth curve of s. caseolaris saplings under low saline conditions shows the highest growth rate after the first five weeks (figure 1g). growth curves of the other species show that the growth of saplings of a. officinalis and b.sexagula under high saline conditions was greatly reduced whilst that of b. gymnorrhiza and r. apiculata was moderately reduced. salinity effects on the height growth of shoots of a. marina and r. mucronata were negligible (figure 1). figure 2 corroborates this result. it shows that rgr values of saplings of each of a. marina and r. mucronata, grown under low saline and high saline conditions are not significantly different whilst those of the other species are significantly different. at the high saline condition, decreases in rgr of a. officinalis, b. gymnorrhiza, b.sexagula and r. apiculata were 51%, 40%, 64% and 32% respectively. the patterns of salinity effects on the dry weight of saplings of the seven mangroves are similar to those shown for rgr. the mean dry weight of a. marina saplings grown under high saline conditions was not significantly different from that of saplings grown under low saline conditions. it is true for r. mucronata also. sapling dry weights of each of a. officinalis, b. gymnorrhiza, b.sexagula and r. apiculata under low saline and high saline conditions were significantly different (figure 3). the percentage water content in three month old saplings of each of a. officinalis, b. sexangula and r. apiculata grown under low saline conditions were significantly different from those grown under high saline conditions. saplings of a. marina, b. gymnorrhiza and r. mucronata did not show such a difference in their water content (figure 4). apart from a. marina, in which leaf size under low saline conditions and high saline conditions was not significantly different, leaves produced by all the other species under high saline conditions differed significantly in size from those produced under low saline conditions (figure 5). 5. discussion in planning the rehabilitation and reconstruction efforts after the recent killer tsunami that hit south-east asia on december 26th, 2004, many countries gave a high priority for the re-establishment of natural barriers against tsunami and other ocean surges. concurrently, mangroves were subjected to extensive discussions as a potential natural barrier against tsunami and some other ocean driven disasters (clarke 2005; dahdouh-guebas et al. 2005; danielsen et al. 2005; liu et al. 2005; williams 2005). mangrove replanting programs were initiated and supported with the help of governmental and non-governmental organizations, particularly in tsunami affected countries. however, the success of such programs depends on the selection of suitable species based on the prevailing edaphic conditions at the replanting site. although true mangrove species show extreme adaptations to harsh environmental conditions in general, most of the species require specific conditions jayatissa and wickramasinghe: guidance for mangrove replanting ... ruhuna journal of science 1, pp. 47–60, (2006) 53 0 1 2 3 4 5 6 7 8 9 10 11 12 13 sh o o t h e ig h t (c m ) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 sh o o t h e ig h t (c m ) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 sh o o t h e ig h t (c m ) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 sh o o t h e ig h t (c m ) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 sh o o t h e ig h t (c m ) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 sh o o t h e ig h t (c m ) figure 1 variations of the mean shoot (i.e. stem height) growth of mangrove saplings grown under two different salinity regimes (low saline = 4-5 ppt; high saline = 25-26 ppt.). note. longitudinal bars indicate standard deviations jayatissa and wickramasinghe: guidance for mangrove replanting ... 54 ruhuna journal of science 1, pp. 47–60, (2006) table 1 interspecific differences in mean dry weight, mean leaf size, % water content and relative growth rate of seven mangrove species, as resulted by one-way anova. (the data from s. caseolaris was not included to the anova as the available data were from plants grown under low salinity level only.) species mean dry mean leaf water rgr weight (g) size (cm2) content (%) a. marina 5.97d 9.39c 76.31b 0.4587bc a. officinalis 7.09 d 20.02bc 79.99a 0.5437bc b. gymnorrhiza 12.09c 23.39b 71.75cd 0.440 bc b. sexangula 5.01d 10.70c 73.93cd 0.2337 c r. apiculata 16.49b 36.58 a 71.20 d 0.5662 bc r. mucronata 23.39a 40.10a 66.86e 1.4962a s. caseolaris 11.65 16.8 84.32 0.8961 ∗ different superscripts with each value denote significantly different (p< 0. 05) groupings according to tukey-kramer hsd test. low salinity high salinity figure 2 variations of relative growth rate (rgr) of 90 day old saplings of six mangrove species (am, avicennia marina; ao, avicennia officinalis, bg, bruguiera gymnorrhiza; bs, bruguiera sexangula; ra, rhizophora apiculata; rm, rhizophora mucronata) grown under two different salinity regimes(low saline = 4-5 ppt; high saline = 25-26 ppt) for three months. note. (longitudinal bars indicate standard deviations. different superscripts with two bars relevant to each species, indicate that rgr values under high saline and low saline conditions are significantly different at p<0.05). and a narrow range of many ecological factors for their optimum growth (kathiresan and bingham 2001). salinity is one of the major environmental factor controlling jayatissa and wickramasinghe: guidance for mangrove replanting ... ruhuna journal of science 1, pp. 47–60, (2006) 55 figure 3 variations of the mean dry weight of saplings of different mangrove species (am, avicennia marina; ao, avicennia officinalis, bg, bruguiera gymnorrhiza; bs, bruguiera sexangula; ra, rhizophora apiculata; rm, rhizophora mucronata; sc, sonneratia caseolaris) grown under two different salinity regimes (low saline = 4-5 ppt; high saline = 25-26 ppt.) for three months. note. (longitudinal bars indicate standard deviations. different superscripts with two bars relevant to each species, indicate that dry weight values under high saline and low saline conditions are significantly different at p<0.05). the growth and survival of mangrove plants (allen et al. 2003; hwang and chen 2001; ye et al. 2005). effects of such factors could be crucial particularly at the early growth of saplings, although the effects of salinity on seedling establishment were not considered here since most replanting programs are started with seedlings established in a nursery where fresh water or low saline water is used for irrigation. in this study, responses of mangrove saplings under low saline and high saline conditions were studied for only 13 weeks. some studies may suggest that such a short period may be insufficient to reveal slowly developing responses (ball et al. 1997). nevertheless, rapid responses to soil salinity during early growth of saplings may be important determinants of their survival. as an example, as b. sexangula under high saline conditions shows a severe inhibition (i.e. 64%) of its rgr after three months of growth, further growth and development is also likely to be impaired, particularly once the initial food reserves in the propagule are exhausted. it is reported that viviparous species perform better than non-viviparous species in seedling establishment, implying the beneficial effects of propagules in early growth (ye et al. 2005) out of the seven species tested in this study, s. caseolaris proved that it is strictly a low saline species, as all the seedlings under high saline conditions died within jayatissa and wickramasinghe: guidance for mangrove replanting ... 56 ruhuna journal of science 1, pp. 47–60, (2006) low salinity high salinity figure 4 variations of the percentage water content of 90 day old saplings of six mangrove species (am, avicennia marina; ao, avicennia officinalis, bg, bruguiera gymnorrhiza; bs, bruguiera sexangula; ra, rhizophora apiculata; rm, rhizophora mucronata) grown under two different salinity regimes (low saline = 4-5 ppt; high saline = 25-26 ppt) for three months. note. (longitudinal bars indicate standard deviations. different superscripts with two bars relevant to each species, indicate that % water content values under high saline and low saline conditions are significantly different at p<0.05). the first two weeks. this suggests that s. caseolaris is not suitable for high saline areas, although mangrove dwellers may favor this species over the other species due to economic uses of the species (jayatissa et al. 2006). the distribution of this particular species in sri lanka is mainly restricted to river estuaries of the wet zone of the country (jayatissa et al. 2002; jayatissa et al. 2006) supporting the fact that it is a low saline species. however, salt tolerance of mature individuals of s. caseolaris could be higher than that of saplings (bhosale 1994). the rest of the six species survived under high saline conditions but showed some differences in growth performances, with the exception of a. marina. the reduction of the leaf size was the first apparent sign of the salt stress in this study, indicating that the leaf size is the most sensitive factor for salinity (parida and das 2005). out of the mangroves tested in this study, a. marina is the only species that did not show a significant reduction of the leaf size under high saline conditions. therefore, a. marina proved to be the species with the highest tolerance to high saline conditions among the species tested in this study. many other studies have also shown that a. marina has a better salinity tolerance (clough 1984; downton 1982; ye et al. 2005). dry matter accumulation, rgr, and water content of r. mucronata saplings grown under high saline condition were not significantly different from those grown jayatissa and wickramasinghe: guidance for mangrove replanting ... ruhuna journal of science 1, pp. 47–60, (2006) 57 low salinity high salinity figure 5 variations of the mean leaf sizes of saplings of six mangrove species (am, avicennia marina; ao, avicennia officinalis, bg, bruguiera gymnorrhiza; bs, bruguiera sexangula; ra, rhizophora apiculata; rm, rhizophora mucronata) grown under two different salinity regimes for three months period (low saline = 4-5 ppt; high saline = 25-26 ppt) for three months. note. (longitudinal bars indicate standard deviations. different superscripts with two bars relevant to each species, indicate that mean leaf sizes under high saline and low saline conditions are significantly different at p<0.05). under low saline conditions. the leaf size was the only exception. therefore, out of the seven species studied, r. mucronata appeared to be the second highest in salinity tolerance. when the decreases or inhibitions of leaf size, rgr and dry matter accumulation under high saline conditions are considered, the species tested in this study can be arranged in descending order of the salinity tolerance as a. marina > r. mucronata > b. gymnorriza, & r. apiculata > a. officinalis, and b. sexangula > s. caseolaris. s. alba was not included in this study as it is a rare species in sri lanka. but documentary evidences say that s. alba can grow well under high saline conditions (ball and pidsley 1995).this grading shows that taxonomically more related species i.e. species in the same genera, may be distant in salinity tolerance. there may be some discrepancies in the optimal salinity for the better performances of mangroves, grown in the field versus in green house conditions (hwang and chen 2001). it could be due to the fact that the salinity level in the field could fluctuate and plants preferentially take up water when the salinity is low. but under the greenhouse conditions plants exposed continuously to the same salinity level. however, when the distribution of these species in sri lanka are considered, it was noted that these pairs of species, which are in the same genus but distant in salinity tolerance, rarely jayatissa and wickramasinghe: guidance for mangrove replanting ... 58 ruhuna journal of science 1, pp. 47–60, (2006) exist in the same micro habitat, although all the species are in the same community, and the distribution of the last three species of the above series, are restricted to mangrove communities with more riverine influence (jayatissa et al. 2002; jayatissa unpublished data). in general, plants growing in habitats with a deficit of water possess adaptations to store water. mangrove plants also have to face to a physiological drought due to the higher soil salinity implying that mangroves growing under higher salinities should have higher water content. in contrast, this study reveals that the percentage water content in saplings of species less tolerant to salinity, i.e. a. officinalis, b. sexangula, rhizophora apiculata, were higher when they were grown under low saline conditions. in a mangrove habitat, the soil salinity is not constant but fluctuates depending mainly on the fresh water inflow and, in the case of sri lanka, blocking of the lagoon mouth. mangrove species, particularly those that are less tolerant to high saline conditions, could be opportunistically absorb and store more water when they are exposed to low saline conditions (kathiresan and bingham 2001). hence, lower values of the water content in plants under high saline conditions may be considered as a sign of salt stress. for a mangrove species, the capacity to invade intertidal habitats as well as the position occupied in a species zonation, depends on their salinity tolerance at early growth (macnay 1968; saurez and medina 2005; ye et al. 2005). as an example, a. marina may occupy a wider range of intertidal habitats as its salinity tolerance is higher (dahdouh-guebas et al. 2002). the tolerance of mangrove species to salinity should be taken into consideration in the selection of planting sites and suitable species for mangrove ecosystem rehabilitation. many attempts to restore mangroves fail completely, as they are poorly planned and managed. planting the wrong species in the wrong place is one of the main reasons for many failures in mangrove rehabilitation (lewis 2005). references allen ja, krauss kw, hauff rd. 2003. factors limiting intertidal distribution of the mangrove species xylocarpus granatum. oecologia, 135:110-121. alongi dm. 2002. present state and future of the world’s mangrove forests. environmental conservation 29(3): 331-349. ball, m.c. and pidsley, s.m. 1995. growth responses to salinity in relation to distribution of two mangrove species, sonneratia alba and s. lanceolata, in northern australia. functional ecology 9(1): 77-85. ball mc. 2002. interactive effects of salinity and irradiance on growth: implications for mangrove forest structure along salinity gradient. trees, 16: 126-139 ball mc, cochrane mj and rawson hm. 1997. growth and water use of the mangroves rhizophora apiculata and r. stylosa in response to salinity and humidity under ambient and elevated concentrations of atmospheric co2. plant cell and environment 20: 1158-1166. jayatissa and wickramasinghe: guidance for mangrove replanting ... ruhuna journal of science 1, pp. 47–60, (2006) 59 bandaranayake wm. 1998. traditional and medicinal uses of mangroves. mangraoves and salt marshes 2: 133-148. bandaranayake, wm. 2002. bioactivities, bioactive compounds and chemical constituents of mangrove plants. wetlands ecology and management 10: 421-452 bhosale lj. 1994. propagation techniques for regeneration of mangrove forests-a new asset. journal of non-timber forest products 1(3-4): 119-122. clarke ac. 2005. rebuilding after tsunami: our key challenges. daily mail 02. 01. 2005, london, u.k. clough bf. 1984. growth and salt balance of the mangrove avicennia maraina (forsk.) vierh. and rhizophora stylosa griff. in relation to salinity. australian journal of plant physiology 11: 419-430. dahdouh-guebas f, jayatissa lp, di nitto d, bosire jo, lo seen d and koedam n. 2005. how effective were mangroves as a defence against the recent tsunami? current biology 15(12): 443-447. dahdouh-guebas f, verneirt m, cannicci s, kairo kg, tack jf and koedam n. 2002.an exploratory study on grapsid crab zonation in kenyan mangroves.wetlands ecology and management 10: 179-187, danielsen f, srensen mk, olwig mf, selvam v. parish f, burgess nd, hiraishi t, karunagaran vm, rasmussen ms, hansen lb, quarto a, suryadiputra n. 2005. the asian tsunami: a protective role for coastal vegetation. science, 310: 643-643. downton wjs. 1982. growth and osmotic relations of the mangrove, avicennia marina, as influenced by salinity. australian journal of plant physiology 9: 519528. dahdouh-guebas, f., jayatissa lp, di nitto d, bosire jo, lo seen d and koedam n. 2005. how effective were mangroves as a defense against the recent tsunami? current biology 15: 443-447 hwang yh, chen sc. 2001. effects of ammonium, phosphate, and salinity on growth, gas exchange characteristics, and ionic contents of seedlings of mangrove kandelia candel (l.) druce. botanical bulletin of academy of singapore 42: 131139. jayatissa, l.p., dahdouh-guebas, f. and koedam, n. 2002. a revision of the floral composition and distribution of mangroves in sri lanka. botanical journal of the linnaean society. 138: 29-43 jayatissa lp, hettiarachi s and dahdouh-guebas f. 2006. an attempt to recover economic losses from decadal changes in two lagoon systems of sri lanka through a newly patented mangrove product. environment, development and sustainability (in press) kathiresan k and bingham bl. 2001. biology of mangroves and mangrove ecosystems . advances in marine biology 40: 81-251. jayatissa and wickramasinghe: guidance for mangrove replanting ... 60 ruhuna journal of science 1, pp. 47–60, (2006) lewis rr. 2005. ecological engineering for successful management and restoration of mangrove forests. ecological engineering 24: 403-418 liu pl-f, lynett p, fernando h, jaffe be, fritz h, higman b, morton r, goff j and synolakis c. 2005. observations by the international tsunami survey team in sri lanka. science 308: 1595. macnae w. 1968. a general account of the fauna and flora of the mangrove swamps and forests in the indu-pacific region. advances of marine biology 6:73-270. parida ak and das ab. 2005. salt tolerance and salinity effects on plants: a review. ecotoxicology and environmental safety, 60:324-349. saurez n and medina e. 2005. salinity effect on plant growth and leaf demography of the mangrove, avicennia germinans l. trees 19: 721-727. tomlinson pb 1986. the botany of mangroves. the cambridge university press, london, uk. tuffers a, naidoo g and von willert dj, 2001. low salinities adversely affect photosynthetic performance of the mangrove avicennia marina. wetland ecology and management 9:225-232. williams n. 2005. tsunami insight to mangrove value. current biology 15(3): r73r73. ye y, tam nfy, lu cy and wong sh. 2005. effects of salinity on germination, seedling growth and physiology of three salt secreting mangrove species. aquatic botany 83:193-205. youssef t and saenger p. 1998. photosynthetic gas exchange and accumulation of phytotoxins in mangrove seedlings in response to soil physico-chemical characteristics associated with waterlogging. tree physiology 18(5): 317-324. zar jh. 1984. biostatistical analysis (second edition). prentice-hall inc. a. simon and schuster company englewood cliffs. new jersey, pp 7 acknowledgments we thank professor gamini senanayake, faculty of agriculture, university of ruhuna, for his help in statistical analysis and dr mark huxham, napier university, uk for his critical comments on the manuscript. ruhuna journal of science vol 13 (2): 150-166, december 2022 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v13i2.122 university of ruhuna faculty of science, university of ruhuna sri lanka 150 diversity of ferns and lycophytes in the mt. malambo, southern philippines lilibeth v. rufila1, fulgent p. coritico1,2*, hannah p. lumista1,2, florfe m. acma1,2, noe p. mendez1,2, joevina c. nobleza1,2 and victor b. amoroso1,2 1center for biodiversity research and extension in mindanao (cebrem), central mindanao university, musuan, maramag, 8714 bukidnon, philippines 2department of biology, college of arts and sciences, central mindanao university, musuan, maramag, 8714 bukidnon, philippines *correspondence: cfulgent@cmu.edu.ph, orcid: https://orcid.org/0000-0003-3876-6610 received: 27th april 2022, revised: 20th november 2022, accepted: 20th december 2022 abstract: this research was carried out to study the diversity and assess the conservation status and endemism of ferns and lycophytes in the mt. malambo, datu salumay, southern philippines. repeated transect walks were done with 20 sampling plots delineated on the site. specimens were collected, identified, and assessed for their conservation status and endemism. data gathered were analyzed employing the shannon-weiner index. the inventory revealed 215 species (202 species of ferns and 13 species of lycophytes) belonging to 74 genera and 23 families. polypodiaceae, aspleniaceae, hymenophyllaceae, and pteridaceae were the species-rich families collected. asplenium nidus l., pneumatopteris costata (brackenr.) holttum, and asplenium thunbergii kunze obtained the highest species importance values. mt. malambo has a diversity value of h’=1.83 which is higher compared to other mountains in mindanao. this study reports 20 philippine endemic and 19 threatened species of ferns and lycophytes in the area. of these, one is critically endangered, 11 are endangered, five are vulnerable, and two are other threatened species. the presence of many threatened and endemic species in the area implies that high priority should be addressed in protecting and conserving these species of ferns and lycophytes in mt. malambo. keywords: diversity assessment, inventory, montane forest, pteridophytes, threatened species 1 introduction the ferns and lycophytes (monilophytes) in the philippines consist of ca. 1,100 species distributed in 154 genera and 34 families. species richness is continuously increasing as a result of discoveries of new species and new species records in the philippines (smith et al. 2008, barcelona et al. 2013, amoroso et al. 2019). with the highest https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v13i2.122 https://creativecommons.org/licenses/by-nc/4.0/ https://orcid.org/0000-0003-3876-6610 l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 151 elevation of 1,354 m above sea level (masl) and classified as a montane forest located at 07°29’87’n and 125°15’22’e, mt. malambo is one of the remaining forested mountains in datu salumay, marilog district but continuously faces anthropogenic disturbances, such as the conversion of forestland into residential and mountain resorts, agriculture for high-value-crops, and overharvesting of plants for ornamentals, food, and handicraft materials for livelihood or household consumption. these anthropogenic activities coupled with accessibility, the rate of forest decline in mt. malambo is fast resulting in a loss of biodiversity. an inventory for ferns and lycophytes has not been done in many places in the philippines, particularly on the island of mindanao. therefore, an inventory of ferns and lycophytes is needed, including their diversity and assessment as a basis to protect and conserve the remaining biodiversity of mt. malambo. this study aimed to prepare an inventory, including diversity, conservation status, and assessment of ferns and lycophytes in mt. malambo, datu salumay, southern philippines. 2 material and methods 2.1 study site the research proposal was presented to the local government officials and to the members of the matigsalug-manobo tribal council for elders city of davao, inc. (mamatripcedi) in mt. malambo, marilog district for their information, and a prior informed consent (pic) was obtained from them after. the pic was used as a supporting document to obtain the gratuitous permit from the department of environment and natural resources (denr) as a requirement to conduct the study. the study site is located in the lower montane rainforest having clay and loam subrates at an elevation of 1,345 masl (acma et al. 2021). 2.2 species inventory and abundance the inventory of ferns and lycophytes in mt. malambo, datu salumay, marilog district in davao, philippines (figure 1, 2) was conducted through repeated transect walks and the establishment of sampling plots. twenty sampling plots of 20 x 20 m (figure 1c, red dots) were established inside the forest of mt. malambo. the total number of individuals in all plots was counted to determine their abundance. 2.3 specimen collection, processing and identification at least four fertile fronds were collected for each species using pruning shears and a trimming cutter. small ferns and lycophytes (ca. 4–20 in.) were collected by uprooting l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 152 the whole plant, removing the earthy matter, and pressing the specimens. for the epiphytic ferns and lycophytes, the whole plants were detached to the trunk and the earthy matter or barks were removed. for tree ferns, each entire frond was collected and cut into five parts: leaf apex, middle pinna, lower pinna, basal pinna and stipe. all specimens were processed using the wet method (hodge 1974). fig 1. study site. a) map of the philippines, b) mt. malambo in the island of mindanao, c) close-up view of mt. malambo and sampling area, d) sampling plots (red marks) and transect walks (white line). identification of the specimens was done using fresh materials from the field and compared to some floras, monographs, field guides, scientific journal articles, such as copeland (1958–1961), co’s digital flora of the philippines of pelser et al. (2011 onwards), and digitized plant specimens available in the global plants on journal l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 153 storage (jstor). the classification system used was based on the pteridophyte phylogeny group (ppg i 2016). these specimens were then processed and included in the prepared voucher specimens. all specimens were accessioned and deposited at the central mindanao university. fig 2. panoramic view of mt. malambo with montane vegetation. 2.4 assessment of conservation status and endemism assessment of conservation status and endemism of the species was based on the recently published book of philippine threatened plants (fernando et al., 2022) which follows the criteria of the international union for the conservation of nature (iucn). this information is important because it serves as a basis for governmental agencies like the protected area management board (pamb), department of environment and natural resources (denr) and local government units (lgus) which formulate policies and guidelines for monitoring and protecting the threatened and endemic species. 2.5 species importance value (siv) and diversity values species richness of ferns and lycophytes was estimated by determining the number of species. calculation for frequency, relative frequency, density, relative density and importance value index (ivi) were derived from curtis & mcintosh (1951). siv or ni = rd + rf + rdom where, rd is relative density, rf is relative frequency, and rdom is relative dominance. the diversity values were computed using the shannon-weiner index (h’) (shannon and weiner 1963) by, l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 154 h = -∑ [ni /n] ln [ni/n] where, ni = number of individuals of each species, n = total number of individuals, ln = the natural log of the number. 3 results and discussion 3.1 species richness repeated transect walks and sampling plots in mt. malambo revealed 215 species from 23 families and 74 genera of ferns and lycophytes (table 1; figure 3). table 1. total number of families, genera and species of ferns and lycophytes in mt. malambo, datu salumay, marilog district. family number of genera number of species lycophytes lycopodiaceae 2 5 selaginellaceae 1 8 ferns aspleniaceae 1 21 athyriaceae 4 18 blechnaceae 1 1 cyatheaceae 2 6 davalliaceae 2 8 dennstaedtiaceae 4 5 dicksoniaceae 2 2 dryopteridaceae 5 15 gleicheniaceae 3 5 hymenophyllaceae 6 19 hypodematiaceae 1 1 lindsaeaceae 3 10 marattiaceae 1 1 nephrolepidaceae 1 5 oleandraceae 1 3 ophioglossaceae 3 3 polypodiaceae 16 38 pteridaceae 5 17 tectariaceae 4 9 thelypteridaceae 7 15 total 74 215 l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 155 of these, 202 species are ferns and 13 are lycophytes. the species richness is ca. 20.8% of the total number of ferns and lycophytes in the philippines and ca. 34.0% of the total number recorded in mindanao island (amoroso et al., 2011). polypodiaceae (32 species) obtained the highest number of species, followed by aspleniaceae (21 species), hymenophyllaceae (19 species), athyriaceae (18 species), pteridaceae (17 species), and dryopteridaceae and thelypteridaceae (both with 15 species). these families have the highest number of species in the entire country (salgado 1990). data on species richness in mt. malambo was comparatively higher compared to the previous works of amoroso et al. (1996) in the marilog forest which documented 183 species of ferns and lycophytes and coritico et al. (2020) in mt. tago range in bukidnon with 203 species, silverio et al. 2021 in mt. sinaka, north cotabato with 163 species, amoroso et al. (2016) in mt. hamiguitan range wildlife sanctuary in davao oriental with 152 species (14.1% compared to the total number of species in the philippines), gonzales (2000) in mt. matutum with 188 species (17.1%), and silverio (2014) in mt. apo with 106 species (9.6%). the high species richness in mt. malambo may be due to its high elevation intact forest and the extensive inventory of ferns and lycophytes in the said area (sosanika et al. 2022). however, the species richness of ferns and lycophytes in mt. malambo is relatively lower than the number of species found in mt. kitanglad, bukidnon with 439 species (39.9%) (amoroso et al. 2011) and mt. malindang in misamis occidental with 371 species (33.7%) (rufila 2016). the high species richness in these mountain forests is due to the presence of diverse habitats, such as the humid lower montane forest. mt. malindang range and mt. kitanglad possess several vegetation types, viz., mossy forest, montane forest, dipterocarp forest, almaciga forest, mixed dipterocarp forest, lowland dipterocarp forest, plantation forest and agroecosystem (amoroso et al. 2006; amoroso et al. 2012). several factors may affect the species richness of local montane forests in the philippines, including the size of the area sampled, climate conditions, soil type, and geographic location (kessler 2010). species richness is also affected by some anthropogenic disturbances, such as the conversion of forests to agricultural or industrial lands and pollution (amoroso et al. 2016). these factors are more likely to affect the variability of species richness reported in the studies on mountains in mindanao, philippines. table 2. species checklist of ferns and lycophytes in mt. malambo, datu salumay, marilog district, philippines. ferns aspleniaceae 1. asplenium affine sw. 2. a. apoense copel.* 3. a. caudatum g forst. 4. a. crinicaule hance 5. a. cymbifolium christ l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 156 table 2 continued. aspleniaceae 6. a. excisum c.presl 7. a. lobulatum mett. 8. a. longissimum blume. 9. a. nigrescens blume 10. a. nidus l. 11. a. normale d. don 12. a. persicifolium j.sm. ex mett. 13. a. phyllitidis d. don. 14. a. polyodon g. forst. 15. a. rhizophyllum l. 16. a. subnormale copel. 17. a. tenerum g forst. 18. a. thunbergii kunze 19. a. unilaterale lam. 20. a. vittaeforme cav. 21. asplenium sp. blechnaceae 22. blechnopsis orientalis (l.) c.presl athyriaceae 23. athyrium brevipinnulum copel. 24. a. elmeri copel. 25. a. puncticaule (blume) moore. 26. cornopteris decurrenti-alata (hook.) nakai. 27. deparia confluens (kunze) m.kato. 28. d. lancea (thunb.) fraser-jenkins 29. d. petersenii (kunze) m.kato. 30. diplazium altum (copel.) c.chr. 31. d. cordifolium blume 32. d. costulisorum (copel.) c.chr. 33. d. davaoense copel.* 34. d. dilatatum blume 35. d. esculentum (retz.) sw. 36. d. forbesii (baker) c.chr. 37. d. geophilum (copel.) alderw. 38. d. melanopodium fée* 39. d. oligosorum copel. 40. d. pallidum (blume) moore 41. d. sorzogonense (c.presl) c.presl cyatheaceae 43. alsophila commutata mett. 44. a. fuliginosa christ* 45. a. loheri christ 46. sphaeropteris elmeri (copel.) r.m.tryon 47. s. glauca (blume) r.m. tryon 48. s. tripinnata (copel) r.m. tryon davalliaceae 49. davallia denticulata (nl burm.) mett., ex kuhn 50. d. hymenophylloides (blume) kuhn 51. d. embolostegia copel. l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 157 table 2. continued. davalliaceae 52. d. pubescens c.w.chen 53. d. repens (l.f.) kuhn/ humata repens (l.f.) diels 54. d. solida (forst.) sw. 55. d. trichomanoides blume. 56. d. wagneriana copel. dennstaedtiaceae 57. dennstaedtia hooveri christ. 58. histiopteris incisa (thunb.) j.sm. 59. pteridium aquilinum (l.) kuhn 60. microlepia protracta copel.* 61. m. speluncae (l.) t.moore. dicksoniaceae 62. calochlaena javanica (blume) m.d.turner & r.a.white 63. dicksonia mollis holttum dryopteridaceae 64. arachniodes aristata (forster) tindale 65. bolbitis heteroclita (c.presl) ching. 66. dryopteris formosana (christ) c.chr. 67. d. nodosa (c.presl) li bing zhang 68. d. pseudocaenopteris (kunze) li bing zhang 69. d. purpurascens (blume) christ. 70. d. sparsa (don) kuntze. 71. dryopteris sp. 72. elaphoglossum blumeanum (fée) j.sm. 73. e. callifolium (bl.) moore. 74. e. latifolium (sw.) j. sm. 75. e. luzonicum copel.* 76. e. petiolatum (sw.) urb. 77. polystichum elmeri copel.* 78. p. nudum copel.* gleicheniaceae 79. dicranopteris curranii copel. 80. d. linearis (burm.) underw. 81. diplopterygium longissimum (blume) nakai 82. sticherus loheri (christ) copel.* 83. s. truncatus (willd.) nakai hymenophyllaceae 84. abrodictyum setaceum (bosch) ebihara & k.iwats 85. a. obscurum (blume) ebihara & k.iwats. 86. callistopteris apiifolia (c presl) copel. 87. cephalomanes extravagans copel. 88. crepidomanes bipunctatum (poir.) copel. 89. c. brevipes (c. presl) copel. 90. hymenophyllum acanthoides (bosch) rosenst. 91. h. angulosum christ 92. h. badium hook. & grev. 93. h. denticulatum sw. 94. h. digitatum (sw.) fosberg 95. h. emarginatum sw. l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 158 table 2. continued. hymenophyllaceae 96. h. fimbriatum j.sm. 97. h. nitidulum (bosch) ebihara & k.iwats. 98. h. pallidum (blume) ebihara & k. iwats. 99. h. paniculiflorum c.presl 100. h. polyanthos (sw.) sw. 101. h. productum kunze 102. vandenboschia maxima (blume) copel. hypodematiaceae 103. leucostegia truncata (d.don) fraser-jenk lindsaeaceae 104. lindsaea adiantoides j.sm. 105. l. apoensis copel.* 106. l. fissa copel.* 107. l. lucida blume 108. l. pulchella (j.sm.) mett. ex kuhn. 109. l. repens (bory) thwaites 110. l. rigida j.sm. ex hook. 111. odontosoria chinensis (l.) j.sm. 112. tapeinidium luzonicum (hook.) kramer 113. t. pinnatum (cav.) c.chr. marattiaceae 114. angiopteris evecta (g.forst.) hoffm. nephrolepidaceae 115. nephrolepis bisserrata (sw.) schott. 116. n. cordifolia (l.) presl. 117. n. exaltata (l.) schott. 118. n. falcata (cav.) c. chr. 119. n. hirsutula (g forst.) c. presl oleandraceae 120. oleandra maquilingensis copel.* 121. o. neriiformis cav. 122. o. nitida (copel.) copel.* ophioglossaceae 123. botrychium daucifolium wall. ex hook. & grev. 124. ophioderma pendulum c.presl 125. ophioglossum ramosii copel.* 126. o. reticulatum l. polypodiaceae 127. aglaomorpha cornucopia (copel.) roos* 128. a. descensa (copel.) hovenkamp & s.linds. 129. a. heraclea (kunze) copel. 130. a. splendens (hook. & bauer) copel. 131. a. sparsisora (desv.) hovenkamp & s. linds. 132. a. quercifolia (l.) hovenkamp & s. linds. 133. calymmodon gracilis (fée) copel. 134. dasygrammitis mollicoma (nees & blume) parris 135. goniophlebium persicifolium (desv.) bedd. 136. g. pseudoconnatum (copel.) copel.* 137. g. subauriculatum (blume) c.presl l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 159 table 2. continued. polypodiaceae 138. lecanopteris deparioides (cesati) baker 139. lemmaphyllum accedens (blume) donk 140. lepisorus mucronatus (fée) li wang 141. l. platyrhynchos (kunze) li wang 142. l. spicatus (l.f.) li wang 143. leptochilus sp. 144. loxogramme scolopendrioides (gaudich.) c.v.morton 145. microsorum commutatum (blume) copel. 146. m. congratifolium (alderw.) holttum 147. m. insigne (blume) copel. 148. m. monstrosum (copel.) copel. 149. m. musifolium (blume) copel. 150. m. phanerophlebium (copel.) copel. 151. oreogrammitis reinwardtii (blume) parris 152. platycerium grande (fée) kunze* 153. prosaptia celebica (blume) tagawa & k iwats 154. p. contigua (forst.) c.presl 155. p. nutans (blume) mett. 156. p. obliquata (blume) mett. 157. pyrrosia adnascens (sw.) ching. 158. p. sphaerotricha (mett.) ching. 159. selliguea albidosquamata (blume) parris 160. s. taeniata (sw.) parris 161. s. triloba (houtt.) m.g.price 162. scleroglossum pusillum (blume) alderw. pteridaceae 163. antrophyum plantagineum (cav.) kaulf. 164. haplopteris alternans (copel.) s.linds. & c.w.chen 165. h. ensiformis (sw.) e.h. crane 166. h. elongata (sw.) e.h. crane 167. h. scolopendrina (bory) e.h. crane 168. pteris blumeana j agardh. 169. p. longipinnula wall. ex j agardh. 170. p. mertensioides willd. 171. p. oppositipinnata fée. 172. p. pacifica hieron. 173. p. schlechteri brause. 174. p. tripartita sw. 175. p. vittata l. 176. syngramma alismifolia (c.presl) j.sm. 177. s. wallichii (hook.) bedd. 178. vaginularia paradoxa (fee) mett. 179. v. trichoidea (j.sm.) fée tectariaceae 180. ctenitis submarginalis (langsdorff & fischer) ching 181. pleocnemia irregularis (c.presl) holttum 182. p. leuzeana (gaudich.) c.presl 183. p. macrodonta (c.presl) holttum 184. pteridrys syrmatica (willd.) c.chr. & ching. l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 160 table 2. continued. tectariaceae 185. tectaria angulata (willd.) copel. 186. t. athyriosora m.g.price. 187. t. decurrens (c.presl) copel. 188. t. griffithii (baker) c.chr. thelypteridaceae 189. chingia ferox (blume) holttum. 190. christella dentata (forssk.) brownsey & jermy. 191. c. hispidula (decne.) holttum. 192. c. parasitica (l.) lév. 193. coryphopteris pubirachis (baker) holttum var. philippinensis holttum* 194. plesioneuron savaiense (baker) holttum 195. pneumatopteris costata (brackenr.) holttum 196. p. laevis (mett.) holttum* 197. p. nitidula (c.presl) holttum* 198. pronephrium amphitrichum holttum* 199. p. clemensiae (copel.) holttum* 200. sphaerostephanos heterocarpus (blume) holttum 201. s. polycarpus (blume) copel. 202. s. unitus (l.) holttum lycophytes lycopodiaceae 203. phlegmariurus nummularifolius (blume) ching 204. p. pinifolius (trevis.) kiew 205. p. salvinioides (herter) ching 206. p. squarrosus (g.forst.) á.löve & d.löve 207. palhinhaea cernua (l.) vasc. & franco selaginellaceae 208. selaginella aristata spring, bull. 209. s. cupressina (willd.) spring 210. s. delicatula (desv.) alston 211. s. eschscholzii hieron. 212. s. flagellifera hieron 212. s. intermedia (blume) spring 214. s. involvens (sw.) spring 215. s. ornata (hook. & grev.) spring legend: asterisk (*) after the scientific names indicates that the particular species is endemic to the philippines. 3.2 species importance values (siv) the five species of ferns that obtained the highest siv include asplenium nidus l., a. thunbergii kunze, lindsaea fissa copel., davallia hymenophylloides (blume) kuhn, oleandra neriiformis cav., and hymenophyllum polyanthos (sw.) sw. (table 3). among these species, a. nidus and d. hymenophylloides are the most frequently l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 161 collected species. these species ultimately play an important role in regulating the ecological stability of the forest ecosystem. some ferns with higher siv conform to the reports of amoroso et al. (2015) and amoroso et al. (2018) revealing nephrolepis hirsutula (g forst.) c. presl and asplenium spp. as the species with high siv in the montane forest of mt. kitanglad, bukidnon and mt. apo, north cotabato. fig 3. some ferns and lycophytes in mt. malambo, datu salumay, southern philippines. a) lindsaea apoensis copel., b) lindsaea fissa copel., c) lindsaea pulchella (j.sm.) mett. ex kuhn., d) sphaeropteris glauca (blume) r.m.tryon, e) angiopteris palmiformis (cav.) c.chr., f) oleandra neriiformis cav., g) botrychium daucifolium wall. ex hook. & grev., h) ophioderma pendula c.presl, i) phlegmariurus salvinioides (herter) ching, j) phlegmariurus squarrosus (g.forst.) á.löve & d.löve, k) palhinhaea cernua (l.) vasc. & franco, l) selaginella involvens (sw.) spring. l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 162 3.3 diversity value mt. malambo is classified as a montane forest based on the apparent change in forest structure and floristic composition. the ferns and lycophytes in the area have a diversity value of h’=1.83 which closely resembles that of the montane forests of mt. malindang with h’=1.80 (rufila 2016). also, the montane forest of mt. kitanglad (amoroso et al. 2011) and mt. apo (silverio 2014) showed the highest diversity values for pteridophytes. the diversity value is relatively higher compared to balinsasayao twin lakes natural park in negros oriental with h’=1.41 (amoroso et al. 2018) and mt. hamiguitan in davao oriental with h’=1.01 (amoroso et al. 2015). this suggests that the montane forest is the most diverse vegetation for the ferns and lycophytes. fern and lycophyte diversity is much higher in mid-elevation of the mountain, usually in the montane forests. this pattern of distribution is affected by different abiotic factors, such as climate and edaphic features. adean tropical forests also revealed that elevational richness patterns for ferns and lycophytes are symmetrically hump-shaped and overall richness is virtually equal along most of the tropical latitudinal gradient (salazar et al. 2013). mehltreter et al. (2018) also mentioned that fern diversity is highest in the mid-elevation with utmost richness and diversity and lesser in the lower and higher elevations. in addition, the peak of fern species richness at mid-elevations has often been interpreted as reflecting the ambient conditions of a balanced climate without extremes, such as drought at low elevations and frost at high elevations (bhattarai et al. 2004; krömer 2007; kluge et al. 2008; kessler et al. 2011). in addition, it has often been linked to an optimal combination of high humidity, rainfall, and moderate temperatures at mid-elevations (lauer et al. 1996). table 3. ferns and lycophytes with highest species importance value (siv) in mt. malambo, datu salumay, southern, philippines. species (%) 1. asplenium nidus l. 17.98 2. pneumatopteris costata (brackenr.) holttum 11.54 3. asplenium thunbergii kunze 11.53 4. davallia hymenophylloides (blume) kuhn 11.21 5. lindsaea fissa copel. 10.43 6. oleandra neriiformis cav. 9.94 7. dryopteris nodosa (c.presl) li bing zhang 9.58 8. polystichum elmeri copel. 8.90 9. nephrolepis hirsutula (g forst.) c. presl 7.94 10. selaginella involvens (sw.) spring 7.76 3.4 endemism and conservation status a total of 20 philippine endemic species of ferns are recorded in mt. malambo. these species include a. apoense copel., d. davaoense copel., d. melanopodium fée, l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 163 alsophila fuliginosa christ, microlepia protracta copel., elaphoglossum luzonicum copel., polystichum elmeri copel., p. nudum copel., sticherus loheri (christ) copel., lindsaea apoensis copel., l. fissa copel., oleandra maquilingensis copel., o. nitida (copel.) copel. ophioglossum ramosii copel., aglaomorpha cornucopia (copel.) roos, goniophlebium pseudoconnatum (copel.) copel., coryphopteris pubirachis (baker) holttum, pneumatopteris laevis (mett.) holttum, p. nitidula (c. presl) holttum, pronephrium amphitrichum holttum and p. clemensiae (copel.) holttum (table 2). among the 215 species of ferns and lycophytes, 19 species were recorded as threatened. of these threatened species, one is critically endangered, 11 are endangered, five are vulnerable and two are other threatened species (table 4). furthermore, of the 24 threatened species, three are philippine endemics namely: a. cornucopia, m. protracta and p. grande. local assessment revealed that there are 37 species classified as very abundant, 21 are abundant, and 76 are rare. local assessment of the mindanao endemic species is rare except for ophioglossum ramosii. table 4. conservation status and endemism of ferns and lycophytes in mt. malambo, datu salumay, marilog district. species conservation status 1. aglaomorpha cornucopia (copel.) m.c. roos vulnerable 2. aglaomorpha heraclea (kunze) copel vulnerable 3. aglaomorpha splendens (hook. & bauer) copel. vulnerable 4. alsophila fuliginosa (h. christ) copel. endangered 5. asplenium vittiforme cav. vulnerable 6. davallia embolostegia copel. other threatened species 7. davallia solida (forst.) sw. other threatened species 8. dicksonia mollis holttum endangered 9. diplazium costulisorum (copel.) c.chr. endangered 10. lecanopteris deparioides (cesati) baker endangered 11. lepisorus platyrhynchos (kunze) li wang endangered 12. microlepia protracta copel. endangered 13. ophioderma pendula c.presl endangered 14. phlegmariurus salvinioides (herter) ching endangered 15. phlegmariurus squarrosus (g.forst.) á.löve & d.löve endangered 16. platycerium grande (fée) kunze critically endangered 17. polystichum nudum copel. endangered 18. sphaeropteris elmeri (copel.) r.m. tryon vulnerable 19. sphaeropteris glauca (blume) r.m. tryon endangered ten (10) endangered species were recorded inside the sampling plots, viz., d. mollis, l. platyrhynchos, a. fuliginosa, s. glauca, p. nudum, o. pendulum, d. costulisorum, p. squarrosa, and p. salvinioides. these threatened species were also documented on different protected areas in mindanao island (coritico & amoroso 2020; amoroso et al. 2018). other endangered species were seen along the trail from the foot of the mountain to the peak of mt. malambo. furthermore, p. grande, a highly priced l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 164 ornamental plant, was recorded while doing the repeated transect walks. p. grande is a mindanao island endemic species and may become extinct in the wild if no conservation measures are adopted in the area. alombro (1999) documented the endangered tmesipteris zamorarum gruezo and amoroso in the area. however, present exploration revealed the absence of this endangered fern species together with phlegmariurus banayanicus (herter) a.r.field & bostck. the absence of t. zamorarum could be due to the loss of its only habitat, which are the tree ferns (alsophila spp. and sphaeropteris spp.). tree ferns are widely collected and sold commercially as a potting medium to grow other plants and as posts. mt. malambo and its vicinity show a high number of threatened and endemic species; therefore, high priority should be given to the protection of these species by the local stakeholders. regulation on the collection and the need to propagate the economically important species should be addressed. further, data from this research can be used as a sound basis in the formulation of policies that are effective for the conservation and protection of the whole mt. malambo. 4 conclusions and recommendations mt. malambo in barangay datu salumay, marilog district is home to 215 species of ferns and lycophytes belonging to 74 genera and 23 families. the species richness of this group of flora is ca. 23% of the total number of species in the philippines and ca. 39% compared with the total number of species in mindanao. mt. malambo has a diversity value of h’=1.83, which is higher compared to the other mountains in mindanao island. twenty endemic species and 24 species are threatened. of these, one species is critically endangered, 12 species are endangered, seven species are vulnerable and four are other threatened species. the high species richness and high diversity of ferns and lycophytes in mt. malambo represent a significant floral resource of the philippines but should be protected and conserved by the stakeholders since some species are placed in the threatened category. acknowledgements this research was supported by ched dare to (commission on higher education discovery-applied research and extension for trans/inter-disciplinary opportunities) research grant. the gratuitous permit from denr region xi and logistic support from the marilog district stakeholders, matigsalug-manobo tribal people council of elders davao, inc. (mamatripcedi), and the central mindanao university (cmu) are gratefully acknowledged. comments from anonymous reviewers are acknowledged. l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 165 references acma fm, mendez np, lagunday ne, amoroso vb. 2021. diversity of understory flowering plants in the forest patches of marilog district, philippines. journal of threatened taxa 13(5): 18247–18256. doi: 10.11609/jott.6278.13.5.18247-18256. alombro nc. 1999. pteridophyte flora of mt. malambo in barangay datu salumay, marilog district, davao city. [ms thesis]. musuan, bukidnon: central mindanao university, philippines. amoroso vb, acma fm, pava hp. 1996. diversity status and ecology of pteridophytes in three forests in mindanao, philippines. in: camus jm and johns rj, editors. pteridology in perspective. royal botanic gardens, kew. amoroso vb, arances jb, gorne nd, ruba rp, rufila lv, opiso gs, demetillo mt, noval kg, van den berg a, stortelder a, kessler pja, gruezo w, lagunzad da, tan bc, co ll, aranico ec, roscom ba, alava cg. 2006. plant diversity in the northern landscape of mt. malindang range and environs, misamis occidental, philippines. college laguna: seamco searca and quezon city: pdm press, philippines. 116 p. amoroso vb, obsioma ld, arlalejo jb, aspiras ra, capili dp, polizon jja, sumile eb. 2009. inventory and conservation of endangered, endemic and economically important flora of hamiguitan range, southern philippines. blumea 54(1–3): 71–76. doi: 10.376/000651909x474113. amoroso vb, laraga sh, calzada bv. 2011. diversity and assessment of plants in mt. kitanglad range natural park, bukidnon, southern philippines. gardens bulletin singapore 63(1/2): 219–239. amoroso vb, amoroso cb, coritico fp. 2012 diversity and status of plants in three mountain ecosystems in southern mindanao, philippines. asian journal of biodiversity 3(1): 50–73. doi: 10.7828/ajob.v3i1.83. amoroso vb, chiou wl, liu yc, callado jrc, briones nng, coritico fp. 2015. threatened endemic, rare, economic species of philippine pteridophytes. central mindanao university, musuan, bukidnon, philippines: instructional materials and development center, 173 pp. amoroso vb, coritico fp, fritsch pw. 2016. species richness and conservation status of ferns and lycophytes in mt. hamiguitan range wildlife sanctuary, davao oriental, philippines. philippine journal of science 145: 127–137. amoroso vb, coritico fp, bucol aa, aribal lg, alcala el. 2018. biodiversity inventory, assessment and monitoring in balinsasayao twin-lakes natural park (btlnp). technical report. central mindanao university. philippines. amoroso vb, acma fm, coritico fp, gorme fs, lagunday ne, salolog mcs, colong rd. 2018. floral diversity assessment of the buffer zones and vicinity of the mt. hamiguitan range wildlife sanctuary (mhrws), davao oriental: basis for inclusion to protected area zone. philippine journal of systematic biology 12(2): 36–51. barcelona jf, nickrent dl, lafrankie jv, callado jrc, pelser pb. 2013. co's digital flora of the philippines: plant identification and conservation through cybertaxonomy. philippine journal of science 142(special issue): 57–67. bhattarai kr, vetaas or, grytnes ja. 2004. fern species richness along a central himalayan elevational gradient, nepal. journal of biogeography 31(3): 389–400. doi: 10.1046/j.0305-0270.2003.01013.x. copeland eb. 1958–1961. fern flora of the philippines. vols. 3. manila: manila bureau of printing, 555 pp. coritico fp, amoroso vb, acma fm, cariño yll, fritsch pw. 2020. ferns and lycophytes of mt. tago range, bukidnon, southern philippines: species richness, distribution and conservation status. philippine journal og science 149(3-a): 773–790. coritico fp and amoroso vb. 2020. threatened lycophytes and ferns in four protected areas of mindanao, philippines. nature conservation research 5(44):78–88. curtis jt, mcintosh rp. 1951. the interrelations of certain analytic and synthetic phytosociological characters. ecology 31: 434–455. fernando es, amoroso vb, tandang dn, banag ci, lapis ab, gruezo ws, coritico fp, pollisco fa jr., palijon am, ching md, lansigan fp, garcia-francisco c, lita nm, tagtag am, de leon jl. 2022. l.v. rufila et al. diversity of ferns and lycophytes in the mt. malambo ruhuna journal of science vol 13 (2): 150-166, december 2022 166 threatened plants of the philippines. department of environment and natural resources. biodiversity management bureau. 149p. gonzales aa. 2000. pteridophytes of mt. matutum, south cotabato. [dissertation]. musuan, bukidnon: central mindanao university, philippines. hodge wh. 1947. the use of alcohol in plant collecting. rhodora 49: 207–210. kessler m. 2010. biogeography of ferns. in: fern ecology. mehltreter k, walker lr, sharpe jm eds. cambridge, united kingdom: cambridge university press, 22–60 pp. kessler m, kluge j, hemp a, ohlemüller r. 2011. a global comparative analysis of elevational species richness patterns of ferns. global ecology and biogeography 20: 868–880. doi: 10.1111/j.14668238.2011.00653.x. kluge j, bach k, kessler m. 2008. elevational distribution and zonation of tropical pteridophyte assemblages in costa rica. basic and applied ecology 9(1): 35–43. doi: 10.1016/j.baae.2006.11.004. krömer t, kessler m, gradstein sr, acebey a. 2005. diversity patterns of vascular epiphytes along an altitudinal gradient in the andes. journal of biogeography 32(10): 1799–1809. doi: 10.1111/j.13652699.2005.01318.x. lauer w, rafiqpoor md, frankenberg p. 1996. die klimate der erde. eine klassifikation auf ökophysiologischer grundlage der realen vegetation. erdkunde 50: 275–300. mehltreter k, walker lr, sharpe jm. 2010 eds. fern ecology. cambridge: cambridge university press, 460 p. pelser pb, barcelona jf, nickrent dl. (eds.). 2011 onwards. co's digital flora of the philippines. accessed at: www.philippineplants.org. ppg i. 2016. a community-derived classification for extant lycophytes and ferns. journal of systematics and evolution 54(6): 563–603. rufila lv. 2016. diversity and distribution of pteridophytes in mt. malindang range natural park, misamis occidental philippines. [dissertation]. musuan, bukidnon: central mindanao university, philippines. salazar l, homeiera j, kessler m, abrahamczykb s, lehnert m, krömer t, kluge j. 2013. diversity patterns of ferns along elevational gradients in andean tropical forests. plant ecology & diversity 1–12. doi: 10.1080/17550874.2013.843036. salgado ea. 1990. a checklist of philippine ferns. philippine journal of science 119: 107–148. shannon ce, weiner w. 1963. the mathematical theory of communication. university of illinois press. urbana, 117 p. silverio fd. 2014. diversity of trees and pteridophytes across vegetation types in mount apo, mindanao. [dissertation]. musuan, bukidnon: central mindanao university, philippines. silverio fd, cadali jan, malik bsa, coritico fp, amoroso vb. 2021. checklist and conservation status of ferns and lycophytes of mt. sinaka, arakan, north cotabato, southern philippines. the thailand natural history museum journal 15(2): 161–175. smith ar, pryer km, schuettpelze e, korall p, schneider h, wolf pg. 2008. fern classification. in: biology and evolution of ferns and lycophytes. ranker ta, haufler ch eds. cambridge, united kingdom: cambridge university press, 417–446 pp. sosanika gl, sule b, fazang k, kiaprani r, damas k, hitofumi a, turia r, dargusch paul, novotny v. 2022. fern species richness and diversity in the forest ecosystems of papua new guinea: a case study along an elevational gradient. case studies in the environment 6(1). doi: 10.1525/cse.2022.1696511. http://www.philippineplants.org/ rjs-2007-erratum-pp-115.dvi ruhuna journal of science vol. 2, september 2007, pp. 115– http://www.ruh.ac.lk/rjs/ i s s n 1800-279x © 2007 faculty of science university of ruhuna. erratum the journal wishes to point out an error in the article “use of antagonistic rhizobacteria and non-pathogenic fusarium oxysporum for suppression of fusarium root and stem rot of cucumis setvins caused by fusarium oxysporum f. sp. cucumerinum” by s. abeysinghe of department of botany, university of ruhuna, matara, sri lanka. (saman@bot.ruh.ac.lk) appeared in the ruhuna journal of science volume 1, pages 24-31. the above title should be corrected as follows: “use of antagonistic rhizobacteria and non-pathogenic fusarium oxysporum for suppression of fusarium root and stem rot of cucumis sativus caused by fusarium oxysporum f. sp. cucumerinum” the editorial board of the ruhuna journal of science apologizes to dr. s. abeysinghe and to the readers for any inconvenience. 114  © 2008 faculty of science university of ruhuna ruhuna journal of science vol. 2, september 2008, pp. 18-25 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x abstract. we model the price prediction in sri lankan stock market using ising model and some recent developments in statistical physics techniques. in contrast to usual agent-models, the influence does not flow inward from the surrounding neighbors to the centre, but spreads outward from the center to the neighbors. monte carlo simulations were used to study this problem. the analysis was based on all share price index, milanka price index in colombo stock exchange and simulated price process. the monthly and daily influences of the above indices to the sri lankan economy were also investigated. the model thus describes the spread of opinions traders. keywords: stock market, price formation, spin model 1. introduction a large set of economic models can be mapped to various versions of the ising model to account for social influence in individual decisions. human beings are not spins, they can learn, adapt the nature and strengthen their interactions with others based on the past experience. we use ising model to study the price prediction in sri lankan financial market. this is a mathematical model where state is described by the accumulation of binary variables and has been frequently used for discovering universality behind various social and economic systems (stauffer, 2000, wedagedara 2006). the spins are interpreted as market participant’s attitude (weron 2000, 2003). the return of 18 modeling asset price dynamics ranasinghe p. k. c. malmini department of mathematics, university of sri jayewardenepura, nugegoda, sri lanka correspondence: malmini@sjp.ac.lk 1. introduction  © 2008 faculty of science university of ruhuna ruhuna journal of science vol. 2, september 2008, pp. 8-17 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x abstract. a semi-empirical statistical physics model for the dynamical behavior of stock prices in sri lankan financial market was analyzed. in this model, the time evolution of a collective set of stock prices was analyzed using the hamiltonian of a nearest neighbor ising model. monte carlo simulations were performed and resultant stylized features of the corresponding system were discussed. keywords: spin model, time series, stock market, stylized features. 1. introduction there has been much recent work applying physics concepts and methods to the study of financial time series( mantegna and stanley 2000, fischer 1991, mandelbolt 1982). in particular, several empirical studies have determined the scale-invariant behavior of the distribution of price changes, the long-range correlation in the absolute value of returns, stock market crashes( mantegna 2000, gopakrishnan, meyer 1999, wedagedara 2006). in this paper we mainly focus on the time evolution of a collective set of stock prices(maskawa 2000) in sri lankan stock market as a case study. we consider the statistics of the sign of price changes. the time series of price changes are coded into a sequence of up and down ising spins. table 1 shows a sample of the coding procedure for virtual data of all share price index and of milanka price index in colombo stock market respectively. let iy be the price of the i th sample. the i th price change is defined as 1− −= i y i y i z and i th code is 8 a spin model for the dynamical behavior of the financial market ranasinghe p. k. c. malmini department of mathematics, university of sri jayewardenepura, nugegoda, sri lanka correspondence: malmini@sjp.ac.lk http://www.ruh.ac.lk/rjs/rjs.html 1. introduction microsoft word rjs-vol-2-lakmini-rupika-kalidasa-1.2.doc.doc © 2007 faculty of science university of ruhuna ruhuna journal of science vol. 2, september 2007, pp. 111-114 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x 111 abstract. the suitability of gliricidia (gliricidia sepium) leaf extract to replace 100%, 75 %, 50 % and 25 % of albert solution in growth of tomato (var. thilina) was tested. the research was conducted under the protected plant house conditions at faculty of agriculture, university of ruhuna. polybags filled with the medium of 1:1 coir dust and sand and the solutions prepared by mixing different composition of gliricidia leaf extract and albert solution were used for the experiment. 100 g ground gliricidia leaves were mixed in 1 l of water and filtered to prepare nutrient solutions. growth performances of plants were recorded 12 weeks post treatments. significantly (p<0.05) lower plant height was observed in plants treated only with gliricidia leaf extract whereas, significantly higher leaf number was observed in plants treated only with albert solution (control). performance of plant growth in respect to time taken for flowering also showed significant differences among treatments where, plants in control took shortest time but the time gap was only 6 days. plants treated with 25% gliricidia extract and 75% albert solution showed significantly higher yield. according to these results it could be concluded that the growth performances of tomato are reduced with increasing level of gliricidia leaf extract but with considering the yield gliricidia leaf extract could replace about 25% of albert solution giving an additional benefits to reduce cost of production. keywords. growth performances, gliricidia leaf extract and tomato introduction protected agriculture is the modification of the natural environment to achieve optimum plant growth. modification can be made to both aerial and root environments to increase the crop yield, extending the growing season, protect form the hazardous and permit plant growth during the any time of the year (molter and a preliminary study on the effects of gliricidia leaf extract on growth performances of tomato (lycopersicon esculentum) department of crop science, faculty of agriculture, university of ruhuna, mapalana, kamburupitiya. correspondence: lakmini077@yahoo.com w.g.d. lakmini, s.t.j. rupika, p.e.kaliadasa 112 lakmini, rupika and kalidasa: m o rp h ol og i c a l he t e r o ge ne i t y . . . r u h u n a j o u r n a l o f sc i e n c e 2 , p p . 111-114 (2 0 07 ) jensen, 1995). albert solution is the common nutrient solution used for all growing crops in protective agriculture. but the cost of production is higher due to higher prices of the albert solution. gliricidia sepium is a leguminous tree and used in many tropical and sub-tropical countries as live fencing. the foliage of gliricidia sepium is used for green manuring and to produce other kind of organic manure due to its higher nutritional composition (gunasena, 1994). number of research have been conducted on the antifungal and antimicrobial properties of gliricidia extracts and those researches proved that gliricidia extracts can inhibit the growth of some fungi, bacteria and nematodes (ganesan, 1994). thus, by using gliricidia leaf extract as a nutrient solution can limit plant disease and pest attacks as well. therefore, the objective of this research was to evaluate the suitability of gliricidia leaf extract to replace albert solution. materials and methods tomato variety “thilina” was used as the planting material. seeds were sown in nursery medium of coir dust in plastic trays and three weeks old seedlings were transplanted in polybags filled with medium of 1:1 coir dust and sand. they were placed in a plant house. the gliricidia leaves were collected form the plants grown naturally in faculty of agriculture, university of ruhuna, mapalana. leaves were ground in a laboratory mill mixing 1 l of water to 100 g of leaves. filtered solutions of the leaf extract were used to prepare different nutrient solutions with different composition of albert solution. albert solution was prepared according to the manufactures instructions. the five different nutrient solutions tested in this experiment were as follows, t1 -albert solution 100% (control) t2 -albert solution 75 % + gliricidia leaf extract 25 % t3 -albert solution 50 % + gliricidia leaf extract 50 % t4 -albert solution 25 % + gliricidia leaf extract 75 % t5 -gliricidia leaf extract 100 % treatment combinations were laid in a crd with four replicates. plant height and number of leaves were measured once in two days during early stages and interval of counting number of leaves was increased during flowering stage. time taken for flowering and final yield was also measured. analysis of variance was performed using proc anova with sas statistical software. test of significance of the difference between treatment means was done using dmrt to identify the best medium out of five treatments been evaluated. lakmini, rupika and kalidasa: m o r p h o l o g i c a l h e t e r o g e n e i t y . . . 1 1 3 r u h un a jo u rna l o f s c i e nc e 2 , pp. 111-114 (2 0 0 7) results and discussion a b 10 30 50 70 90 110 130 14 24 34 44 54 64 74 days aft er plant ing h ei gh t ( cm ) t 1 t 2 t 3 t 4 t 5 0 5 10 15 20 25 14 22 30 38 46 54 62 70 days aft er plant ing n ou m be r of le av es t 1 t 2 t 3 t 4 t 5 c d 46 48 50 52 54 56 58 t 1 t 2 t 3 t 4 t 5 t reat ment d ay s af te r pl an tin g d c b b a 0 50 100 150 200 250 300 t 1 t 2 t 3 t 4 t 5 t reat ment y ie ld /p la nt ( g) a a b b b figure 1. effect of different composition of gliricidia leaf extract on (a) plant height, (b) number of leaves, (c) time taken for flowering and (d) final yield. there was no significant difference among treatments in plant height until 19 days after transplanting (fig. 1a). there after treatment 5 showed significantly lower plant height than did other four treatments indicating there is no significant effect on plant height by replacing albert solution with gliricidia leaf extract. 114 lakmini, rupika and kalidasa: m o rp h ol og i c a l he t e r o ge ne i t y . . . r u h u n a j o u r n a l o f sc i e n c e 2 , p p . 111-114 (2 0 07 ) number of leaves per plant was more or less similar in all treatments during early stage of transplanting but after three weeks plants in control showed significantly higher number of leaves (fig. 1b). there was no statistically significant difference among four other treatments showing the ability to use gliricidia leaf extract. significant difference was observed in time taken to flowering (fig. 1c). plants treated only with gliricidia leaf extract showed delay in flowering and plants treated only with albert solution showed early flowering. whereas, among treatment 3 and 4 there were no statistically significant difference. yield is the most important parameter of any nutrient trial. there was a statistically significant difference in treatment 1 and 2 than other three treatments (fig. 1d). significantly higher yield was observed in plants grown in control (200g/plt) and the plants treated with 25% gliricidia extract and 75% albert solution (250g/plt). but these figures are comparatively lower than the total yield obtained from farmers who use only albert solution under normal environmental conditions and also the size of fruits was less. these changes might be the results of high temperature (42 oc) in the plant house during the experimental period. conclusion according to the results obtained, it could be concluded that the growth performances of tomato reduce with increasing composition of gliricidia leaf extract as nutrient solution therefore, gliricidia leaf extract alone can not be used as a nutrient solution for growing tomato. but with considering the yield albert solution can be replaced approximately by 25 % of gliricidia leaf extract and also it would be more economical. moreover, further research should be conducted on nutrient analysis of tomato grown under these nutrient solutions before come in to a final conclusion. references ganesan, t. 1994. antifungal properties of wild plants. adv. plant sci.7(1): 185187. gunasena, h.p.m. (1994). multipurpose tree species in sri lanka. proceedings, 4th regional workshop on multipurpose trees, kandy, sri lanka. molter, a.j and jenson, m.h., 1995. world bank technical paper no. 253. protected agriculture global review. 25-59. ruhuna journal of science vol 10(1): 32-50, june 2019 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v10i1.49 university of ruhuna  faculty of science, university of ruhuna sri lanka 32 efficacy of gum arabic as an esculent film on shelf life extension of tomato (solanum lycopersicum l) fruit t. a. fashanu 1 *, a. t. oladiji 2 , o. a. peters 1 1 nigerian stored products research institute, km 3, asa-dam road p. m. b 1489, ilorin, nigeria. 2 department of biochemistry, university of ilorin, p.m.b 1515, ilorin, nigeria *correspondence: titifashanu@gmail.com; https://orcid.org/0000-0001-6022-6149 received: 30 th june 2018, revised: 19 th november 2018, accepted: 8 th march 2019 abstract effect of gum arabic (ga) as a non-toxic outer layer to prolong the shelf-life of tomato was carried out in the present study. tomatoes were coated with 0% (control) (a), 5% w/v (b), 10% w/v (c), 15% w/v (d) and 20% w/v (e) ga and stored at ambient (33°c; 70%). the polygalacturonase and β-galacturonase activities were carried out. weight loss, titratable acidity, ph, lycopene, β-carotene and antioxidant property of the fruits were determined at day 0, 5, 10, 15 and 20 during the experimental duration. results indicated that polygalacturonase activity ranged from 0.753-1.138 unit/mg, β-galacturonase activity ranged from 1.55-2.78 x 10 3 miller units, β-carotene ranged from 1.78-23.89 mg/g, antioxidant property ranged from 66.26-192.12μg/g, titratable acidity was from 0.595-0.95% while the lycopene ranged from 0.77-17.62 mg/g. these results showed that control had significant rapid weight loss with faster rate of softening while the tomatoes coated with ga demonstrated a significant (p<0.05) deferment in ripening as well as maintained its nutritional quality up to 20 days. this indicates that ga may be utilized as an edible coating for tomato. key words: edible film, gum arabic, ripening, shelf life. 1 introduction the process of gas exchange continues in fresh fruits and vegetables even when they have been harvested and are thus extremely short-lived and vulnerable to various diseased conditions (chien et al. 2007). swift maturity and senescence which promotes desiccation or water loss is as a result of continuous respiratory activity which in turn leads to quality loss of the produce (maftoonazard et al. 2008). tomato originated from elevated region of peru and ecuador, and it is usually cultivated in mild weather conditions. it ranks next to potato and sweet potato with respect to world vegetable mailto:titifashanu@gmail.com https://orcid.org/0000-0001-6022-6149 t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 33 production. it is widely cultivated in tropical, subtropical and temperate climates, and thus ranks third in terms of world vegetable production (gebisa et al. 2016). the appearance of a good fruit must have a hard, uniform and shiny colour, and no signs of mechanical damage, shriveling, or rot. principal causes for postharvest losses are decay, external damage encountered during harvest and handling at an inappropriate stage (sargent and moretti 2002). during the early phase of ripening, tomatoes are responsive to chilling injury thus the use of low temperature for preservation is not encouraged (harold et al. 2007). it is usually advised to harvest tomato fruits at mature green phase to reduce damage during post-harvest handling. the major loss of tomatoes occurr as a result of deterioration during handling and distribution, and this damage is intense when they are at the breaker and progressive stage of ripeness (moneruzzaman et al. 2008). edible food packaging is of major interest. the increasing concern and inquiry in edible packaging have been motivated by the current awareness of the unfavourable effect of non-decomposable packaging waste and the rise in consumer preference for secure, simple and consistent food (krochta 2002). gum arabic is an anhydrous discharge from the stalk and shoot of acacia senegal trees which is grown in sudan as a cash crop. gum arabic is known globally in the comestible, beverage and pharmaceutical sector as a competent supplement to produce a stable system, film building and an outer layer/ encapsulating agent, oxidation inhibitor, stabilizer, emulsifier, texturant, clouding, clarifying agents and food adhesive (murwan et al. 2008). gum arabic evolves and exists within the confine of the ecosystem, and is made up of carbohydrate characterized with high affinity for water and water hating protein components (fao 1990). the hydrophobic protein function in keeping an emulsion from separating by accumulating on a surface of oil droplets while the water loving carbohydrate component prevents suspension of particle in a liquid and merging of two molecules (anderson and weiping 1990). the development of edible coating produced from two composites, such as polysaccharides, proteins and/or lipids increasing the functionality of the coating has received attention in a recent review (hassan et al. 2018). this is because each coating material has unique but limited functions. polymers such as polysaccharides and protein, which are hydrophilic in nature are excellent thin layer mold with good flavour and serve as obstacles to lipids at low relative humidity. however, they cannot act as barriers to water. these materials are better gas barriers rather than preventing moisture loss. nevertheless, some polysaccharides can retard moisture loss by serving as molecules to be destroyed rather than moisture barrier (kester and fennema 1986). films made from proteins are fragile and can easily break due to its strong conformity, hence, the addition of a material that can interact cohesively with protein makes it more flexible (sothornvit and krochta 2005). protein film is also a poor moisture barrier like a polysaccharide film (lin and t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 34 zhao 2007). ali et al. (2010) observed that, in banana, gum arabic has not shown any antifungal activity against colletotrichum musae (the causal agent of anthracnose) but mixture of coatings produced from gum arabic with chitosan showed remarkable antifungal effects compared to only chitosan. a study by ali et al. (2013) explained the operation of a new coating produced from gum arabic on tomato fruits, it was observed that compared to uncoated fruits, tomatoes coated with 10% gum arabic presented significantly lower rates of changes in weight, colour, firmness, titratable acidity, soluble solids content, ascorbic acid content, and decay percentage. the aim of the present research is therefore to formulate comestible film with gum arabic which is readily available and cost effective (in nigeria) to prolong the shelf life of tomato fruit, and to monitor the activity of the softening enzymes and nutritional qualities of the tomato fruit during storage. 2 materials and methods 2.1 reagents all reagents used were of analytical grade, and most of them were products of sigma-aldrich, germany and bdh, england. 2.2 plant materials fresh tomato (abindi kerewa var.) was harvested from a farm within university of ilorin campus, ilorin, nigeria and was transported in stackable plastic crates to the chemistry/ biochemistry laboratory of nigerian stored products research institute (nspri) headquarters, ilorin, nigeria. the sample was identified and mature green tomato with trace of yellow according to the usda standard tomato colour classification chart (usda, 1991) and of consistent size with no insect damage were used in the present study. tomatoes were grouped into five, and placed separately in storage cartons. the initial weight of each group was noted. group a was the control while groups b, c, d and e were coated with 5%, 10%, 15% and 20% (w/v) gum arabic (loba chemie, india) respectively. 2.3 production of gum arabic (ga) solution and treatment of sample with coating ali et al (2013b) method was adopted for the preparation of ga with little adjustment. ga solutions of 5%, 10%, 15% and 20% were prepared by melting 5 g, 10 g, 15 g and 20 g of its pulverized form in 100 ml of distilled t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 35 water. the mixtures were agitated with low heat (40°c) for 60 min using a magnetic stirrer/hot plate. glycerol monostearate (1.0%) (fisher chemicals) was annexed as a plasticizer after cooling to 30°c in order to enhance the potency and elasticity of the coating mixtures. 1m naoh was used to adjust the ph of the mixture to 5.8. the tomatoes in five different groups were subjected to five distinctive treatments, and each treatment was administered in triplicate. fruits were soaked in each concentration of ga solution for 1-2 min and assuring a uniform and consistent coating of the whole surface. the control fruits were immersed in distilled water. the fruits were then air-dried, loaded in storage carton, and stocked at ambient temperature (33°c) at 70– 80% relative humidity. the data were collected before treatment (day 0) and at 5 day intervals for 20 days. 2.4 enzymatic assay of polygalacturonase the enzymatic assay of polygalacturonase was done following the method explained by somogyi (1952). principle: calculations: ( )( ) ( )( ) df = dilution factor 10 = time of assay (in minutes) as per the unit definition 0.1 = volume (in milliliters) of enzyme used 2.5 determination of β-galactosidase activity this was carried out according to miller method (1972). the level of βgalactosidase enzyme was calculated in miller units as: ( ) 2.6 estimation of β-carotene and lycopene β-carotene was estimated using the dried methanol extract following kumar et al. (2001) method. a volume of 10 ml of acetone-hexane mixture (4:6) was used to extract 100 mg of sample for 1 min and filtered using whatman filter paper no 1. the absorbance was recorded using uv-vis spectrophotometer t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 36 (searchtech, england) at three different wavelengths (453, 505 and 663 nm). the lycopene content was estimated by: ( ) ( ) 2.7 extraction for antioxidant and total phenolics determination norsyamimi et al. (2014) method was adopted for the extraction with little modification. drying of tomato was done using hot-air oven (drying oven searchtech, england) at 40°c and then pulverized using hammer mill (polymix® switzerland). the dried and pulverized samples (10 g) were extracted in 90% methanol solvent at ambient temperature (33 ° c) and it was left to stand for 24 hrs, the solutions were sifted using whatman filter paper no 1. the filtrate collected was left to evaporate to produce a thick mass. the extracts were stored at 0-4°c for subsequent analyses. 2.8 estimation of antioxidant capacity in terms of ferric reducing antioxidant property the ferric reducing antioxidant capacity was determined by the method explained by ali et al. 2013b. a mixture of 40 μl of tomato fruit extract and 3 ml of frap reagent (sigma-aldrich) was prepared. the resulting solution was incubated for 4 min at 37 °c. absorbance was recorded with uv-vis spectrophotometer (searchtech england) at 593 nm, and the outcome were given as the concentration of antioxidants having a ferric reducing action equivalent to 1 mg −1 g feso4 of fresh weight of fruit sample. 2.9 determination of total phenolic contents makkar et al. (1997) method was used for the determination of total phenolic content. portions of the tomato extract in a test tube were filled to the volume of 1 ml with distilled water. 2.5 ml of naco3 solution (20%) and 0.5ml of folin-ciocalteu reagent (loba chemie, india) (1:1 with water) were added sequentially to the test tube, the resulting solution was shaken and the tubes were positioned away from light for 40 min, and the absorbance using uvvis spectrophotometer (searchtech. england) was taken at 725 nm against the reagent blank. a standard curve was plotted using gallic t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 37 monohydrate and the total phenolic content was estimated and expressed as gallic acid equivalent in mg/g of extract. 2.10 determination of ph sharoba (2009) method was used to determine the ph with little modification as follows; 10 g of sample was homogenized and centrifuged (centurion scientific limited united kingdom) (5000g, for 20 min), at 4°c. the supernatant was recovered for ph measurement. the ph was measured at 20 °c with a ph meter (searchtech phs-3c england). 2.11 determination of titratable acidity aoac (2002) method was used to estimate titratable acidity. a volume of 25 cm 3 of distilled water was added to 10 g of the sample and the mixture was agitated. whatman (no. 1) filter paper was used to filter the mixture, and 10 ml of filtrate was taken into a conical flask, and two drops of phenolphthalein added as indicator. titration of the solution was done against 0.1m naoh until the colour changes to pink. titratable acidity was represented as percentage citric acid, where v= volume of 0.1 m naoh used, m= molarity of naoh, and f= factor of citric acid (0.064). 2.12 determination of weight loss the change in weight was monitored for each group during the storage period with the use of top loading balance (snowrex electronic scale, london). this was done by checking the weight of tomato fruits before taking samples for analyses. 2.13 statistical analysis the experiments were arranged in completely randomized design (crd) with three replicates, data were subjected to analysis of variance (anova), and tested for significance among treatments by duncan’s multiple range test (dmrt) at (p<0.05) using spss version 21. t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 38 3. results and discussion 3.1 polygalacturonase activity effect of gum arabic (ga) on the polygalacturonase activity of tomatoes is as illustrated in figure 1. the enzyme level increased in the control and all the treated groups till day 15, however at day 20, group e (20% ga) showed the highest enzyme level of 1.138 unit/mg followed by group a (control) which was 1.052 unit/mg with groups b and d being 0.866 unit/mg and 0.929 unit/mg respectively while group c showed the lowest level of the enzyme (0.753 unit/mg). fig. 1. effect of gum arabic (ga) treatment on polygalacturonase activity of each treatment during the storage period. bars represent means of triplicate readings (n=3). bars with shared alphabet are not significantly different (p<0.05). error bars represent standard error (se) of the mean. a= control (0% ga); b= 5% ga; c=10% ga; d=15% ga; e=20% ga. the polygalacturonase [pg; poly (1,4-α-d-galacturonide) glycanhydrolase; ec 3.2.1.15] is specific in tomato only during the ripening stage of fruit maturity. pg plays an essential role in cell wall breakdown, becomes bulky during ripening, and has a role in fruit softening too (raymond et al. 1988). there were significant differences between group a and c (p<0.05) at day 20 while there was no significant difference between groups b and d (p>0.05). there was also no significant difference (p>0.05) between groups a and e. a a a a a b bc bc a ab b b ab a a c c ab a ab ab c d bc a t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 39 the higher enzyme level recorded in group a and e might be due to loss of fruit firmness as a result of increase hydration of cell wall occasioned by modification of pectin rich middle lamella. the changes in the arrangement of pectin gel govern the loosening of the gel, and the cell can separate from one another (alexander and grierson 2002). raymond et al. (1988) reported similar results where they affirmed that polygalacturonase activity increased during ripening and perform an essential role in cell wall degeneration and fruit softening. earlier report by anurag et al. (2009) also stated that the polygalacturonase enzyme level increases as fruit ripens as observed in the present study. 3.2 β-galactosidase activity the level of β-galactosidase activity of tomatoes coated with ga is shown in figure 2. the level of β-galactosidase increased in all the test groups. in group a, the enzyme increased from 1.54-2.78 x 10 3 miller units, for group b, it increased from 1.53-2.69 x 10 3 miller units, group c increased from 1.552.59 x 10 3 miller units, group d increased from 1.57-2.782 x 10 3 miller units while group e increased from 1.58-2.75 x 10 3 miller units during the storage period. there was no significance difference (p>0.05) in the enzyme level at day 20. fig. 2. effect of gum arabic (ga) treatment on β-galactosidase activity of each treatment during the storage period. bars represent means of triplicate readings (n=3). bars with shared alphabet are not significantly different (p<0.05). error bars represent standard error (se) of the mean. a= control (0% ga); b= 5% ga; c=10% ga; d=15% ga; e=20% ga. 0 500 1000 1500 2000 2500 3000 3500 0 5 10 15 20 β -g al ac to si d as e( m ill er u n it ) storage tme (days) a b c d e c c c c a c c c c c c b b b ab a b b ab a ab a a a a a http://jxb.oxfordjournals.org/search?author1=lucille+alexander&sortspec=date&submit=submit http://jxb.oxfordjournals.org/search?author1=don+grierson&sortspec=date&submit=submit t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 40 in group a, there was an increase in the enzyme level up to day 15, however at day 20, the level of the enzyme increased rapidly with group c showing lowest level of the enzyme indicating that the fruits in this group retain its firmness. anowar et al. (2014) reported that β-galactosidase could be an important enzyme mainly accredited for cell wall alteration and changes in fruit during ripening as it is known to be a glycosidase that acts on short chain oligomers of galactose residues present either as homo/ hetero polysaccharides, glycoproteins, or glycolipid. β-galactosidase activity increased in all the groups as the activity of the enzyme is seen in all the groups; however the lowest activity recorded in group c might be as a result of the reduction of cell wall modification. as observed in the present study similar findings were reported by emmadeldin et al. (2012) where they reported that βgalactosidase plays a major role in fruit softening. 3.3 changes in the β-carotene content the effect of ga coating on β-carotene content present in each group during the storage period is shown in figure 3. β-carotene content in group a (control) increased from 1.78 mg/g at day 5 to 21.79 mg/g at day 15 after which it dropped sharply to 3.54 mg/g at day 20. in all the treated groups (b, c, d and e) the β-carotene content also increased up to day 15 and decreased at day 20. group b (5% ga) showed the highest β-carotene content of 23.89 mg/g at day 15 while group e (20%) showed the lowest βcarotene content of 14.47 mg/g at day 15. fig. 3. effect of gum arabic (ga) treatment on β-carotene content of each treatment during the storage period. bars represent means of triplicate readings (n=3). bars with shared alphabet are not significantly different (p<0.05). error bars represent standard error (se) of the mean. a= control (0% ga); b= 5% ga; c= 10% ga; d= 15% ga; e= 20% ga. -5 0 5 10 15 20 25 30 0 5 10 15 20 β -c ar o te n e ( m g/ g) storage time(days) a b c d e t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 41 the pro-vitamin a activity of β-carotene is of special interest. the inadequate retinol supply in notable parts of the world is being balanced by β-carotene intake (tilman et al. 2010). there was a significant difference (p<0.05) between group b (5%) and e (20%) at day 15 while there were no significant differences (p>0.05) among groups a, b, c and d. the sharp increase in βcarotene content in group a might be due to ripening of the tomato fruit compared to group d and e. the result is similar to what was reported by meredith and young in (1971) where β-carotene content in tomatoes decreased as it reaches the peak of ripening stage. ali et al. (2013) also reported similar results where β-carotene increased in the control up to 8 days; however the change in the current study may be connected with geographical location. 3.4 changes in the lycopene content the effect of ga treatment on the lycopene contents of treated tomato are illustrated in figure 4. the lycopene content increased in the control and the treated groups. the lycopene content in group a (control) increased from 0.77 mg/g at day 0 to 15.69 mg/g at day 15 and the group b reached peak of 16.33 mg/g at day 15. the group c had the highest lycopene content of 17.62 mg/g at day 15 which dropped sharply at day 20. the lycopene content of group d and group e were comparably low to those of groups a, b and c at day 15. fig. 4. effect of gum arabic (ga) treatment on lycopene content of each treatment during the storage period. bars represent means of triplicate readings (n=3). bars with shared alphabet are not significantly different (p<0.05). error bars represent standard error (se) of the mean. a= control (0% ga); b= 5% ga; c=10% ga; d=15% ga; e=20% ga. -5 0 5 10 15 20 25 0 5 10 15 20 ly co p e n e (m g/ g) storage tme (days) a b c d e t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 42 compared to other carotenoids, lycopene had a very potent antioxidant activity and demonstrate the strongest physical quenching rate constant with reactive oxygen species (abdul-hammed et al. 2015). there were no significant differences (p>0.05) in lycopene content of the groups a, b, c and d while group e showed a significant difference (p<0.05) at day 15. lycopene content increased as fruit matures (gil-randez et al. 1998). the initial increase in lycopene content as storage days increased might be due to increased respiratory rate of fruit which in turn increased the breakdown of lycopene due to cellular breakdown thereby resulting in a decrease in lycopene content towards the end of the experiment. the immediate increase in the lycopene content in groups a, b and c may be due to ripening of the fruit compared to those groups treated with higher concentration. ali et al. (2013) reported similar results when increase in lycopene content was attributed to faster ripening. abdul-hammed et al.(2015) also reported similar results where the concentrations of lycopene, the ripening and antioxidant index of tomatoes becomes abundant from the breaker stage to the light red and the decrease when fully red. 3.5 change in the antioxidant properties the effect of ga coating on the antioxidant capacity determined in terms of frap (ferric reducing antioxidant power) of tomatoes is shown in figure 5. fig. 5. effect of gum arabic (ga) treatment on the antioxidant property of each treatment during the storage period. bars represent means of triplicate readings (n=3). bars with shared alphabet are not significantly different (p<0.05). error bars represent standard error (se) of the mean. a= control (0% ga); b= 5% ga; c=10% ga; d=15% ga; e=20% ga. 0 50 100 150 200 250 0 5 10 15 20 fr a p ( μ g/ g) storage tlme (days) frap a b c d e bcd cd b bc c d bc a bc b a a a a ab c c a b b d c a b b t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 43 group a (control) showed the highest antioxidant capacity of 165.59 µg/g at day 10 and dropped at day 15 and 20 to 84.62 µg/g and 55.50 µg/g respectively. groups b and c also showed the highest antioxidant capacities at day 10 and 15 respectively while groups d and e showed highest antioxidant capacity at day 10. group c compared to others showed highest antioxidant capacity of 192.12 µg/g at day 15. the advantage of consuming fruit and vegetables cannot be ascribed to a lone compound but to the stimulating and additive effects between different phytochemicals. free radicals scavenging abilities in fresh fruits and vegetables rely on several factors such as species, cultivars, environmental conditions, geographical origin and analytical methods (chun et al. 2005). there was a significant difference (p<0.05) among group c, d and e while groups a and b showed no significant difference (p>0.05) at day 15. also groups d and e showed no significant difference (p>0.05). there is usually a correlation between the antioxidant activity with the total phenolic and flavonoid contents (tehrani et al. 2011). this corresponding characteristic is an indication that phenolics and flavone compounds are possibly the main compound responsible for the high antioxidant in fruits (stratil et al. 2007). reyes and cisneros-zevallos (2003) also reported corresponding characteristics between total phenolics and total antioxidant activity. similar result was reported by bhandari and lee (2016) where antioxidant capacity increase to a certain ripening stage in some cultivars of tomato and later decreased. 3.6 change in the total phenolics content the change in the total phenolic contents in all the groups of tomatoes as affected by ga treatment are shown in figure 6. the total phenolic contents increased in group a from 227.44 µg/kg to 3563.00 µg/kg at day 15 and thereafter decreased to 1101.89 µg/kg at day 20. also in the treated groups (b, c, d, e) the total phenolic contents also increased up to day 15 and decreased at day 20. although the highest phenolic content was reached in all the groups on day 15, the group c (10% g.a) showed the highest phenolics at 3825.73 µg/kg. the total phenolic contents reduced at day 20 in all the groups with the group a being the lowest at 1101.89 µg/kg. t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 44 fig. 6. effect of (gum arabic) ga treatment on total phenolics content of each treatment during the storage period. bars represent means of triplicate readings (n=3). bars with shared alphabet are not significantly different (p<0.05). error bars represent standard error (se) of the mean. a= control (0% ga); b= 5% ga; c=10% ga; d=15% ga; e=20% ga.; d=15% ga; e=20% ga. the health promoting effects of fruits and vegetables is attributed to phenolic compounds; essential secondary plant metabolites, the most important of which is the antioxidant activity associated with diminished danger of cancer and cardiovascular diseases. phenolic compounds contribute about 60–70% of the ability to prevent oxidation of other chemicals of tomato extracts (bhandari and lee 2016). there was no significant difference (p>0.05) among the groups at day 20. the maximum total phenolic content shown in group c (10% ga) is an indication that the group retained higher amount of antioxidant than the other groups. the effect of ethylene, a phytohormone which simulate the activity of phenylalanine ammonium lyase (pal) an important enzyme in phenol biosynthesis could result in increase in phenol content during ripening (anne and michaela 2009). the senescence and breakdown of cell structure during storage might results in the reduction in total phenolics content recorded towards the end of the experiment in all the groups (macheix et al. 1990). similar results were reported by ghasemnezhad et al. (2010) and ali et al. (2010), where a decrease in total phenolics was recorded at higher concentration of chitosan. 3.7 change in titratable acidity the effect of ga titratable acidity values of tomatoes are as illustrated in figure 7. titratable acidity values increased in all the groups up to 15 days of 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 5 10 15 20 t o ta l p h e n o li cs (µ g /k g ) storage time (days) a b c d e de d e e e de de d e e e d d a de b e ab a a a b c c c c c t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 45 storage with fruits in group c showing highest value of 0.95g/ml at day 15. the titratable acidity thereafter decreased in all the groups with group e showing the lowest value of 0.61g/ml at day 20. the titratable acidity of the tomato increased significantly (p<0.05) from 10 days of storage; this could be potentially attributed to more organic acids being produced during the storage period. in contrary, lugwisha et al. (2016) reported a decrease in titratable acidity in tanzania sugar apple fruits from 0.28 to 0.12% during ripening, which was predicted to be associated with the usage of component acids in the respiratory process. fig. 7. effect of gum arabic (ga) treatment on titratable acidity of each treatment during the storage period. bars represent means of triplicate readings (n=3). bars with shared alphabet are not significantly different (p<0.05). error bars represent standard error (se) of the mean. a= control (0% ga); b= 5% ga; c=10% ga; d=15% ga; e=20% ga. 3.8 changes in the ph value the effect of ga treatment on ph values of tomatoes during the storage period are shown in figure 8. the ph of group a falls within the normal range for tomato (3.7-4.5). the ph of the treated groups (b, c, d, e) also falls within the normal range of tomatoes with an exception in group e which was 4.70 at day 20. 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 0 5 10 15 20 t it ra ta b le a ci d it y ( g / m l) storage time(days) a b c d e http://www.sciencedirect.com/science/article/pii/s187853521630106x#b0135 http://www.sciencedirect.com/science/article/pii/s187853521630106x#b0135 t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 46 fig. 8. effect of gum arabic (ga) treatment on ph of each treatment during the storage period. bars represent means of triplicate readings (n=3). bars with shared alphabet are not significantly different (p<0.05). error bars represent standard error (se) of the mean. a= control (0% ga); b= 5% ga; c=10% ga; d=15% ga; e= 20% ga. the ph of the control and treated groups were fluctuating all through the storage period and they were all within the range acceptable for tomato (3.7 4.5), however at day 20 the ph of group e exceeded 4.5, this phenomenon could possibly be due to oxidation of acid during storage resulting in higher ph. the previous report available on the ph values of tomatoes was strengthened in this result. for instance, mohammed et al. (1999) reported that though the ph of ripe tomatoes may be greater than 4.6, tomato products are widely grouped as acidic foods with the ph below 4.5 being the acceptable value as it stops microorganism breeding. 3.9 weight changes the effect of ga treatment on change in weight of tomato fruits in each group is presented in table 2. the weight decrease in all the groups, for group a, the average weight reduced from 311.67 g to 275.00 g, group b reduced from 306.67 g to 272.66 g, group c also dropped from 309.67 g to 278.33 g, group reduced from 283.00 g to 246.00 g, and group e reduced from 262.67 g to 221.33 g. 3.5 3.7 3.9 4.1 4.3 4.5 4.7 4.9 0 5 10 15 20 c h a n g e i n p h storage days a b c d e t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 47 table 2. effect of gum arabic (ga) on weight changes (g) of coated tomato during the storage period (all the values are mean ± sem of triplicates. a= control (0% ga); b= 5% ga; c=10% ga; d=15% ga; e=20% ga). group initial day 5 day 10 day 15 day 20 a 311.67±4.33 a 291.67±4.09 ab 287.00±2.03 c 282.67±4.37 b 275.00±6.08 b b 306.67±4.91 c 291.00±1.53 b 287.33±0.83 b 280.00±1.15 c 272.66±2.33 c c 309.67±4.91 b 296.00±4.58 a 293.67±4.7 a 290.33±4.91 a 278.33±4.48 a d 283.00±13.32 ad 269.33±13.38 c 264.00±14.19 ad 254.33±14.43 ab 246.00±13.20 ab e 262.67±5.04 ac 253.33±4.41 ab 240.00±3.05 ac 229.33±3.18b c 221.33±3.71 ac generally tomato stored at ambient condition tends to lose water which may results in loss of weight. singh and reddy (2006) also attributed the loss of weight to the change in soluble sugar concentration resulting from the usage of monosaccharides for respiratory purposes during storage. the loss of water from fruit is normally through diffusion from the fruit skin to the atmosphere, the extent of transpiration depends on external and environmental factors such as temperature, relative humidity, air movement and atmospheric pressure (joyce and patterson, 1994). the weight loss observed in this study is in agreement with the report of singh and reddy (2006). 4 conclusions comparative analysis on the effects of several coating concentrations of gum arabic and storage condition on the quality and the shelf-life of tomato fruits, revealed that the group coated with 10% gum arabic and stored at ambient temperature was the best treatment for maintaining the quality and extending the shelf-life of tomato fruit over other treatments or control. it showed lower polygalacturonase activity, lower β-galactosidase as well as higher lycopene and β-carotene, higher antioxidant capacity in terms of ferric reducing antioxidant power, higher total phenolics and higher titratable acidity, reduced decay incidence. results suggest that gum arabic is effective in reducing the activity of the ripening enzyme while maintaining the quality and extending the shelf-life of tomato fruit. references abdul-hammed m, oke mo, bolarinwa if. 2015. carotenoid accumulation pattern and nutritional indices of cherry-nasmata and var-10 tomato varieties. international food research journal 22(2): 761-767. alexander l, grierson d. 2002. ethylene biosynthesis and action in tomato: a model for climacteric fruit ripening. journal experimental botany 53 (377): 2039-2055. http://jxb.oxfordjournals.org/search?author1=lucille+alexander&sortspec=date&submit=submit http://jxb.oxfordjournals.org/search?author1=don+grierson&sortspec=date&submit=submit t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 48 ali a, mahmud tm, sijam k, siddiqui y. 2010. potential of chitosan coating in delaying the postharvest anthracnose (colletotrichum gloeosporioides penz.) of eksotika ii papaya. international journal of food science and technology 45: 2134–2140. ali a, maqbool m, alderson pg, noosheen z. 2013a. effect of gum arabic as an edible coating on antioxidant capacity of tomato (solanum lycopersicum l.) fruit during storage. postharvest biology and technology elsevier 76: 119–124. ali a, zahid n, manickam s, siddiqui y, alderson pg, maqbool m. 2013b. effectiveness of submicron chitosan dispersions in controlling anthracnose and maintaining quality of dragon fruit. postharvest biology and technology 86:147–153. anderson dmw and weiping w. 1990. the characterization of acacia paolii gum and four commercial acacia gums from kenya. food hydrocolloids 3(6):475-484. anne m, michaela s. 2009. polyphenol content and antioxidant capacity of apple fruit: effect of cultivar and storage conditions. journal of applied botany and food quality 82: 152 – 157. anowar h, masud r, yoshinobu k, hairul ar. 2014. changes in biochemical characteristics and activities of ripening associated enzymes in mango fruit during the storage at different temperatures. biomed research international volume 2014, article id 232969, 11 pages. anurag p, nagendra nm, ana lsc, randhir s. 2009. biochemistry of fruit softening: an overview, physiology and molecular biology of plants 15(2):103-13. aoac 2002. official method of analysis of aoac international, 17th ed. published by the association of official analytical chemists international, suite 400 2200, wilson boulevard, arlington, virginia 2220 –3301, usa. gebisa b, gezu d, alemayehu b, fikadu t. 2017. performance evaluation of tomato (lycopersicon esculentum mill.) varieties under supplemental irrigation at erer valley, babile district, ethiopia. journal of plant sciences. 5 (1): 1-5. doi: 10.11648/j.jps.20170501.11 bhandari sr, lee jg. 2016. ripening-dependent changes in antioxidants, color attributes, and antioxidant activity of seven tomato (solanum lycopersicum l.) cultivars. journal of analytical methods in chemistry vol. 2016, aricle id 5498618, 13 pages. https//doi.org/10.1155/2016/5498618. miller jh. 1972. experiments in molecular genetics, cold spring harbor, ny: cold spring harbor laboratory press chien pj, sheu f, lin hr. 2007. coating citrus (murcott tangor) fruit with low molecular weight chitosan increases postharvest quality and shelf life. food chemistry 100:1160– 1164. chun ok, kim d, smith n, schroeder d, han jt, lee cy. 2005. daily consumption of phenolics and total antioxidant capacity from fruit and vegetables in the american diet. journal of the science of food and agriculture 85(10): 1715-1724. emmadeldin hek, mathilde c, mireille f, 2012. cell wall glycosidase activities and protein content variations during fruit development and ripening in three texture contrasted tomato cultivars. saudi journal of biological sciences 19(3): 277–283. fao. 1990. food and agriculture organization of the united nations. the design of agricultural investment projects lessons from experience. technical paper no. 5, investment centre, fao, rome, italy. ghasemnezhad mm, shiri a, sanavi m. 2010. effect of chitosan coatings on some quality indices of apricot (prunus armeniaca l.) during cold storage. caspian journal of environmental science 8(1):25-33. gil-randez f, herrero p, sanz p, prieto ja and moreno f. 1998. hexokinase pii has a double cytosolic-nuclear localisation in saccharomyces cerevisiae. febs letter 425(3):475-8. harold cp, ioannis ck, penelope jb and dimitrios s. 2007. a review of recent research on tomato nutrition, breeding and post-harvest technology with reference to fruit quality. the european journal of plant science and biotechnology ©2007 global science books. https://www.researchgate.net/researcher/72755356_dmw_anderson https://www.researchgate.net/researcher/74236041_wang_weiping https://www.researchgate.net/journal/0268-005x_food_hydrocolloids https://www.hindawi.com/31965371/ https://www.hindawi.com/94549105/ https://www.hindawi.com/14642596/ https://www.hindawi.com/36515737/ https://www.researchgate.net/journal/0971-5894_physiology_and_molecular_biology_of_plants http://www.sciencedirect.com/science/article/pii/s1319562x12000344 http://www.sciencedirect.com/science/article/pii/s1319562x12000344 http://www.sciencedirect.com/science/article/pii/s1319562x12000344 http://www.sciencedirect.com/science/journal/1319562x http://www.sciencedirect.com/science/journal/1319562x http://www.yeastgenome.org/author/herrero_p/overview http://www.yeastgenome.org/author/sanz_p/overview http://www.yeastgenome.org/author/prieto_ja/overview http://www.yeastgenome.org/author/moreno_f/overview t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 49 hassan b, shahzad asc, abdullah ih, khalid mz, naseem a. 2018. recent advances on polysaccharides, lipids and protein based edible films and coatings:a review. international journal of biological macromolecules 109 1095–1107. joyce d, patterson b. 1994. postharvest water relations in horticultural crops: principles and problems. aciar proceedings 50: 228–238. kester jj, fennema or. 1986. edible films and coatings: a review. food technology 40:47– 59. krochta jm. 2002. proteins as raw materials for films and coatings: definitions, current status and opportunities print isbn: 978-1-58716-107-0; ebook isbn: 978-1-42003198-0. kumar r, gupta np. 2001. lycopene therapy in idiopathic male infertility: results of a clinical trial presented at: 34th annual conference of the urological society of india. lin d, zhao y. 2007. innovations in the development and application of edible coatings for fresh and minimally processed fruits and vegetables. comprehensive reviews in food science and food safety-crfsfs 6 (3): 60-75. lugwisha ehc, fabian c, othman oc. 2016. postharvest changes in physicochemical properties and llevels of some inorganic elements in sugar apple (annona squamosa l.) fruits of coast region, tanzania journal of food nutrition science 4: 41–48. macheix jj, fleuriet a, billot j. 1990. fruit phenolics. pp.378, crc press, incorporated. maftoonazad n, ramaswamy hs. 2008. effect of pectin based coating on the kinetics of quality change associated with stored avocados. journal of food processing and preservation 32(4):621-643. makkar hps, becker k, abel h, pawelzik e. 1997. nutrient contents, rumen protein degradability and antinutritional factors in some colourand white-flowering cultivars of vicia faba beans. journalof science food agriculture 75 (4): 511-520. meredith fi, young rh. 1971. changes in lycopene and carotene content of ‘rebblush’ grapefruit exposed to high temperature. horticulural science 6: 233-234. mohammed m, wilson la, gomes pl. 1999. postharvest sensory and physiochemical attributes of processing and non-processing tomato cultivar. journal of food quality 22:167–182. moneruzzaman km, hossain abms, sani w, saifuddin m. 2008. effect of stages of maturity and ripening conditions on the biochemical characteristics of tomato. american journal of biochemistry and biotechnology 4 (4): 336-344. murwan k, sabah e, abu e, yagoub a, asma aab. 2008. emulsion-stabilizing effect of gum from acacia senegal (l) willd. the role of quality and grade of gum, oil type, temperature, stirring time and concentration. pakistan journal of nutrition 7 (3): 395399. norsyamimi h, masturah m, nurina a, syarul nb. 2014. solvent selection in extraction of essential oil and bioactive compounds from polygonum minus. journal of applied sciences 14: 1440-1444. raymond es, matthew k, william rh. 1988. reduction of polygalacturonase activity in tomato fruit by antisense rna. proceedings of the national academy of science usa. 85(23): 8805–8809. reye lf, cisneros-zevallos l. 2003. wounding stress increases the phenolic content and antioxidant capacity of purple-flesh potatoes (solanum tuberosum l.). journal of agricultural and food chemistry. 51(18): 5296-5300. sargent a, moretti cl. 2002. tomato. in: gross, k. wang, c. and saltveit, m. (eds.). tomato. in: gross k, wang c, saltveit m. (eds.).the commercial storage of fruits, vegetables and florist and nursery stocks. united states department of agriculture handbook 66. sharoba am. 2009. producing and evaluation of red pepper pastes as new food product. annals of agricultural science moshbohor 47(2): 151-165. singh kk, reddy bs. 2006. post-harvest physicomechanical properties of orange peel and fruit. journal of food engineering 73:112-120. somogyi m. 1952. notes on sugar determination. journal biological chemistry 195: 19–23. https://www.cabdirect.org/cabdirect/search/?q=au%3a%22macheix%2c+j.+j.%22 https://www.cabdirect.org/cabdirect/search/?q=au%3a%22billot%2c+j.%22 https://www.cabdirect.org/cabdirect/search/?q=do%3a%22fruit+phenolics.%22 https://www.cabdirect.org/cabdirect/search/?q=pb%3a%22crc+press%2c+inc.%22 https://www.ncbi.nlm.nih.gov/pubmed/?term=sheehy%20re%5bauthor%5d&cauthor=true&cauthor_uid=16593997 https://www.ncbi.nlm.nih.gov/pubmed/?term=kramer%20m%5bauthor%5d&cauthor=true&cauthor_uid=16593997 https://www.ncbi.nlm.nih.gov/pubmed/?term=hiatt%20wr%5bauthor%5d&cauthor=true&cauthor_uid=16593997 https://www.ncbi.nlm.nih.gov/pmc/articles/pmc282595/ https://www.researchgate.net/profile/lou_reyes/publication/10606861_wounding_stress_increases_the_phenolic_content_and_antioxidant_capacity_of_purple-flesh_potatoes_solanum_tuberosum_l/links/5428e0060cf26120b7b5720f.pdf https://www.researchgate.net/profile/lou_reyes/publication/10606861_wounding_stress_increases_the_phenolic_content_and_antioxidant_capacity_of_purple-flesh_potatoes_solanum_tuberosum_l/links/5428e0060cf26120b7b5720f.pdf t.a. fashanu et al. efficacy of gum arabic on shelf life of tomato ruhuna journal of science vol 10(1): 32-50, june 2019 50 sothornvit r, krochta jm. 2005. plasticizers in edible films and coatings. in: innovations in food packaging, pp. 403-433. doi: 10.1016/b978-012311632-1/50055-3 stratil p, klejdus b, kuban v. 2007. determination of phenolic compounds and their antioxidant activity in fruits and cereals. talanta. 71:1741-1751. tehrani ms, chandran ab, hossain ms, nasrulhaq-boyce a. 2011. postharvest physicochemical and mechanical changes in jambu air (syzygium aqueum alston) fruits. australian journal of crop science: 5(1):32-38. tilman g, georg l, andreu p, catharine ra, wilhelm s, guangweng t, david t, shi-an y, hans kb. 2010. β-carotene is an important vitamin a source for humans. journal nutrition 140 (12) 2268s-2285s. usda 1991. united states standards for grades of fresh tomatoes. usda, agricultural market service,washington, dc. on 29042017@6:32pm. http://jn.nutrition.org/search?author1=tilman+grune&sortspec=date&submit=submit http://jn.nutrition.org/search?author1=georg+lietz&sortspec=date&submit=submit http://jn.nutrition.org/search?author1=andreu+palou&sortspec=date&submit=submit http://jn.nutrition.org/search?author1=a.+catharine+ross&sortspec=date&submit=submit http://jn.nutrition.org/search?author1=wilhelm+stahl&sortspec=date&submit=submit http://jn.nutrition.org/search?author1=guangweng+tang&sortspec=date&submit=submit http://jn.nutrition.org/search?author1=david+thurnham&sortspec=date&submit=submit http://jn.nutrition.org/search?author1=shi-an+yin&sortspec=date&submit=submit http://jn.nutrition.org/search?author1=hans+k.+biesalski&sortspec=date&submit=submit ruhuna journal of science vol 8: 76-89, december 2017 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v8i2.28 university of ruhuna  faculty of science, university of ruhuna 76 sri lanka determination of the microbiological quality and proximate composition of fermented cassava food products sold in ilorin-west local government area, nigeria c.o. adetunji1*, s.a. akande 2, a.k. oladipo 2, r.a. salawu 2, a.f. onyegbula 2 1landmark university, department of microbiology, laboratory of microbiology, biotechnology and nanotechnology, omu-aran, nigeria. 2nigerian stored products research institute, km 3, asa-dam road, p.m.b. 1489, ilorin, nigeria. *correspondence: adetunjicharles@gmail.com received: 19th september 2017, revised: 21st november 2017, accepted: 25th december 2017 abstract in the present study, the microbiological safety and the proximate analyses of five urban markets within ilorin-west local government area, kwara state, nigeria were carried out using standard protocols. the bacterial load of fermented staple products from cassava ranged from 0.1 to 10.9×105 cfu/g while the fungi and yeast content ranged from 1.1 to 8.2×105 cfu/g. the isolates of bacteria from all the markets include the following; pseudomonas aeruginosa, staphylococcus aureus, staphylococcus epidermis, streptococus faecalis, lactobacillus species, acetobacter spp.., bacillus cereus, escherichia coli, bacillus subtilis, lactobacillus species while the isolated fungi include fusarium oxysporium, aspergillus niger, a.flavus, a. fumigatus, saccharomyces cerevisae, candida albicans, penicillum spp., rhizopus stolonifera, mucor spp..the results of the proximate composition showed that moisture content of fermented staple products of cassava ranged from 6.21% (garri ijebu from market a and lebu from market c) to 72.25% (fufu from market c) while dry matter content ranged between 27.75% (fufu from market c) to 93.79% (garri ijebu market a and lebu from market c). ash content ranged from 0.23% (tapioca from market a) to 1.96% (lebu from market a), crude fibre content ranged between 1.13% (fufu from market c) and 5.28% (abacha from market d), and the carbohydrate content of the fermented staple products from cassava ranged from 18.61% (fufu from market c) to 81.44% (tapioca from market a). even though some potential pathogenic bacteria like e.coli and bacillus were isolated from cassava fermented products, the minimum microbial load obtained could not impose any health risk. keywords: cassava, fermented products, microbial loads, proximate analysis. c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 77 1 introduction in africa, cassava has been reported among the most crucial food (fao, 1999), and nigeria is the largest producer of cassava in the world (fao, 2017). oyewole (1991) stated that cassava root needs to be processed. moreover, in nigeria, majority of people prefer to consume cassava in fried, baked or in boiled form after fermenting the raw cassava crop. burns et al. (2012) stated that cassava root needs to be processed so as to remove available poisonous substances in the tuber. examples of staple fermented foods obtained from processed cassava tuber include garri, fufu, lafun, pupuru and tapioca (lancaster et al., 1982). previous researchers discovered that diverse microorganisms are important in the process of fermenting the raw cassava root tuber including bacteria, yeasts and moulds (tamang et al. 2016; capozzi et al. 2017). oyewole and odunfa (1988) discovered that bacillus spp., lactobacillus spp., geotrichum spp. and aspergillus spp. were present in the fermentation of lafun. moreover, hahn (1989) discovered that several fungi including aspergillus spp., penicillium spp., mucor spp., fusarium spp., curvularia spp. and cladosporium spp. were isolated from garri during storage. their study, therefore, necessitates the great need in ensuring the microbiological safety of the staple fermented foods obtained from processed cassava tuber. the quality of food defines the amount of available nutrients that could be derived from staple food products. majorly, most of the fermented staple food obtained from cassava include garri, fufu and lafun (oyewole and odunfa 1988). garri, a dry food is largely consumed by most people from africa without cooking and can be consumed with the addition of sweeteners (oyewole et al. 2001). the preparation of garri includes the detaching of the root, washing, crushing, fermentation, removal of water and, roasting (oyewole and sanni 1995). fufu is another major staple fermented food from cassava that is consumed by most people in africa (lancaster et al. 1982, okafor et al. 1984, sanni 1989, longe 1990, oyewole 1995), and the processing includes peeling, breaking into small sizes, soaking, pulping, screening, sedimentation, removal of water and production of cassava paste (oyewole et al. 2001). the processing of lafun is almost similar to the preparation of fufu but the fermented cassava needs to be sun-dried and milled before consumption (oyewole and sanni 1995) therefore, the present study intends to establish the micro flora of these popular cassava products in nigeria, and to determine their proximate content and microbiological safety to consumers. in addition, the obtained result will assist the policy makers in formulating necessary quality hazard, storage techniques as well as processing line necessary for the production of fermented staple food from cassava in nigeria and west africa at large. c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 78 2 material and methods 2.1 reagents and test samples all reagents used were of analytical grade obtained from sigma-aldrich, germany and bdh, england. these include; sodium hydroxide pellets, tetraoxosulphate (vi) acid, boric acid, anhydrous sodium tetraoxosulphate (vi), copper tetraoxosulphate (vi) pentahydrate, selenium powder, hydrochloric acid and ethanol (95% v/v). all cassava products displayed in each markets were sampled using simple random sampling techniques, these include; white garri (white; cassava flake) yellow garri (yellow; cassava flake), ‘lebu’ (fine cassava flake), ‘garri ijebu’(another variety of cassava flake), ‘igbodo’ (edible cassava flour, whole root), ‘lafun’ (edible cassava flour, pelletized), tapioca (starch extract), ‘fufu’ (a dough from cassava) and ‘abacha’ (african salad). five urban markets (designated as a, b, c, d and e) within ilorin-west local government area, kwara state in nigeria were used for sample collection. a total of 256 samples (a=70, b=57, c=50, d=55 and e=24) were collected from all the markets depending on their sizes and numbers of available sellers. prior to analyses in the laboratory, each product was bulked together to give a representative sample from each market, after which 29 samples were obtained for microbiological and proximate analyses. 2.2 determination of the proximate composition proximate analyses including moisture content, ash, and crude fibre were carried out on the fermented staple products from cassava samples following the aoac (2000) methods while the total carbohydrates content was analyzed following the method described by osborn and vogt (1978). the dry matter content was calculated from moisture content (mc) using the relation; dry matter (%) = 100 – mc 2.3 microbiological analyses isolation of microorganisms five gram of the selected fermented staple products from cassava were weighed and mixed with 250 ml of sterilized normal saline followed by serial dilution. aliquot (1 ml) from the resulting mixture was placed in a sterile petri dish using pour plate method in duplicate. nutrient agar (product of hi media, ref. moo2-5005) was used for the determination of bacteria, incubated at 37oc for 24–48 h while potato dextrose agar (product of hi media, ref. gm096-500g) was used for the determination of fungus in the c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 79 samples, incubated at 27oc for 48–72 h. they were procured from lab trade nigeria ltd, ilorin, kwara state, nigeria. the total number of available colonies obtained after incubation was counted and expressed in cfu/g. the pure cultures of the isolates were acquired by sub-culturing on newly prepared agar plates. identification of microbial isolates the cultural and biochemical characteristics of the pure culture of the isolates obtained were enumerated using the protocol developed by adetunji and adejumo, 2017; adetunji et al. (2012) and uzeh et al. (2009). the following cultural, morphological and biochemical features was determined; cellular shape, colonial elevation, colonial edge, colonial opacity, colonial pigmentation, cellular arrangement gram’s staining, motility test, spore staining, capsule staining, catalase test, methyl red test, starch hydrolysis, citrate utilization, and oxygen reaction. 2.4 statistical analysis of data results were expressed as mean ± sd of triplicates (n = 3) determinations. all data generated were analyzed by one-way analysis of variance (anova) using the spss statistics for windows version 20.0.0 (ibm spss statistics, ibm corporation 2011, armonk, ny. usa). the means were separated using new duncan multiple range tests. significance was accepted at 5% probability level (p<0.05). 3 results and discussion the results of proximate composition (table 1) showed that the moisture contents of fermented cassava food products from ilorin-west urban markets ranged from 6.21% (garri ijebu from market a and lebu from market c) to 72.25% (fufu from market c), while dry matter content ranged from 27.75% (fufu from market c) to 93.79% (garri ijebu from market a and lebu from market c). ash contents ranged from 0.23% (tapioca from market a) to 1.96% (lebu from market a), crude fibre content ranged from 1.13% (fufu from market c) to 5.28% (abacha from market d), the total available carbohydrate contents of the fermented cassava food products ranged from 18.61% (fufu from market c) to 81.44 (tapioca from market a). c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 80 table 1: proximate composition of fermented cassava food products sold in ilorin-west urban markets, nigeria (wet basis). market sample name moisture content (%) ash (%) crude fibre (%) carbohydrates (%) dry matter (%) a yellow garri 9.45 ± 0.07 gh 1.39 ± 0.03 efg 1.51 ± 0.01 n 75.65 ± 0.32 fg 90.55 ± 0.07 hi white garri 8.83 ± 0.18 ij 1.88 ± 0.09 ab 1.78 ± 0.01 k 77.17 ± 0.54 de 91.17 ± 0.18 g igbodo 8.38 ± 0.23 jk 1.52 ± 0.40 cdef 1.48 ± 0.08 no 76.81 ± 0.31 ef 91.63 ± 0.23 f lafun 8.36 ± 0.11 jk 1.14 ± 0.05 hi 1.03 ± 0.05 s 77.27 ± 0.57 de 91.64 ± 0.11f lebu 6.69 ± 0.13 op 1.96 ± 0.16 a 1.39 ± 0.10 p 77.30 ± 0.66 de 93.31 ± 0.13 b fufu 69.39 ± 0.59 c 0.80 ± 0.00 j 1.57 ± 0.03 m 23.44 ± 0.03 o 30.61 ± 0.59 m gari ijebu 6.21 ± 0.10 p 1.79 ± 0.05 abc 1.57 ± 0.03 m 73.10 ± 0.33 jk 93.79 ± 0.10 a tapioca 6.74 ± 0.49 o 0.25 ± 0.03 m nd 81.44 ± 0.32 a 93.26 ± 0.49 b b yellow gari 13.35 ± 0.06 e 1.03 ± 0.43 ij 1.64 ± 0.04 l 70.93 ± 0.57 l 86.65 ± 0.06 k white garri 9.01 ± 0.20 hi 1.47 ± 0.10 efg 3.69 ± 0.01 c 71.30 ± 0.58 l 90.99 ± 0.20 gh igbodo 9.34 ± 0.20 ghi 1.62 ± 0.04 cde 2.65 ± 0.03 e 77.15 ± 0.55 de 90.66 ± 0.20 hi lafun 9.76 ± 0.11 g 1.51 ± 0.06 cdef 2.09 ± 0.02 h 73.47 ± 2.50 ijk 90.24 ± 0.11i fufu 70.81 ± 0.57 b 0.53 ± 0.12 k 1.25 ± 0.05 q 21.04 ± 0.41 p 29.53 ± 0.01n tapioca 7.79 ± 0.35 lm 0.32 ± 0.01 kl nd 80.85 ± 0.56 a 92.21 ± 0.35 de c yellow garri 9.37 ± 0.55 gh 0.30 ± 0.02 kl 1.64 ± 0.05 l 68.75 ± 0.54 m 90.63 ± 0.55 hi white garri 11.40 ± 0.10 f 0.24 ± 0.05 m 3.94 ± 0.02 b 72.66 ± 0.59 k 88.60 ± 0.10 j igbodo 8.35 ± 0.14 jk 1.60 ± 0.01 cde 2.63 ± 0.03 e 80.68 ± 0.57 a 91.65 ± 0.14 f lafun 9.13 ± 0.36 hi 1.22 ± 0.04 ghi 2.43 ± 0.08 g 75.03 ± 1.14 gh 90.87 ± 0.36 gh fufu 72.25 ± 0.40 a 0.49 ± 0.09 kl 1.13 ± 0.02 r 18.61 ± 0.02 q 27.75 ± 0.40 o lebu 6.21 ± 0.17 p 1.59 ± 0.14 cde 1.83 ± 0.01 jk 78.07 ± 0.62 cde 93..79 ± 0.17 a d yellow garri 8.28 ± 0.18 kl 1.65 ± 0.21 bcde 1.89 ± 0.01 i 74.77 ± 0.63ghi 91.72 ± 0.18 f white garri 8.21± 0.15 klm 1.54 ± 0.06 cdef 3.65 ± 0.01 c 80.37 ± 0.85 ab 91.79 ± 0.15 ef igbodo 7.27 ± 0.06 n 1.75 ± 0.07 abcd 2.55 ± 0.05 f 80.74 ± 0.32 a 92.73 ± 0.06 c lafun 7.94 ± 0.14 klm 1.03 ± 0.06 ij 2.89 ± 0.02 d 74.46 ± 0.57 ghij 92.06 ± 0.14 def abacha 9.00 ± 0.10 hi 0.32 ± 0.07 kl 5.28 ± 0.03 a 78.59 ± 0.33 cd 91.00 ± 0.10 gh fufu 67.72 ± 0.29 d 0.28 ± 0.02 kl 1.86 ± 0.02 ij 28.18 ± 2.16 n 32.28 ± 0.29 i e white garri 7.72 ± 0.07 mn 1.48 ± 0.09 defg 1.66 ± 0.02 l 78.45 ± 0.32 cd 92.28 ± 0.07 d igbodo 9.15 ± 0.08 hi 1.60 ± 0.01 cde 1.61 ± 0.03 lm 79.14 ± 0.31 bc 90.85 ± 0.08 gh lafun 8.22 ± 0.55 klm 1.32 ± 0.27 fgh 1.43 ± 0.02op 74.05 ± 0.87 hijk 91.78 ± 0.55 ef *means within the same column with unshared superscript letters are significantly different (p<0.05), nd= not detected. c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 81 the carbohydrate content obtained from white garri from market d had the highest value of 80.37 % among all the garri samples tested during this study. however, a higher carbohydrate content of 85.8 % was reported by okolie et al. (2012), from garri samples sold in lagos metropolis, nigeria. the difference in the carbohydrate content might be due to the preliminary loss of soluble carbohydrate in the previous unit processing of the cassava tubers. the reduction in the carbohydrate content from the garri samples might also be due to the fact that the isolated microorganisms have the tendency to utilize the garri sample as source of carbon source for their usage and sustainability (colehour et al. 2014, oyeyiola et al. 2014). fufu (market c) was significantly higher (p<0.05) in moisture content of 72.25 % than all the other fermented cassava food products. this is expected to be so because fufu is a wet product with rapid spoilage in 3 to 4 days. this same product (fufu from market c) was significantly (p<0.05) lower in crude fibre, total carbohydrates and dry matter contents. during this study, fufu from market c recorded a higher moisture content of 72.25% compared to the moisture content of 58.80% recorded by omosuli et al. (2017), from fufu prepared from cassava flour in their study. the different variation recorded in the moisture content of fufu from the various studies might be due to the fact that some level of transpiration occurs after the processing of cassava tubers into fufu (ikujenlola and opawale 2007). the variation in the moisture content observed from the various fermented cassava product might be due to their ability to absorb moisture during storage condition which consequently supports the colonization of these spoilage fungi as well as increases their deteriorative capabilities. the moisture absorbed by the fermented cassava products enhanced the biodegradability potential of these microorganisms (jonathan et al. 2016). the nigerian industrial standards (nis) 1988, state that the moisture content of garri should not exceed 7% (w/w), ash content should not exceed 1.5% (w/w) and crude fibre should not exceed 2% (w/w) (sanni et al. 2015). only three garri products were within the stipulated limit and they are; lebu (market a and c) and garri ijebu (market a). the nis standards for cassava chips stipulate that the moisture content should not exceed 13% (w/w), crude fibre 2.0% maximum and ash content 3.0% maximum. the cassava chips samples in the present project (lafun i and ii) showed some level of compliance to the nis standards, except in a few cases where the crude fibre contents were above the maximum content of 2.0%, these are; lafun i (market b, c and d), lafun ii (market b, c and d). the results of microbiological characteristics of the fermented cassava food samples from ilorin-west local government urban markets is as shown in table 2. https://www.ncbi.nlm.nih.gov/pubmed/?term=colehour%20am%5bauthor%5d&cauthor=true&cauthor_uid=25071997 c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 82 table 2: microbial load of fermented cassava food products sold from ilorin-west urban markets (a-e), nigeria (mean ± sd, n=3 samples; na= nutrient agar, pda = potato dextrose agar). market sample name na (×104 cfu/g) pda (×105 cfu/g) a yellow gari 2.0 ± 0.1 d 1.9 ± 0.1 f white gari 1.9 ± 0.1 de 1.5 ± 0.8 g igbodo 0.5 ± 0.0 hi 3.0 ± 0.1e lafun 0.3 ± 0.0 ijkl 5.0 ±0.0 c lebu 0.9 ± 0.1g 3.0 ± 0.0 e fufu 5.5 ± 0.1 b 1.0 ± 0.1 gh gari ijebu 0.3 ± 0.0 ijkl 0.1 ± 0.0 m tapioca 0.4 ± 0.1 hijk 1.0 ± 0.1 gh b yellow gari 0.5 ± 0.1 hi 5.9 ± 0.1 b white gari 1.0 ± 0.0 g 3.1 ± 0.1 e igbodo 0.2 ± 0.0 lm 2.0 ± 0.1 f lafun 0.3 ± 0.0 ijkl 2.1 ± 0.1 f fufu 0.8 ± 0.1g 4.1 ± 0.2 d tapioca 2.9 ± 0.1c 5.2 ± 0.3 c c yellow gari 0.9 ± 0.1g 2.0 ± 0.0 f white gari 0.6 ± 0.1 h 1.1 ± 0.1 gh igbodo 0.3 ± 0.1 ijkl 2.9 ± 0.1 e lafun 0.2 ± 0.1klm 8.2 ±0.3 a fufu 0.3 ± 0.1 jklm 0.9 ± 0.1 h lebu gari 0.1± 0.0 m 2.0 ± 0.1 f d yellow gari 0.3 ± 0.1 jklm 1.1 ± 0.1 gh white gari 2.0 ± 0.1 d 4.1 ± 0.1 d igbodo 0.6 ± 0.1 h 1.2 ± 0.2 gh lafun 5.4 ± 0.1 b 5.0 ± 0.1 c fufu 10.9 ± 0.1 a 8.1 ± 0.1 a abacha 10.8 ± 0.0 a 6.0 ± 0.1 b e white gari 1.8 ± 0.1 e 6.1 ± 0.1 b igbodo 0.5 ± 0.1 hij 4.0 ± 0.1 d lafun 1.3 ± 0.1 f 6.1 ± 0.1 b *means within the same column with unshared superscript letters are significantly different (p<0.05) the bacterial load of fermented food samples ranged from 0.1–10.9×104 cfu/g, while the fungi and yeast ranged 1.1–8.2×105 cfu/g. the highest bacterial load was observed from market d from lafun with microbial loads of 5.4± 0.1×104 cfu/g, while the lowest was observed from market c from lebu garri with bacterial counts of 0.1±0.0×104 cfu/g (table 2). the bacterial load observed during this study were of lower value when compared with the findings of adebayo-oyetoro et al. (2013), who reported a bacterial load of 8.1×106 cfu/g from the fermented cassava flour sampled from the different market during their study. also, the fungi and yeast count found during this study showed a lower value when compared to the value of 3.5×106 cfu/g observed by oyeyiola et al. (2014), from fermented cassava food products sold in oyo town, oyo state, nigeria. c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 83 table 3: m a rk e t s a m p le s c e ll u la r sh a p e (r : r o d , c : c o c c i) c o lo n ia l e le v a ti o n ( r : r a is e d , f : f la t) c o lo n ia l e d g e ( e : e n ti re , l : l o b a te ) c o lo n ia l o p a c it y (t : t ra n sl u c e n t, o : o p a q u e ) c o lo n ia l su rf a c e (s : s m o o th , r : r o u g h , d : d u ll ) c o lo n ia l p ig m e n ta ti o n ( y : y e ll o w , c : c re a m , c w : c re a m y w h it e , y c : y e ll o w is h c re a m , w : w h it e ) c e ll u la r a rr a n g e m e n t (c : c h a in , c l : c lu st e rs , s : s in g le ) g ra m ’s st a in in g m o ti li ty t e st s p o re s ta in in g c a p su le s ta in in g c a ta la se t e st m e th y l re d t e st s ta rc h h y d ro ly si s c it ra te u ti li z a ti o n o x y g e n r e a c ti o n action on simple carbohydrates probable microorganism l a c to se g lu c o se s u c ro se m a lt o se f ru c to se a yellow garri r r e t s yc c + + ae a a a ag pseudomonas aeruginosa white garri c r e o s cw cl + + + + fan ag a a a staphylococcus aureus igbodo c r e o s cw cl + + + + fan ag a a a s. aureus lafun c r e o s cw cl + + + + fan ag a a a s. aureus lebu c r e o s cw cl + + + + fan ag a a a s. epidermis fufu c r e o s cw cl + + + + fan ag a a a streptococcus faecalis garri ijebu r r l o s cw cl + fan a a a a ag lactobacillus sp tapioca r r l t r c s an a a acetobacter sp b yellow garri c r e o s cw cl + + + + fan ag a a a s. aureus white garri c r e o s cw cl + + + + fan ag a a a s. epidermis igbodo c r e o s cw cl + + + + fan ag a a a s. aureus lafun c r e o s cw cl + + + + fan ag a a a s. epidermis fufu c r e o s cw cl + + + + fan ag a a a streptococcus faecalis tapioca r r l t d c c + + + + + + + ae ag a ag ag bacillus cereus continued.. c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 84 table 3 continued. m a rk e t s a m p le s c e ll u la r sh a p e (r : r o d , c : c o c c i) c o lo n ia l e le v a ti o n ( r : r a is e d , f : f la t) c o lo n ia l e d g e ( e : e n ti re , l : l o b a te ) c o lo n ia l o p a c it y (t : t ra n sl u c e n t, o : o p a q u e ) c o lo n ia l su rf a c e (s : s m o o th , r : r o u g h , d : d u ll ) c o lo n ia l p ig m e n ta ti o n ( y : y e ll o w , c : c re a m , c w : c re a m y w h it e , y c : y e ll o w is h c re a m , w : w h it e ) c e ll u la r a rr a n g e m e n t (c : c h a in , c l : c lu st e rs , s : s in g le ) g ra m ’s st a in in g m o ti li ty t e st s p o re s ta in in g c a p su le s ta in in g c a ta la se t e st m e th y l re d t e st s ta rc h h y d ro ly si s c it ra te u ti li z a ti o n o x y g e n r e a c ti o n action on simple carbohydrates probable microorganism l a c to se g lu c o se s u c ro se m a lt o se f ru c to se c yellow garri c r e o s cw cl + + + + fan ag a a a s. aureus white garri r r e t s c c + + + fan ag ag ag e.coli igbodo c r e o s cw cl + + + + fan ag a a a s. aureus lafun c r e o s cw cl + + + + fan ag a a a s. epidermis fufu c r e o s cw cl + + + + fan ag a a a s. aureus lebu garri c r e o s cw cl + + + + fan ag a a a s. epidermis d yellow garri c r e o s cw cl + + + + fan ag a a a s. aureus white garri r f l o d w cl + + + + + + fan ag a a a a bacillus subtilis igbodo r r e o s cw cl + + + + fan ag a a a streptococus faecalis lafun r r e t s c c + + + fan ag ag ag e.coli fufu r r l o s cw cl + fan a a a a ag lactobacillus sp abacha r r e o s y s + fan a a a a salmonella spp. e white garri c r e o s cw cl + + + + fan ag a a a s. aureus igbodo c r e o s cw cl + + + + fan ag a a a s. epidermis lafun c r e o s cw cl + + + + fan ag a a a s. aureus c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 85 the isolates of bacteria from all the market included pseudomonas aeruginosa, staphylococcus aureus, staphylococcus epidermis, streptococcus faecalis, lactobacillus species, acetobacter spp., bacillus cereus, e. coli, bacillus subtilis, lactobacillus spp. (table 3). the isolated microorganisms present in the fermented cassava food product have also been observed by ogiehor and ikenebomeh (2005). this might be resulting from processing and poor handling practices where these products are sold and can further increase microbial hazards from food products from cassava (oyeyiola et al. 2014). it was observed that staphylococcus spp. was isolated from all the markets, and similar trends have been reported earlier by researchers that had worked on different varieties of garri from nigeria (ogiehor et al. 2007, adetunji et al. 2012, olopade et al. 2014). they are normally associated with individual hygiene of the people who handle them in market places. (aboloma 2008, oyeyi and lum-nwi 2008, shamsuddeen and ameh 2008, wada-kura et al. 2009). kim et al. (2009) also discovered that p. aeruginosa, an opportunistic pathogen, which is responsible for bacteremia and gastrointestinal infections in affected individuals. e. coli has been documented to be responsible for diarrhoea in human being (nweze 2010). some bacillus spp. are known to be accountable for food poisoning and their presence as a biological hazard in garri products which are consumed raw have raised many concerns with cassava products (aboloma 2008). the isolated fungi included fusarium oxysporium, aspergillus niger, a. flavus, a. fumigatus, saccharomyces cerevisae, candida albicans, penicillum spp., rhizopus stolonifera and mucor spp. (table 4 and 5). some of the isolated fungi were identified to be spoilage inducing on foods during storage (homer et al. 1994). the microbiological attributes of the raw cassava to be processed should be clean and of good quality. the usage of starter cultures that have antimicrobial properties with the capability to detoxify, in addition to sustaining adequate and clean surrounding is recommended. also, sterilised packaging material should be used in packing the processed fermented food from cassava (ikujenlola and opawale 2007). the presence of fungi on various fermented cassava food products examined during this study might be due to the fact that the environmental condition favors their rate of sporulation (olopade et al. 2014). also, oranusi and braide (2012), reported in their study that some fungus species like penicillium, fusarium, and aspergillus isolated from different samples during this study could produce poisonous substances which may be toxigenic when exposed to a favorable environmental condition. the occurrence of aspergillus species found in most of the fermented cassava food products from all the markets during this study might be due to the various methods used in the cassava processing. the same observations was reported by c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 86 jonathan et al. (2017) who discovered the presence of aspergillus species from cassava products. table 4: distribution of different isolated fungus from fermented cassava food products sold from ilorin-west urban markets, nigeria. market sample name distribution of fungus from different market a yellow garri fusarium spp., mucor spp., a. niger, saccharomyces spp., a. fumigatus rhizopus spp., penicillum spp., aspergillus spp. aspergillus niger, a. fumigatus, fusarium spp., mucor spp. aspergillus niger, a. flavus, rhizopus spp., penicillum spp., mucor spp. fusarium spp., rhizopus spp., penicillum spp., aspergillus spp. saccharomyces spp., candida spp. a. niger, saccharomyces spp., a. fumigatus, mucor spp. rhizopus spp., penicillum spp., aspergillus spp., fusarium spp. aspergillus niger, a. fumigatus, penicillum spp., fusarium spp. penicillum spp., aspergillus spp., mucor spp. aspergillus niger, a. flavus, rhizopus spp., penicillum spp. penicillum spp., aspergillus spp., fusarium spp. saccharomyces spp., candida spp., aspergillus niger penicillum spp., aspergillus spp., fusarium spp., saccharomyces spp. mucor spp., saccharomyces spp., aspergillus fumigatus rhizopus spp., penicillum spp., aspergillus spp. mucor spp., aspergillus niger, a. fumigatus, fusarium spp., aspergillus flavus, rhizopus spp., penicillum spp., mucor spp. saccharomyces spp., candida spp. rhizopus spp., penicillum spp., aspergillus spp. aspergillus fumigatus, mucor spp., a. niger fusarium spp., penicillum spp., aspergillus spp., mucor spp. aspergillus niger, a. fumigatus, fusarium spp., mucor spp. rhizopus spp., penicillum spp. saccharomyces spp., candida spp. aspergillus fumigatus, a. niger, saccharomyces spp., mucor spp. rhizopus spp., penicillum spp., aspergillus spp. aspergillus flavus, rhizopus spp., penicillum spp. aspergillus flavus, rhizopus spp., penicillum spp. white garri igbodo lafun lebu fufu garri ijebu tapioca b yellow garri white garri igbodo lafun fufu tapioca c yellow garri white garri igbodo lafun fufu lebu garri d yellow garri white garri igbodo lafun fufu abacha e white garri igbodo lafun 4 conclusions this study has established the microbiological quality and proximate composition of fermented cassava food products sold in ilorin-west local government area, nigeria. during this study, various pathogenic microorganisms constituting biological hazards were highlighted from different markets where fermented cassava food products were surveyed. therefore, there is a need to prevent these foods from contaminants especially the biological hazards from air and during processing stages. the seller should wear gloves during production and sales. more studies with applications of modern techniques that involve utilization of nanotechnology and irradiation for preservation of fermented foods are required to ensure food safety further. c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 87 table 5: cultural and morphological characterization of the isolated fungi from fermented cassava food products sold from ilorin-west urban markets, nigeria. description probable fungus colonies have aerial mycelium with whitish or peach colour; conidiophores are usually short branched on phialides. fusarium oxysporium the colony usually contains black conidiophore. conidial heads, radiate. conidiophore stipe smooth-walled, hyaline but often in brown colour. vesicles globose to sub-globose. aspergillus niger colonies have yellow-green conidiophores. conidiophores have stipes with smooth-walled hyaline but often in brown colour. aspergillus flavus colonies have dense felt of yellow-green conidiophores. conidia heads typically radiate latter splitting in several loose columns, yellow-green becoming dark yellow-green. sclerotia are brown to black. aspergillus fumigatus colonies extent quickly and developed within three days. they have flat, moist, glittering or dull, and cream in color. blastoconidia are present. saccharomyces cerevisae colonies are whitish-cream in color, smooth, glabrous and yeast-like in appearance. presence of spherical to sub spherical blastoconidia. candida albicans colonies grow and sporulate with yellow or brown-green conidiophores with 3-6 phalides. phalides often solitary, cylindrical with a short neck. penicillum spp. colonies have whitish color becoming grayish-brownish. sporangiophores are colourless to dark brown, rough-walled stolons opposite the branched rhizoids. it has sporangia with sub-globose, ovoid, with blackish-brown color at maturity. rhizopus stolonifera the colony was white and woolly. the hypha were thick and non-septate, columella were round. the sporagiopores departs laterally from mycelium, the sporangia were filled with spores. mucor spp. good agricultural and manufacturing practices should be encouraged during post-harvest, processing, and marketing of raw cassava crop. the food processor should adopt good and general hygiene and apply haccp during production of fermented cassava food products. the local cassava processing centers should encourage the usage of standard operating procedure (sop) to minimize the level of biological hazards and any available contamination from the raw material and the environment. references aoac. 2000. official methods of analysis of aoac international, 17th ed. published by the association of official of analytical chemists international suite 400 2200, wilson boulevard, arlington, virginia 22201-3301 usa. aboloma ri. 2008. microbiological analysis of bread samples from bakery to sale points in ado-ekiti, ekiti state, nigeria. biological and environmental sciences journal for the tropics 5(3): 77 – 81. adetunji co, adejumo io. 2017. nutritional assessment of mycomeat produced from different agricultural substrates using wild and mutant strains from pleurotus sajor-caju during solid state fermentation. animal feed science and technology 224:14-19. 10.1016/j.anifeedsci.2016.12.004 https://en.wikipedia.org/w/index.php?title=blastoconidia&action=edit&redlink=1 c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 88 adetunji co, olaleye oo, fawole ob, williams jo, oloke jk, adetunji jb, kazeem mo. 2012. microbiology and moisture content studies of gari stored in two different packaging materials. science focus 17(1):110-115. adebayo-oyetoro ao, oyewole ob, obadina ao, omemu ma. 2013. microbiological safety assessment of fermented cassava flour ‘‘lafun’’ available in ogun and oyo states of nigeria. international journal of food science http://dx.doi.org/10.1155/2013/845324 burns ae, gleadow rm, zacarias am, cuambe ce, miller re, cavagnaro tr. 2012. variations in the chemical composition of cassava (manihot esculenta crantz) leaves and roots as affected by genotypic and environmental variation. journal of agricultural and food chemistry 60: 4946−4956. dx.doi.org/10.1021/jf2047288. capozzi v, fragasso m, romaniello r, berbegal c, russo p, spano g. 2017. spontaneous food fermentations and potential risks for human health. fermentation 3(49): 1-19. doi: 10.3390/fermentation3040049 colehour am, meadow jf, liebert ma, cepon-robins tj, gildner te, urlacher ss, bohannan bjm, snodgrass jj, lawrence ss. 2014. local domestication of lactic acid bacteria via cassava beer fermentation. peer j 2: e479; 1-18. doi 10.7717/peerj.479. fao. 1999. fao world book. food and agricultural organization of the united nations, rome. fao. 2017. fao in nigeria, nigeria at a glance. accessed on november 17, 2017. http://www.fao.org/nigeria/fao-in-nigeria/nigeria-at-a-glance/en/ hahn sk. 1989. an overview of africa traditional cassava processing utilization. outlook on agriculture 18: 110 – 118. homer we, lehrer sb, salvaggio je, banks de, weissmmanb dn. 1994. ‘fungi’. indoor air pollution: an allergy perspective. immunology and allergy clinics of north america 14: 551 – 566. ikujenlola av, opawale bo. 2007. effects of processing on the yield and physico-chemical properties of cassava products. advanced material research, 18: 165 -170. jonathan sg, adeniyi ma, asemoloye md. 2016. nutrient value, fungal biodeterioration and aflatoxin contamination of suya spices a novel nigerian indigenous snack. hindawi scientifica 2016: id 4602036. jonathan sg, bello ts, asemoloye md. 2017. food values, spoilage moulds and aflatoxin detection in ‘attiéké’ (a cassava fermented product). journal of microbial and biochemical technology 9:244-248. doi: 10.4172/1948-5948. 1000372. kim ky, kimb ht, kim d, nakajimad j, higuchi t. 2009. distribution characteristics of airborne bacteria and fungi in the feedstuff manufacturing factories. journal of hazardous materials 169: 1054 – 1060 pp. lancaster pa, ingram js, lim my, coursey dg. 1982. traditional fermented staple products from cassava foods; survey of processing techniques. economic botany 36(1): 12 – 48 pp. longe og. 1990. effects of processing on the chemical composition and energy value of cassava. nutrition reports international 21(6): 819 – 828 pp. nigerian industrial standard. 1988. standard for garri. standard organization of nigeria. federal ministry of industries, lagos 10. nweze ei. 2010. aetiology of diarrhoea and virulence properties of diarrhoea genic escherichia coli among patients and healthy subjects in southeast nigeria. journal of health, population and nutrition 28 (3): 245 – 252 pp. ogiehor is, ikenebomeh mj. 2005. extension of shelf life of gari by hygienic handling and sodium benzoate treatment. african journal of biotechnology 4(7): 744 – 748 pp. ogiehor is, ikenebomeh mj, ekundayo ao. 2007. the bioload and aflatoxin content of market garri from some selected states in southern nigeria: public health significance. african health sciences 7(4): 223–227. okafor n, ijioma b, oyolu c. 1984. studies of microbiology of cassava retting for foo-foo production. journal of applied bacteriology 56(1): 1 – 13. c.o. adetunji et al. fermented cassava food products sold in ilorin-west, nigeria ruhuna journal of science vol 8: 76-89, december 2017 89 okolie np, brai m, natoyebi om. 2012. comparative study on some selected garri samples sold in lagos metropolis. journal of food studies 1(1):1-10. olopade bk, oranusi s, ajala r, olorunsola sj. 2014. microbiological quality of fermented cassava (gari) sold in ota ogun state nigeria. international journal of current microbiology and applied sciences 3(3): 888-895. omosuli sv, ikujenlola av, abisuwa at. 2017. quality assessment of stored fresh cassava roots and ‘fufu’ flour produced from stored roots. journal of food science and nutrition therapy 3(1): 9-13. oranusi us, braide w. 2012. a study of microbial safety of ready-to eat foods vended on highways: onitsha-owerri, south east nigeria. international research journal of microbiology 32: 66-71. oyewole ob, odunfa sa. 1988. microbiological studies on cassava fermentation for lafun production. food microbiology 5: 125 – 133. oyewole ob, sanni o, dipeolu ao, adebayo k, ayinde ia, pearce dm, white jl, tomlins k, westby a. 2001. improved processing technology for the fermentation of cassava to fufu. 11th world congress of food technologists. seoul, korea. april 23-27. oyewole ob, sanni lo. 1995. constraints in traditional cassava processing: a case of ‘fufu’ production. in: transformation alimentaire du manioc (cassava food processing). editors: tom agbor egbe, alainbrauman, dany griffon, serge treche. institut francais de recherche scientifique pour le development en cooperation (orstom), paris, france. isbn 2-7099-1279-1. pp. 523-529. sanni mo. 1989. the mycoflora of gari. journal of applied bacteriology 67: 239 – 242. sanni la, odukogbe oo, faborode mo, ibrahim ro. 2015. optimization of process parameters of a conductive rotary dryer for garri production. in: proceedings of international conference on national capacity building strategy for sustainable development and poverty alleviation. 26 – 28 may 2015. cr1006, block 10, the american university in the emirates, dubai, uae. 527-532. oyeyi ti, lum-nwi mef. 2008. bacteriological quality of some street ended foods in bayero university campuses, kano, nigeria. biological and environmental sciences journal for the tropics 5(4): 239 – 243. oyeyiola gp, oyeniyi or, arekemase mo, ahmed rn. 2014. a comparative study on the microbiological and nutritional properties of stored chips and flours. world journal of biological research 6(2):1-6. shamsuddeen u, ameh jb. 2008. survey on the possible critical control points during the production of ‘balangu’ in kano. bayero journal of pure and applied sciences, 1(1): 76 – 79. tamang jp, watanabe k and holzapfel wh .2016. review: diversity of microorganisms in global fermented foods and beverages. frontiers in microbiology 7:377. doi: 10.3389/fmicb.2016.00377. uzeh re, alade fa, bankole m. 2009. the microbial quality of pre-packed mixed vegetable salad in some retail outlets in lagos, nigeria. african journal of food science 3(9): 270 – 272. wada-kura a, maxwell rg, sadiq h-y, tijjani mb, abdullahi io, aliyu ms, adetunji oa. 2009. microbiological quality of some ready-to-eat foods and fomites in some cafeterias in ahmadu bello university, zaria. biological and environmental sciences journal for the tropics 6(1): 6 – 9 pp. microsoft word rjs-vol-ii-wickramasekara-new-1.2.doc © 2007 faculty of science university of ruhuna ruhuna journal of science vol. 2, september 2007, pp. 89-95 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x 89 abstract. aphidius rhopalosiphi is a parasitoid, which parasitizes the aphid rhopalosiphum padi on wheat and is a potential candidate in biological control programmes. the effect of honeydew excreted by host aphids a. rhopalosiphi and metapolophium dirhodum, and honeydew extracted by non-host aphids, mizus percicae and brevicoryne brassicae on searching and walking time of a. rhopalosiphi was tested. water was used as the control. walking time of a. rhopalosiphi on honeydew treated area was also investigated. when parasitoids were released, the time taken to reach the honeydew treated discs was significantly shorter than the time taken, to reach the control discs (p ≤ 0.05). the time taken by parasitoids to reach areas treated with honeydew of different aphid species did not differ significantly among the species (p ≥ 0.05). the time spent on different honeydew treated discs was significantly higher than controls (p ≤ 0.05). the time spent on non-host honeydew was significantly shorter than the time spent on the host honeydew (p ≤ 0.05). speed of movements of a. rhopalosiphi on treated area was significantly higher than nontreated area (p ≤ 0.05). these results revealed that honeydew of both hosts and non -hosts act as an attractant. distinct behaviour pattern suggests that host honeydew act as a stimulant as well. keywords: aphidius rhopalosiphi, searching behaviour, host honeydew, non-host honeydew 1. introduction aphids (hemiptera: aphidae) are one of the major groups of insect pests of wheat (emden, 1972). aphidius rhopalosiphi l. (hymenoptera: aphidiidae) is a well-known parasitoid of the aphid rhopalosiphum padi (hemiptera: aphidae) which feeds on wheat. searching behaviour of the aphid parasitoid aphidius rhopalosiphi (hymenoptera : aphidiidae) in response to honeydew excreted by aphids on different host plants m.g.v. wickremasinghe department of zoology, faculty of science, university of ruhuna, matara, sri lanka correspondence: vineetha@zoo.ruh.ac.lk 90 wickramasinghe:searching behaviour of the aphid… r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 89-95 ( 2 0 0 7 ) biological control has drawn the attention of scientists as well as policy makers (martin, et al., 2003) as an alternative to chemical control that brings about harmful effects on the environment as well as on human health. therefore, it is very important to investigate the searching behaviour of a parasitoid to find the host. searching time of a parasitoid is one of the indicators of its performances (vinson, 1978). the olfactometer studies revealed (wickremasinghe and van emden, 1992) that several species of aphid parasitoids have been attracted to the odour of the plant which is host of their aphids, more than to the odour of host aphid or their honeydew, in the absence of the host plant. however, a combination of host plant and host aphid odour proved the most attractive stimulus, of those tested. although not tested directly the magnitude of the responses suggests that honeydew was a stronger attractant than host aphids. honeydew is excess sap excreted by aphids following feeding on the host plant phloem sap. the behaviour of parasitoids on honeydew contaminated plants suggested that honeydew could be an arrestant as well as a searching stimulant. an arrestant may be defined as a chemical that congregates insects as a result of undirected kinetic reactions, either the slowing down or stopping locomotion (orthokinesis) (shorey, 1977). to investigate this further, effect of honeydew from host and non-host aphid species on searching and walking times of the parasitoids were investigated under the laboratory conditions. 2. materials and methods 2.1 aphids and host plants the four different aphid species and their host plants selected for the study were metapolophium dirhodum and rhopalosiphum padi found on wheat, brevicoryne brassicae found on brussels sprouts and myzus persicae found on tomato plants. each aphid species was reared on the relevant host plant. plants were kept in culture cages (25×35×60 cm) covered in muslin. wheat, brussels sprouts and tomato were grown in separate pots (12 cm diameter) and were infested with the relevant aphid species. 2.2 collection of honeydew to collect honeydew 10 aphids from each species were caged separately on leaves of its host plant in para film lined clip cages for 24 hours. honeydew excreted by each aphid species was collected on para film and was removed using capillary tubes. immediately after collection 2µl of honeydew was applied on to para film discs (3.5 cm diameter) and placed inside a petri dish. a hole of 1.5 diameter was made in the center of the lid of each petri dish and a perspex tube (1.5 cm wide and 6 cm long) was fixed to make a passage for introducing parasitoids in to each test chamber. 2.3 rearing of aphidius rhopalosiphi aphidius rhopalosiphi was reared on r. padi infesting the wheat variety hobbit. wheat was grown in two separate cages (25×35×60 cm). parasitized aphids were collected using a wet paint brush and placed in specimen tubes lined with cotton wool to prevent desiccation of the parasitized aphids. parasitoids, 24-48 h after emergence, were used in this experiment. wickramasinghe:searching behaviour of the aphid… 9 1 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 89-95 ( 2 0 0 7 ) behaviour of parasitoids was observed in an observation chamber. a petri dish containing a honeydew-treated para film disc was placed inside the observation chamber. the top and the four sides of this chamber were covered with opaque oil paper to prevent entry light which could disturb the behaviour of parasitoids. experiment was started by releasing a single parasitoid at the top of the tube leading to petri dish in the test chamber. a parasitoid was released by inverting the specimen tube containing the parasitoid over the opening of the central tube. time taken by parasitoids to reach the honey dew treated disc and the time spent for searching inside the honeydew treated disc was recorded. ten parasitoids were allowed to search on each honeydew treated para film disc. eight replicates were used for each treatment. discs treated with 20 µl of water were used as controls. in another experiment, whether the honeydew acts as a searching stimulant was tested. the movement of a. rhopalosiphi was observed in the treated para film discs. twenty microliters of honeydew solution was applied to a 4 cm diameter spot in 0.5µl droplets. treated discs were air dried before use. eight such discs were prepared. at the beginning of the experiment a petri dish with a honeydew-treated filter paper was placed inside the observation chamber and a single parasitoid was released into the arena, and the behaviour was observed. the movement of the parasitoid was measured by tracing the path followed by the parasitoid on the filter paper. a transparent paper was placed above the top of the petri dish and trace of the path was marked directly on the transparent paper. the time spent outside and inside honeydew treated area by each parasitoid was measured and the distance travelled was measured using the trace. ten parasitoids were allowed to search on treated discs. eight treated discs were used in each experiment. 3. results the time taken since the release of the parasitoid into the tube above the arena to reach the treated part of the test disc was significantly shorter with honeydew treated than the control (p ≤ 0.05) to water treated discs 44% of the parasitoids tested failed to react at all. the time taken by parasitoids to reach areas treated with honeydew produced by different aphid speciesdid not differ significantly (p ≤ 0.05) (table 1). table 1 : time spent on searching on host/non host honeydew by a. rhopalosiphi honeydew producing spp. time spent on searching x ± se(s) time taken to reach the test disc x ± se(s) % parasitoids reached the test chamber r. padi on wheat 39.342 ± 2.14c 42.93 ± 7.65a 100 92 wickramasinghe:searching behaviour of the aphid… r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 89-95 ( 2 0 0 7 ) m. dirhodum on wheat 36.56 ± 2.15c 44.65 ± 7.47a 100 m. persicae on tomato 14.03 ± 1.24b 51.87 ± 6.95a 100 b. brassicae on brussels sprouts 15.34 ± 1.43b 48.84 ± 5.57a 100 control (water) 9.22 ± 0.76a 181.81 ± 28.3b 56 means with different supper script letters in each column are significantly different (one way anova using log transformed data). means separated by lsd at p ≤ 0.05. searching time of a parasitoid is a good indicator of its performances. however, the time spent for searching on honeydew of r. padi and m. persicae fed on wheat were similar and significantly shorter than the time spent searching on the honeydew of b. brassicae fed on brussels sprouts and m. persicae on tomato (p ≤ 0.05). the speed of movement of a. rhopalosiphi inside the honeydew treated area was significantly lower than the out side (p ≤ 0.001) (table 2). table 2 : rate of movement of a. rhopalosiphi in the presence of honeydew location movement (mm/s) a) outside the treated area 14.14 ± 0.31a b) inside the treated area 3.24 ± 0.47b means followed by different supper script letters are significantly different at p ≤ 0.05. when a parasitoid was introduced in to the tube leading to the arena it did not immediately orient itself to the arena containing honeydew but took some time to walk in to the arena. parasitoids walking towards the area appeared excited, probing the substrate with the antennae as they walked. the time taken by the parasitoid to reach the honeydew treated disc and their behavior inside the inverted tube was similar regardless of the type of honeydew. in the control, parasitoids were not active; many of them remained on the top of the inverted specimen tube without moving out. only 56% walked inside the tube, but even those took about four times longer duration to reach the water treated compared to honeydew treated discs. all parasitoids responded positively to the honeydew treated discs, indicating the presence of an attractant. the parasitoids searched for discs treated with honeydew of their hosts (r. padi and m. dirhodum feeding on wheat) took about half a minute. abdominal protractions and frequent antennation were observed. some of the parasitoids tried to oviposit on the honeydew droplets. when parasitoids searched discs with the honeydew of m. persicae and b. brassicae fed on tomatoes and brussels sprouts respectively, frequent antennation could be observed but abdominal protractions were rarely observed. this suggests that the host honeydew could be a stimulant most probably for oviposting. wickramasinghe:searching behaviour of the aphid… 9 3 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 89-95 ( 2 0 0 7 ) 4. discussion from the observations and responses measured, it is apparent that a. rhopalosiphi responds to odour of non-host and host honeydew and could discriminate host from non host honeydew upon contact. the time spent searching discs with honeydew of r. padi and m. dirhodum was about 2.5 times greater than the time spent searching with honeydew of m. persicae and b. brassicae. in the second experiment, parasitoids walking across the untreated area showed distinct behaviourial changes upon contact with the honeydew contaminated area. when parasitoids were released on to the test disc, firstly they walked rapidly; did not show any abdominal movements or antennation and they rarely turned back. however, when it entered the honeydew treated area, frequent abdominal protractions and zigzag movements (angle turns) increased. parasitoids stopped very frequently and touched the substrate with the antennae. parasitoids moving out of honeydew treated area touched the filter paper with antennae and quickly turned back towards the area. apparently the overall retention of parasitoids on honeydew is reflected through several different mechanisms shown by the parasitoids as arrest, decreased walking speed (orthokinesis) and increased turnings (klinokinesis). honeydew therefore appears to be an arrestant, which keeps the parasitoid in the honeydew contaminated area, increasing the chance of contact of host aphids. the attraction of a. nigripes toward both non-host and hosts aphids has been reported by bouchard and cloutier (1984). a. nigrips responded positively to the odour of the honeydew of the preferred host macrosiphum euphobiae but also that of less preferred hosts such as myzus perssicae, apis nasturtii and a non host rhopalasiphum maidis. hagen et al. (1976) reported that chrysopa carnea, a predator of theiroaphis trifolii was attracted to the odour of the host honeydew; van emden and hagen (1976) subsequently found that a break down product of the tryptophan in the honeydew was the attractant, which hagen et al. (1976) reported as sucrose the arrestant. behavior of a. rhopalosiphi suggestes that attractant may be present in both host and non-host honeydew, but the stimulant is present only in host honeydew. frequent abdominal protractions by a. rhopalosiphi at host honeydew indicated that an oviposition stimulant may be present in the honeydew and this response would cause the parasitoid to increase its searching time. that abdominal protractions were rare on nonhost honeydew indicated that the oviposition stimulant was either absent or most probably present in smaller amounts. such differences between host and non-host honeydew might be expected to arise from diversity of chemicals present in the honeydew of aphids feeding on different plants (auclair, 1958, 1963; sidhu and patton, 1970). similar changes in the locomotory behaviour following contact with kairomones have been found for other parasitoids (chiri and leghan, 1982; strand and vinson (1982); vet and van der hoven, 1984; waage, 1978). strand and vinson (1982) reported increased walking speed of cardiochiles nigriceps. a behavioural response similar to that shown by a. rhopalosiphi in this study, has been reported by waage (1978) for locomotory response of the ichneumoid nemeritis canescence to the mandibular gland secretion of its host plodia interpucntella. in response to the presentation of the host chemical on a surface, a walking nemeritis exhibited a complex orthokinetic responses involving stopping, walking at a 94 wickramasinghe:searching behaviour of the aphid… r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 89-95 ( 2 0 0 7 ) reduced speed and probing with the ovipositor. the wasps loosing contact with the treated area exhibited a klinokinetic response and returned back to the treated area. two leptopilina species such as l. fembriata and l. heterotoma, which attacks drosophila spp. reduce their walking speed and increase their probing frequency (chiri and lengner, 1982). a significant increase in the number of both oviposition reactions (host stinging) and egg laying has been observed in aphidius eivi when the dummies of aphids are coated with conical secretion (larocca, 2007). however this enhancement has not been observed when the aphid dummies contained distilled water. budenburg & powell (1997) has reported that episyrphus balteatus (deg) landed more frequently on wheat ears contaminated with host aphid honeydew than on clean ears suggesting a response to honeydew volatiles. the number of eggs laid by e. baltetus has been increased with increasing honeydew concentration. honeydew of host aphids attracts the parasitoids from a distance (bouchard & cloutier, 1984) then arrests the parasitoid in the contaminated area. slowing down the increased antennae examination would increase the possibility of the parasitoid encountering the host. the klinokinatic response would increase intense searching throughout the contaminated area, further increasing the chance of finding host. references auclair, j.l. (1958). honeydew excretion in the pea aphid acyrthosiphon pisum (homoptera; aphiidae) j. insect physio.2, 330-337. bouchard, and clouher, (1984). honeydew as a source of host searching kairomone for the aphid parasitoid a. negriceps (hymenoptera; aphidiodae). canad. j. 2001. 62, 1513-1520. bouch, g.m., gaerawald, r.l. & miyasaki, s. (1969). development of a chemically defined diet for adults of the apple maggot based on amino acid analysis of honeydew. ann. ent. soc. am. 62, 19-21. budenburg, w.j. and powell, w. (1997). ent. exp. et appl. 64, 57-61. chiri, a.a. & leghan, e.f. (1982). host searching kairomones alter behavior of chelonus sp., a hymenopterous parasite of the pink bollworm, pechinophora gossypiella (saunders) environ. ent 11, 452-455. emden v. h.f., hagen, k.s. (1976). olfactory researches of green lacewing, chzysopa carnea to tryptophan and certain breakdown products. environ. ent. 5, 469-473. hagen, k.s., shawall, e.f. and tassan, r.l. (1976). tryptophan in artificial honeydew as a source of an attractant for adult chzysopa cavaea. environ. ent. 5, 458-468. larocca, m (2007). functional bases of host acceptance behaviour in the aphid parasitoids aphidius ervi. physiol. ent. 32, 305-312. martin, h, schmidt, andreaslaeur, tobias and carlsten thies (2003). relative importance of predators and parasitoids for cereal aphid control. environ. ent. 270, 1527, 1905-1909. sidhu, h.s., & patton, r.l. (1970). carbohydrates and nitrogenous compounds in the honeydew of the mustard aphid, lipaphis ezysimi. j. inset physiolo. 16. 1339-1348. shorey, h.h. (1977). interactions of insect with their chemical environment. in chemical control of insect behavior, edited by h.h. shorey, j.j. shorey & j.j. mckelvey. pp1-5. wiley, new york. wickramasinghe:searching behaviour of the aphid… 9 5 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 89-95 ( 2 0 0 7 ) strand, m.r. & vinson s.b. (1982). behavioral responses of the parasitoid casdiochiles nigriceps to a kairomone. ent. exp. appl. 31 308-315. vet, l.e.m. & van der hoeven, r. (1984). comparison of the behavioral response of two leptopilicina species (hymenoptera; eucoilidae) living in different microhabitats, to kairomone of their host (drosophilidae). neth. j. 2001. 34 220-227. vinson, s.b., haclan, d.p.& hart, g.w. (1978). responses of the parasitoid, microtestes flavus to the brown spot scale and its honeydew. environ. ent.7, 874-878. waage, j.k. (1978). arresment responses of the parasitoid, nemuritis canescens, to a contact chemical produced by its host, plodina intelpunctella. physiolo. ent. 3, 135-146. wickremasinghe, m.g.v & van emden, h.f. (1992). reactions of adult female parasitoids, particularly aphidius rhopalosiphi, to volatile chemical cues from the host plants of their aphid prey. physiolo. ent.17,297-304. ruhuna journal of science vol 11 (2): 157-165, december 2020 eissn: 2536-8400 © faculty of science doi: http://doi.org/10.4038/rjs.v11i2.94 university of ruhuna © faculty of science, university of ruhuna sri lanka 157 short paper ethnobotanical study of buru community forest, taraba state, nigeria aderopo akinsoji 1, doris omoigui 2, lanre ogunyebi *3 1department of botany, university of lagos, lagos, nigeria 2 department of botany and biotechnology, university of benin, benin, nigeria 3department of cell biology and genetics, university of lagos, nigeria *correspondence: logunyebi@unilag.edu.ng; orcid: https://orcid.org/0000-0003-2315-470x received: 18th june 2019, revised: 08th june 2020, accepted: 29th november 2020 abstract an ethnobotanical study of buru community forest was conducted using participatory rural appraisal techniques. a total of 91 species of plants belonging to 43 families with high endemicity were recorded. the dominant families were fabaceae, euphorbiaceae, and meliaceae. many species such as tetrapleura tetrapetra, phyllanthus mullerianus, sarcocephalus latifolius and aframomum melegueta had multiple uses. the three major uses of the species are for medicinal (39 species), edible (33 species) and construction purposes (30 species). keywords: buru community forest, ethnobotanical, participatory rural appraisal technique. 1 introduction buru is a small, remote rural village in kurmi local government area of taraba state, nigeria. the village is at the edge of a lowland rainforest called buru community forest (bcf). buru has six other associated hamlets which collectively form buru community (obot and inahoro 2004). by a participatory forest management arrangement with taraba state forestry department, buru community forest is being maintained and managed by the buru community. bcf is located between 60 5’ to 70 05’n and 100 81’ to 100 96’ e (figure 1) at an altitude of 314 asl in the foothills of mambilla highlands and covers an area of 10,800 hectares (akinsoji 2013). the area faces the rain-laden wind from the atlantic coast (chapman and chapman 2010) with a mean annual rainfall of 290 mm (bawdwen and tuley 1969) with bimodal peaks in july and september (akinsoji 2013). the dry season runs between november and march with a brief spell of dry http://doi.org/10.4038/rjs.v11i2.94 https://creativecommons.org/licenses/by-nc/4.0/ https://orcid.org/0000-0003-2315-470x https://orcid.org/0000-0003-2315-470x https://orcid.org/0000-0002-4761-4558 aderopo akinsoji et al. ethnobotanical study of buru community forest in nigeria ruhuna journal of science vol 11 (2): 157-165, december 2020 158 and cold harmattan wind in december (akinsoji 2013). bcf is a fragmented part of the guinea forest biodiversity hotspot which extends from sierra leone to congo, and it is known to be one of the 36 biodiversity hotspots of global significance for conservation priorities (myers et al. 2002, noss 2016). it is rich in biodiversity and it has been designated as an important birding area (ezealor 2002). the soil is of volcanic origin comprising various ratios of clay mixture hence the ground is characterized by hills and depressions which make the terrain hilly and rugged (obot and inahoro 2004). the forest is a lowland rain forest with characteristic vertical stratification and dense canopy coverage draped with lianas and climbers. the forest harbours some endangered species and many iucn red data list plants. the common forest species include khaya grandifoliola, milicia excelsa, terminalia superba, ceiba pentandra, cola gigantea, bosqueia angolensis, khaya ivorensis, entandrophragma utile, tetrapleura tetrapetra and zanthoxylum zanthoxyloides. buru community is an agrarian community, and the main occupation is farming. over the years, parcels of the forest had experienced structural changes. for instance, the traditional slash and burn followed by shifting cultivation had turned some of the farmed forest parcels into derived savanna. some common species of the derived savanna include hymenocardia acida, anogeissus lieocarpus, uapaca togoensis, terminalia avicennioides, t. laxiflora, combretum spp., crossopteryx febrifuga, sarcocephalus latifolius and detarium microcarpum. buru is a small settlement of about 600 people dominated by the tigun ethnic group who are the land-owners (akinsoji 2013). the others are ndoros, kakas, and mambillas who are migrant farmers. they have lived and survived in this environment for generations with minimal contact with the rest of the world depending only on forest resources to meet their livelihood needs. the major components of the resources are plants. this study was carried out to document how the people use plants for their survival. 2 material and methods the survey was carried out using participatory rural appraisal (pra) techniques (martin 1995, mccracken et al. 1998, akinsoji 2003). focus group discussions were conducted in buru and the satellite hamlets. at each site, three groups comprising men, women and youths were engaged in discussions. two teachers from buru primary school acted as interpreters. in addition, one-on-one interviews were also held with certain individuals who were passive during the group discussions. quantitative data collected for buru and the hamlets were similar, so they were pooled. the validity of information gathered was verified by triangulation (walter 1998, akinsoji 2003). the data compiled was further discussed with the participants. plant specimens were identified and recorded. those that could not be immediately identified were recorded with their indigenous (hausa) names. their botanical aderopo akinsoji et al. ethnobotanical study of buru community forest in nigeria ruhuna journal of science vol 11 (2): 157-165, december 2020 159 nomenclature was deciphered using gbile (1980). those whose hausa names were not known were taken to forestry herbarium in ibadan where they were identified and confirmed. all the plant specimens in healthy condition were then deposited in gashaka herbarium located in gashaka gumti national park. fig.1: map of nigeria showing buru community forest 3 results and discussion a total of 91 species of plants belonging to 43 families were sampled in buru community forest (bcf), and interestingly they were reported as endemic to this community. the family fabaceae dominated with 14 species. the families euphorbiaceae and meliaceae had six species each while the families apocynaceae and rubiaceae had five species each. the above five families constituted more than aderopo akinsoji et al. ethnobotanical study of buru community forest in nigeria ruhuna journal of science vol 11 (2): 157-165, december 2020 160 one-third of all species encountered. some species such as tetrapluera tetrapetra and aframomum melegueta are used for more than one purpose. t. tetrapetra plant is used to treat skin rash and chest pain, while the edible fruit is added to pepper soup for flavor. it is also burnt to give an aroma which is believed to drive away evil spirits. such multiple purpose plants have been recorded in gashaka gumti national park, nigeria (akinsoji 2003), malawi (maghembe and seyani 1991) and in phillipines (rondolo 2000). the major uses for plants are medicinal, edible and construction purposes. 3.1 medicinal plants thirty-nine species belonging to 23 families of medicinal plants were recorded in the buru community forest (bcf). the two families fabaceae and euphorbiaceae dominated with five species each while the family apocynaceae had four species. the remaining families had one or two species each (table 1). the part of plants that is mostly used for medicinal purpose is the leaf. this is probably due to its function as the site of production of biological molecules that have bioactive properties. similar findings were also reported by anbarashan and padmavathy (2010) in india and phannel et al. (2010) in western kenya. nineteen (48.7%) species had their leaves used for treating ailments. five species had their fruits used for medicinal purposes, while four species had their roots, and four species had their seeds respectively. three species, tetrapleura tetrapetra, anselia gigantea and crinum jagus are used for magical/mythical purposes. t. tetrapetra fruit is used to drive away evil spirits, a. gigantea is used as a love charm while c. jagus is used as a protective charm. this corroborates walker (1999) claim that some aspects of traditional medicine are linked with magic. table 1: medicinal plants of buru community forest species family part used uses acanthus montanus (nees) t. anders acanthaceae leaf typhoid alchornea cordifolia (schum &thonn.) mull.arg. euphorbiaceae leaf malaria ansellia gigantea rchb f. orchidaceae pseudobulb love charm cenchrus aralioides roxb. poaceae dried plant impotence chromolaena odorata (linn.) king & robinson asteraceae leaf wound dressing cissus aralioides (welw.ex bak.) planch ampelidaceae root dizziness celtis ferruginea dc. connaraceae fruit oral hygiene commelina sp linn. commelinaceae leaf wound treatment costus dubius (afzel.) k. schum costaceae leaf cough, mouth sores crinum jagum (themps.) dandy amaryllidaceae leaf protective charm culcasia scandens p. beauv araceae leaf purgative erythrina vogelli hook. f. fabaceae leaf malaria euphorbia hirta linn. euphorbiaceae leaf skin disease aderopo akinsoji et al. ethnobotanical study of buru community forest in nigeria ruhuna journal of science vol 11 (2): 157-165, december 2020 161 table 1. continued species family part used uses garcinia kola heckel guttiferae seed cough ixora brachypoda dc. rubiaceae sap wound dressing jatropha podagrica hook. euphorbiaceae leaf wound treatment kigelia africana (lam.) benth bignoniaceae bark body pains lophira alata banks ex.gaertn.f ochnaceae leaf heart pain macaranga barteri muell. arg. euphorbiaceae napoleona imperialis p.beauv. lecythidaceae root fever olax subscorpioidea oliv. olacaceae stem sap antiseptic phyllanthus mullerianus (kuntze) ex. euphorbiaceae leaf malaria piper guineense schum & thonn. piperaceae seed skin rash piper umbellatum linn. piperaceae seed skin rash rauvolfia vomitora afzel. apocynaceae leaf malaria sarcocephalus latifolius (sm.) bruce rubiaceae root gonorrhea, stomach pain senna alata (linn.) roxb. fabaceae leaf skin diseases senna siamea mill. fabaceae flower activate lactation solanum torvum sw solanaceae leaf rib pain solenostemon monostachyus p. beauv. lamiaceae stereospermum kunthianum cham. bignoniaceae bark dysentery/ stomachache strombosia pustulata oliv. olacaceae strophanthus hispidus dc. apocynaceae leaf navel pain tabernaemontana pachysiphon stapf. apocynaceae fruit std tamarindus indica linn. fabaceae fruit aphrodisiac tetrapleura tetrapetra (schum&thonn) taub. fabaceae seed chest pain, skin rash vernonia amygdalina dcl. asteraceae leaf fever, general tonic voacanga africana stapf. apocynaceae fruit std xylopia aethiopica (dunal) a.rich. annonaceae fruit skin rash zanthoxylum zanthoxyloides (lam) zepern.& timler rutaceae root sickle cell 3.2 edible plants thirty-three species of edible plants belonging to 23 families were recorded in bfc (table 2). the family fabaceae dominated with 6 species while the other species had one or two species each. twenty-five of the species were trees. the most common edible part was the fruit (19 species) which are known to be source of nutrients and vitamins hence they are good food supplements. the other edible plants were seed (10 species) and leaf (7 species). the seeds are used as spices to add flavour and aroma to foods/ soups while the leaves are eaten as vegetables as well as spices. some of these have been reported to be sold in lagos markets (akinsoji 2017). hence, lagos and some other towns could be potential outlets for these plants to aderopo akinsoji et al. ethnobotanical study of buru community forest in nigeria ruhuna journal of science vol 11 (2): 157-165, december 2020 162 serve as a source of supplementary family income for buru inhabitants. elaies guineensis is remarkable because at least three of its parts are edible. its palm oil is used for frying and making of stews, the kernel oil is edible and used in making soap while the sap is taken as palm wine. these edible plants are important in the rural economy by increasing household income for the family as they are harvested and taken to markets for sale. akinsoji (2003) and campbell (1987) made similar observations in gashaka (nigeria) and zimbabwe, respectively. table 2: edible plants of buru community forest species family leaf fruit seed tuber aframomum melegueta k. schum zingiberaceae + annona senegalensis pers. annonaceae + beilschmiedia manni (meisn.) benth.& thonn.f. lauraceae + blighia sapida konig sapindaceae + brachystegia eurycoma harms fabaceae + chrysophyllum albidum g.don sapotaceae + crassocephalum rubens (juss. ex. jacq.) s. moore asteraceae + diallium guineense wild. fabaceae + dacryodes edulis (g. don) hj. lam burseraceae + elaeis guinensis jacq. arecaceae + + + irvingia gabonensis aubryleconte ex o rorke) baill irvingiaceae + + kigelia africana (lam.) benth bignoniaceae + landolphia owariensis p. beauv. apocynaceae + maesobotrya dusenii (pax) hutch euphorbiaceae + mangifera indica linn. anacardiaceae + moringa oleifera lam. moringaceae + + + musanga cecropioides muell. arg moraceae + napoloena imperialis p.beauv. lecythidaceae + ocimum gratissimum linn. labiatae + parkia biglobosa jacq benth fabaceae + + persea americana mill. lauraceae + piper guineense schum. &thonn. piperaceae + piper umbellatum linn. piperaceae + psidium guajava linn. myrtaceae + pterocarpus erinaceus poir fabaceae + ricinodendron heudelotii (baill.) pierre euphorbiaceae + tamarindus indica linn. fabaceae + tetrapleura tetrapetra (schum. & thonn.) taub. fabaceae + trichilia preuriana a. juss meliaceae + vernonia amygdalina del. asteraceae + vitex simplicifolia oliv. verbenaceae + xylopia aethiopica (dunal) a.rich. annonaceae + + zanthoxylum zanthoxyloides (lam.) zepernick &timber rutaceae + aderopo akinsoji et al. ethnobotanical study of buru community forest in nigeria ruhuna journal of science vol 11 (2): 157-165, december 2020 163 3.3 construction plants table 3 shows that thirty species of plants are used for construction purposes. these plants belong to thirteen families dominated by families meliaceae, fabaceae and poaceae. seventeen of these are timber species while the remaining thirteen are either soft wooded trees (2) or herbaceous species belonging to the family poaceae (e.g. phragmites and palisota). timber species are used mainly for house construction (roofing, doors and furniture) while the soft wooded species are used for making agricultural implements such as hoe or knife, and cutlass handles. grasses are used mainly to thatch roofs of huts but some are also used to make baskets, mats, door curtains and chairs which can be sold in markets to support family income. akinsoji (2003) reported similar results for gashaka gumti national park. entada barks are stripped and used as ropes to tie wooden beams together in thatching roofs with grasses. the timber species are conserved through a joint management agreement with taraba state forestry department. the agreement vests the ownership and management of bcc on buru community and does not permit commercial logging. only community members can log for personal building purposes and permission to log must be obtained from forest management committee of the community. table 3: construction plants of buru community forest, nigeria species family parts used use a) timber afzelia africana sm. fabaceae wood (hard) roofing of buildings, furniture, wooden houses etc, albizia zygia (dc.) j.f. macbr. fabaceae + + aubrevillea kerstingii (harms) pellegr. fabaceae + + bombax buonopozense p.beauv. brachystegia eurycoma harms bombacaceae + + canarium schwenfurthii linn. burseraceae + + celtis zenkeri engl. diospyros dendo welw.ex.hein funtumia africana meliaceae ebenaceae apocynaceae + + hallea ciliate aubr.& pellegr. rubiaceae + + khaya ivorensis c. dc. meliaceae + + khaya grandifoliola c. dc. meliaceae + + khaya senegalensis (desr.) a.juss meliaceae + + lovoa trichidioides harms. meliaceae + + milicia excelsa (nelw.) c.c. bery moraceae + + piptadeniastrum africanum (hook. f.) brenan fabaceae + + sterculia rhinopetala k. schum sterculiaceae + + trilepisium madagascariense dc moraceae + + aderopo akinsoji et al. ethnobotanical study of buru community forest in nigeria ruhuna journal of science vol 11 (2): 157-165, december 2020 164 table 3 continued. family parts used use b) non-timber species borassus aethiopum mert. arecaceae stem and frond roof beams and thatching elaeis guineensis jacq. arecaceae frond making baskets and furniture entada purseantha dc. fabaceae bark rope for tying roof beams kigelia africana (lam.) benth bignoniaceae soft wood making wooden handles for knives and agricultural implements laccosperma secundiflorum (p.beaux) o kuntze arecaceae culm cane furniture oxytenanthera abyssinica (a. rich,) munro poaceae stem making huts, fences and furniture pandanus crassicaulis huynh pandanaceae leaves thatching/roofing palisota hirsuta (thunb.) k. schum poaceae whole plant thatching of huts panicum maximum jacq. poaceae culm thatching phragmites karka (retz,) trin. ex.steud). poaceae culm beds, chairs and thatching rothmania hispida (k.schum) rubiaceae soft wood wooden handles for knives and agricultural implements (e.g. hoe) vitex simplicifolia stapf verbenaceae soft wood wooden handles for knives and agricultural implements (e.g. hoe) 4 conclusions the study documents how the people in the buru community in kurmi local government area use plants for their survival. it revealed high abundance of plant resources in this area and how traditional farming system of slash and burn impacted on the plant resources, though the management option of buru community forest as a result of the agreement between the taraba state forestry department and the community help in adequate protection and sustainability of the forest. this study recommends that inhabitants should be educated on the importance of conservation as continuous exploitation of the plant resouces without adequate conservation strategies can lead to loss of some of these important plant resources. aderopo akinsoji et al. ethnobotanical study of buru community forest in nigeria ruhuna journal of science vol 11 (2): 157-165, december 2020 165 acknowledgements we thank mr. emma hamzat, the divisional forest officer for granting us the permission to carry out the study and introducing us to buru community. we are grateful to chief joel emma and the buru community for their hospitality and cooperation. mr. emmah lucas and mrs. mercy m. luka served as interpreters gratis. anonymous reviewers of rjs are acknowledged. references akinsoji a. 2003. vegetation studies of gashaka gumti national park, nigeria 1: ethnobotany. the nigerian field 68:124-144 akinsoji a. 2013. community-based forest management in buru, taraba state, nigeria. journal of environment and earth science 3(12): 146-151 akinsoji a. 2017. market survey of spices and vegetables obtained from four markets in lagos metropolis. fuw trends in science and technology 2(2): 788-791 anbarashan m, padmavathy a. 2010. ethno-medicinal plants of five sacred groves in andalore district, tamilnadu, india. ethnobotanical leaflets 14: 774-780 bawden mg, tuley p. 1969. the land resources of s. sardauna and s. adamawa provinces, northern nigeria. land resources division, tolworth, surrey. chapman jd, chapman hm. 2001. the forests of taraba and adamawa states, nigeria. an ecological account and plant species checklist. university of canterbury, christchurch, new zealand. ezealor au. 2002. critical sites for biodiversity conservation in nigeria. nigerian conservation foundation, lagos. 110pp. gbile zo. 1980. vernacular names of nigerian plants (hausa). federal department of forestry, lagos. 63pp. martin gj. 1995. ethnobotany. chapman and hall, london. 268pp mccracken ja, preety jn, conway gr. 1998. an introduction to rapid rural appraisal for agricultural development. internationl institute for environment and development, london. myers n, mittermeier cg, de fonseca gab, kent j. 2002. biodiversity hotspots for priorities. nature 403: 1045 noss r. 2016. announcing the world’s 36th biodiversity hotspot: the north american coastal plain. retrieved on 08/16/2020, https://www.cepf.net/stories/announcing-worlds-36th-biodiversity-hotspotnorth-american-coastal-plain obot ea, inahoro i. 2004. natural resources management plan for buru community forest. nigerian conservation foundation, lagos. 58pp. phannel as, nyunja ro, onyago jc. 2010. plant species in the folk medicine of kit mikayi region, western kenya. ethnobotanical leaflets 10: 836-840 walter s. 1998. the utilization of non-timber forest products in the rainforest of madagascar. a case study. plants research and development 48: 121-144 https://www.cepf.net/stories/announcing-worlds-36th-biodiversity-hotspot-north-american-coastal-plain https://www.cepf.net/stories/announcing-worlds-36th-biodiversity-hotspot-north-american-coastal-plain microsoft word rjs-2007-s.ramanayaka-1.2.doc © 2007 faculty of science university of ruhuna ruhuna journal of science vol. 2, september 2007, pp. 48-69 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x 48 abstract a simulation model is presented for cost effective paddy product transportation in sri lanka. paddy production in sri lanka is assumed to be sufficient to meet the entire requirement of the country, and distributing among the administrative districts is taken to be proportional to their respective populations. this simulation problem is solved by chance constraint stochastic transportation method. suppliers and consumers are determined by their production and their population by assuming paddy production to be independent and normally distributed. keywords: maha season, yala season, simulation, transportation problem, stochastic transportation problem. 1. introduction rice is the staple food in sri lanka. it is produced from paddy, which is harvested in the two cultivation seasons ‘maha’ and ‘yala’, which are agricultural periods based on monsoon rains. time period of these seasons are from september/october to march/april for the ‘maha’ season, and from april/may to august/september for the ‘yala’ season. total rice requirement for human consumption can be produced in sri lanka. however, in some years rice is being imported to meet the demands. for instance in 2003, total rice requirement for human consumption is 1,923 thousand metric tons of which 1,888 thousand metric tons had been supplied from domestic source and, 35 thousand metric tons from imports (www.statistics.gov.lk). paddy harvests vary highly among the districts of sri lanka and consumptions also vary according to the human population of the districts. because of this unbalanced paddy production and consumption, it is required to transport paddy from higher production areas to low production areas. however, due to various reasons, price of rice is relatively high and is steadily going up. one of the factors that cause increase in price of rice is ad-hoc simulated model for cost effective paddy product distribution in sri lanka s ramanayake1 and g.t.f. de silva2 1advanecd technological institute, dehiwala, 2department of mathematics, university of moratuwa1 sudarshana@email.com 2gtfdes1@yahoo.com ramanayake and de silva: simulated model for... 49 ruhuna journal of science 2, pp.48-69 (2007) transportation. in an earlier work on this matter, de silva et al. (1979) have solved a simple transportation problem of paddy supplier districts to consumer districts. a new transportation strategy is attempted in this work where instead of ad-hoc or classical transportation method, the problem is solved by using stochastic transportation problem (stp), which is a special class of stochastic programming problem (spp). there are various classes of spp such as single objective stochastic programming problem (sospp), multi objective stochastic programming problem (mospp) and stochastic linear programming problem (slpp), and they are classified according to the treatment of objectives and that of the constraints (mohan et. al. 1997). a very common approach of chance constraint programming simplifying to deterministic equation is used for this study (hamdy 1999) due its simplicity and the comparatively less amount of calculations involved. the sequel of this paper is organized as follows: in the section 2 we present the methods and materials used in the investigation with deterministic equivalent of stochastic transportation problem that derives model equations. in the third section, results are presented followed by discussion and conclusion. 2. methods and materials the transportation problem deals with commodity shipped from a source to a destination. the objective is to determine the amounts shipped, from each source or supplier district to each destination or consumer district that minimizes the total transportation cost while satisfying both supply limits and the demand requirements (hamdy 1999a, harvey 1999b). this model assumes that the transportation cost on a given route is directly proportional to the number of units shipped on that route and is taken to be proportional to the distance between administrative capitals of the districts. stochastic paddy transportation problem which aims at finding out cost benefit transportation strategy can be described as below: if cij and xij are transportation cost and number of units to be transported from ith supplier of ‘m’ number of suppliers to the jth consumer of ‘n’ number of consumers respectively, then the problem is to minimize the total transportation cost z given by ∑∑= i j ijij xcz , (1) subject to mixap i n j iji ,...2,1 ˆ 1 =≥       ≥ ∑ = α , and (2) 50 ramanayake and de silva: simulated model for... ruhuna journal of science 2, pp.48-69 (2007) njxbp j m i ijj ,...2,1 1ˆ 1 =−≥       ≤ ∑ = β , (3) where p in the above stands for the probability, iâ , supply amount, is independent, normally distributed random variable with mean e( iâ ) and variance var( iâ ) and similarly jb̂ , demand amount, is also independent normally distributed with mean e( jb̂ ) and variance var( jb̂ ). the minimum probabilities that satisfy i th supplier constraints and jth demand constraints are iα and 1jβ respectively. paddy production is uncertain in each season and year due to various reasons as well as climatic conditions. the supply and demand of the paddy production vary from season to season and year to year due to various reasons. for instance, the uncertainty in the weather conditions is one of the obvious ones. deterministic equivalent of stochastic transportation problem in the stochastic chance constrained transportation problem described above, supply and demand constraints depend on probabilities of at least iα and jβ−1 respectively. these chance constraints can be converted to equivalent deterministic form as follows (hamdy 1999a). consider the ith supply constraint, mixap i n j iji ,...2,1 ,ˆ 1 =≥       ≥ ∑ = α . based on the normality assumption, one can easily see that this ith constraint reduces to ( ) ( ) ( ) i k a aex i i n j ij αφ≤             − φ ∑ = ˆvar ˆ 1 ramanayake and de silva: simulated model for... 51 ruhuna journal of science 2, pp.48-69 (2007) where ( ) iik αα −=φ 1 and ().φ is the cumulative distribution function of a standard normal distribution. this gives, m . . . 3, 2, 1,i ;)ˆ()ˆvar( 1 =+≤∑ = ii n j ij aeakx iα (4) similarly the second type of chance constraints (demand chance constraints) can be written as equivalent to the deterministic type as n . . . 3, 2, 1,j ;)ˆ()ˆvar( 1 =+≥∑ = jj m i ij bebkx jβ . (5) in many research, solution approach of spp is to find its deterministic equivalent. another way of approaches is fuzzifying approach to cope with vagueness appearing in the cost functions and constraints. in this approach decision maker has specified a fuzzy aspiration level of probability to the stochastic constraints and objective functions and then get the deterministic equivalent of spp (mohan 1997). here, it is adapted the earlier one due to it’s simplicity and less interaction of decision maker. after getting deterministic equivalent of spp, it can be solved first by taking a feasible solution and then by performing iterations. least-cost first rule is adapted to get feasible solution (hamdy 1999, harvey 1999). in this study, a simulation has been used to find out the best transportation strategy for paddy production in sri lanka. since district-wise paddy production is unbalanced, it is needed to transport the paddy production from surplus areas to consumer areas. the paddy production is taken with the reasonable assumption that the total amount would be sufficient to meet the entire requirement of the country. the total paddy production of a cultivation year is designed to be distributed among administrative districts proportional to the population density. on this basis that each district requirement an amount proportional to its population, certain districts which have surpluses after meeting its own needs were identified as supplier with the amounts of supply and those who are in deficit were identified as consumer along with their requirements. therefore paddy consumption per person (γ ) is defined by considering available data of each district. ( ) ( ){ } ( )∑ ∑ × + = i ii i ii w populationdistrict production yalaproduction maha γ where 52 ramanayake and de silva: simulated model for... ruhuna journal of science 2, pp.48-69 (2007)    = seasons theof onein available is data if 1 seasonsboth in available is data if 2 iw these weights are assigned according to the availability of paddy harvest data in cultivation seasons. therefore, once there is an absence of data in any cultivation season of a particular district, it is omitted from the transportation problem. similarly district-wise paddy consumption per person ( iγ ) is taken as a simple ratio of paddy harvest to the population in each season as follows, idistrict of population idistrict in season a of productionpaddy =iγ . a particular district will be either a supplier or a consumer. supplier or consumer districts are determined based on the figures iγ and γ . that is, if iγ is greater than γ then the district ‘i’ is considered as a supplier district of that cultivation season and if iγ is less than γ then the district ‘i’ is considered as a consumer district of that cultivation season. thus the supplier or consumer amount of each district for a season is given by ( ) ( )idistrict of population×−γγ i . the calculation procedure is shown in figure 1. figure 1: flow chart to find out supplier and consumer districts if a particular district has surplus paddy production in ‘maha’ but needs some more paddy to fulfill its requirement in ‘yala’, the excess production of ‘maha’ will be start find γ calculate γi for particular season of ith district if γi>=γ in this season, i th district is labeled as supplier calculate the supplier amount ( )*di i l i calculate the consumer amount (γ-γi)*district population end yes no in this season, ith district is labeled as consumer ramanayake and de silva: simulated model for... 53 ruhuna journal of science 2, pp.48-69 (2007) no yes no yes star if one season supplier & if supplier amount>cons end supplier amt =supplier amt-consumer amt consumer amt = consumer amt-supplier end allocated to be used in the ‘yala’ season, instead of transporting to another district. if ‘maha’ production is large enough to fulfill ‘yala’ requirement, the remaining amount after allocating for ‘yala’ seasonal demand, can be transported. otherwise, if ‘maha’ production is not large enough, all the surpluses are allocated for ‘yala’ requirement and the rest will be taken from another district. therefore, if a particular district has a surplus production, it may be transported or not, which is decided by ‘yala’ production (figure 2). however, if both seasons have an excess production they are transported without any adjustment. again if both seasons need more paddy to fulfill their requirements they are taken from another supplier district. moreover, as total supplier amount and consumer amount are not the same, fictitious consumers or fictitious suppliers are introduced to make balance transportation problem. figure 2. rearranging the supply and demand amount according to the 'maha' and 'yala' requirement supplier and consumer districts are determined based on their average values of concerned period (1989-2003) to construct simulation model. supplier and consumer amounts of each district for each season is calculated based on right hand side values of deterministic equivalence of chanced constraint equations (equation 4 and 5). three simulation cases are presented by assuming that the minimum 54 ramanayake and de silva: simulated model for... ruhuna journal of science 2, pp.48-69 (2007) probabilities to hold the supplier constraints (αi’s) are 0.15, 0.05, and 0.01 and the minimum probabilities to hold demand constraints (βj’s) are 0.85, 0.95 and 0.99. these simulated supplier and consumer amount calculating procedure are shown in the following flow chart of figure 3 for this study, data from year 1989 to 2004 from the annual report of the central bank sri lanka, are considered and the simulated model is presented for the year 2004. initial transportation tables and optimum tables were prepared using ms excel and matlab packages. figure 3: flow chart for the forecast supply or consumer amount (where, kα is found from standard normal table) as described above, initial transportation tables of the year 2004 of both seasons are presented in table a.1 (a) – (b). in these tables, supply and demand amounts of supplier and consumer districts are presented by thousands of metric tons of a right most column and a bottom row respectively. the cost of transport from a supplier to a consumer is taken to be proportional to the distance between administrative capitals of district. the relative transportation cost is presented in the initial transportation tables by row and column deduction. similarly, simulated initial transportation tables of the year 2004 are shown in table a.2 (a) – (c) and tables a.3 (a) – (c). the decision maker (dm) can decide the probability level required to hold demand and supply constraints ( eq. (2) & eq. (3)) and then can decide upon various transportation strategies for the problem. among those simulated solutions dm can adopt best solution based on decision rules. no yes start read if district is consumer demand amount = ( ) ( )bebk ˆˆ end district is supplier supply amount = ( ) ( )aeak ˆˆvar + ramanayake and de silva: simulated model for... 55 ruhuna journal of science 2, pp.48-69 (2007) 3. results in both ‘maha’ and ‘yala’ seasons, colombo district is the main consumer district. as the main consumer district the average paddy demands of the colombo district is 151,180 mt.(table 1) and 161,880 mt.(table 2) in ‘maha’ and ‘yala’ seasons respectively. polonnaruwa district is recorded as the highest supplier in ‘maha’ season whilst ampara district is the highest supplier in ‘yala’ season with average supply of 261,040 mt.(table 1) and 146,120 mt.(table 2), respectively. badulla, kurunagala, matale, monaragala, mannar, anuradhapura, polonnaruwa, trincomalee, batticaloa, ampara and hambantota districts are the average suppliers whilst colombo, kalutara, galle, matara, kegalle, ratnapura, kandy, nuwaraeliya, puttalum, and jaffna are the consumer districts in ‘maha’ seasons. however, in ‘yala’ season, only few districts function as suppliers namely polonnaruwa, trincomalee, batticaloa, ampara, and hambantota and others are consumers (jaffna district ‘yala’ data are not available). according to the paddy transportation strategies for the considered years obtained by solving the classical transportation problems, kurunagala district is the main supplier for the colombo for all the years in ‘maha’ season except for the years 1996, 1999 and 2004. moreover kurunagala and polonnaruwa are the only suppliers of paddy to the colombo district in maha season. the highest supplier in the ‘maha’ season is the polonnaruwa and frequently it supplies to colombo, kalutara, kegalle, kandy & jaffna districts and sometimes it supplies to puttalum, vauniya & mannar. moreover every year polonnaruwa supplies a large amount of paddy to the fictitious consumer. it means that polonnaruwa can store a large amount of paddy of its harvest for the ‘yala’ season, which is comparatively low harvest. second highest supplier in this season is ampara district. ampara district frequently supplies its excess productions to kandy and nuwaraeliya and sometime supplies to ratnapura districts. every year and season to season, there is a common pattern between suppliers and consumers. transportation strategies for the year 2004 are shown in table a.4 and in table a.5 for ‘maha’ and ‘yala’ respectively. in next season, “yala” supplies are polonnaruwa, trincomalee, batticaloa ampara, and hambantota. almost all suppliers are from dry zone of northeast, east and southeast areas. however the major consumers are colombo, kalutara, galle, matara, kegalle, ratnapura, kandy and nuwaraeliya, the same as in the ‘maha’ season. again colombo is the highest consumer and there is not a fixed supplier but most of the time it meets the requirement from the fictitious supplier. however when polonnaruwa or kurunagala play as the supplier in this season, colombo district receives its needy by them. sometimes it receives paddy from anuradhapura and hambantota too. 56 ramanayake and de silva: simulated model for... ruhuna journal of science 2, pp.48-69 (2007) during the ‘yala’ season, as it is not much produced like the ‘maha’ season most of the consumers especially major consumers get their need from fictitious suppliers. moreover, the amount they receive from fictitious supplier match with the amounts store at fictitious consumers in ‘maha’ season of the relevant year. therefore fictitious supplier may be from the same district or a fixed supplier of the ‘maha’ season. there is an interesting relationship among the supplier districts and the consumer districts in ‘maha’ season according to the solutions of transportation problems. that is all major consumers have regular suppliers. suppliers of colombo are polonnaruwa and kurunagala. suppliers of kalutara district are hambantota and polonaruwa. moreover hambantota district regularly supplies paddy to galle, matara and sometimes to ratnapura districts too. monaragala district also regularly supplies to the ratnapura district. suppliers to kandy district are ampara, matale and polonnaruwa. kegalle and nuwaraeliya get their need from polonnaruwa, troncomalee respectively. these observations are same for the year 2004. in ‘yala’ season, there is no clearly shown regular supplier to the particular consumer as shown in ‘maha’ season. however ampara, polonnaruwa and hambantota which are the highest producers in this season have regular consumer districts. ampara district supplies paddy to kandy, nuwaraeliya, badulla and monaragala districts while hambantota supplies to kalutara and galle. sometime polonnaruwa and kurunagala play as the suppliers of colombo district in this season. but in year 2004 kurunagala has not been the supplier. further certain amounts of its (colombo) needs come from fictitious supplier. as in ‘maha’ season these common observations are shown in year 2004 (table a.5) always ‘maha’ season has a fictitious consumer to stock its excess product and ‘yala’ season needed fictitious supplier to fulfill the demand its consumers. in ‘maha’ season ampara and polonnaruwa regularly supply to fictitious consumer. further sometimes baticaloa, matale, kurunagala anuradapura and trincomalee also supply to the fictitious consumer in this season. therefore these districts can stock paddy harvest to use next ‘yala’ season and as it is in ‘yala’ season, most of the consumers such as colombo, kalutara, galle, matara, kegalle, ratnapura receive their needed amounts from fictitious suppliers. however, in year 2004 ‘maha’ season actual transportation strategy show that batticaloa ampara and trincomalee are supplied their excess product to the fictitious consumer of 72.38, 65.45, 49.08 thousand meric tons of paddy respectively and in ‘yala’ season, galle, colombo, puttalam, ratnapura, matara, kalutara are received their needy from fictitious suppliers. ramanayake and de silva: simulated model for... 57 ruhuna journal of science 2, pp.48-69 (2007) according to the paddy supply and demand statistics, it shows that averagely they are of high variation from year to year as well as season to season. generally, most of the supplier districts, their supply amounts too are of high variation than the consumer districts. kurunagala, anurahapura polonnaruwa, trincomalee, batticaloa and ampara are the main supplier districts and comparatively variation is higher than the other districts. however in year 2004, ‘maha’ season badulla, matale, monaragala, vauniya, manar, anuradhapura, trincomalee, baticaloa, ampara, and hambantota districts are the suppliers. but in ‘yala’ season of this year only polonnaruwa, baticaloa, ampara and hambantota are the suppliers. the estimated values of simulated transportation model of the year 2004 are shown in the table 1 and table 2 of both seasons. due to the variations in the paddy supply or consumption amounts, the simulated paddy supply or consumption amounts of those districts are deviated from the actual values. for instance in ‘maha’ season, kurunagala district plays as a consumer actually whereas in the simulated result it shows that it is a supplier. moreover, the simulated values for jaffna, anuradhapura, ampara and hambantota are deviated from actual values. all three simulation cases of ‘yala’ season, the anuradhapura district is neither supplier nor consumer whereas actually it supplies 143.38 thousands metric tons of paddy. however, there is a significant deviation of ‘maha’ season. in ‘yala’ season in the actual case kurunagalla district plays neither as supplier nor a consumer. however, the simulation shows that it is a supplier. it supplies paddy to colombo district as same as in the deterministic transportation problem of each and every year. the simulation results of the kurunagala district are noticeably deviated from the actual values of both seasons. moreover, suppliers and the consumers of simulation are the same as in deterministic cases. -300 -200 -100 0 100 200 300 400 c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a b ad ul la p ut ta la m k ur un ag al a m at al e m on ar ag al a ja ffn a v au ni ya m an na r a nu ra dh ap ur a p ol on na ru w a t rin co m al ee b at tic al oa a m pa ra h am ba nt ot a 00 0' m t actual alpha = 0.15 beta = 0.15 alpha = 0.05 beta = 0.05 alpha =0.01 beta = 0.01 supply demand figure 4. simulated and actual supply or demand amounts of paddy in year 2004 ‘maha’ 58 ramanayake and de silva: simulated model for... ruhuna journal of science 2, pp.48-69 (2007) according to the simulated results, the excess production of matale, monaragala, vauniya, mannar and anuradhapura districts in the “maha’ season are stored for the consumption in the ‘yala’ season. this is same for all three cases considered in this study. however, excess production of all those districts except anuradhapura are not enough to completely fulfill the requirement of ‘yala” season. the excess production of anuradhapura district is sufficient to cover the demand of ‘yala’ season and hence the rest is transported. -300 -200 -100 0 100 200 300 c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a b ad ul la p ut ta la m k ur un ag al a m at al e m on ar ag al a ja ffn a v au ni ya m an na r a nu ra dh ap ur a p ol on na ru w a t rin co m al ee b at tic al oa a m pa ra h am ba nt ot a 00 0' m t actual alpha = 0.15 beta = 0.15 alpha = 0.05 beta = 0.05 alpha =0.01 beta = 0.01 supply demand figure 5. simulated and actual supply or demand amounts of paddy in year 2004 'yala' table 1: simulated supply and demand (minus) amounts for “maha” season of the year 2004. the α and β probabilities are 0.15, 0.05 and 0.01. district std dev variance alpha = 0.15 alpha = 0.05 alpha =0.01 average paddy productio n ‘89-‘03 (‘000 mt) beta = 0.15 beta = 0.05 beta = 0.01 colombo -151.18 25.29 639.48 -177.229 -192.782 -210.005 kalutara -37.8 12.27 150.62 -50.4381 -57.9842 -66.34 galle -36.67 9.5 90.33 -46.455 -52.2975 -58.767 matara -18.17 8.35 69.75 -26.7705 -31.9058 -37.5921 kegalle -27.43 11.5 132.3 -39.275 -46.3475 -54.179 ratnapura -38.27 11.32 128.23 -49.9296 -56.8914 -64.6003 kandy -58.28 15.74 247.64 -74.4922 -84.1723 -94.8912 nuwaraeliya -33.22 9.93 98.66 -43.4479 -49.5549 -56.3172 badulla 6.06 7.24 52.47 13.5172 17.9698 22.9002 4 ramanayake and de silva: simulated model for... 59 ruhuna journal of science 2, pp.48-69 (2007) district std dev variance alpha = 0.15 alpha = 0.05 alpha =0.01 puttalam -22.29 10.12 102.41 -32.7136 -38.9374 -45.8291 kurunagala 96.74 35.97 1293.89 133.789 1 155.910 7 180.406 2 matale 14.49 5.69 32.38 20.3507 23.8500 5 27.7249 4 monaragala 11.84 6.32 39.93 18.3496 22.2364 26.5403 2 jaffna -49.85 12.36 152.83 -62.5808 -70.1822 -78.5994 vauniya 3.87 7.13 50.83 11.2139 15.5988 5 20.4543 8 mannar 6.21 9.06 82.05 15.5418 21.1137 27.2835 6 anuradhapura 79.43 57.54 3310.74 138.696 2 174.083 3 213.268 polonnaruwa 261.04 31.38 984.55 293.361 4 312.660 1 334.029 9 trincomalee 25.93 15.96 254.65 42.3688 52.1842 63.0529 6 batticaloa 36.27 24.94 622.21 61.9582 77.2963 94.2804 4 ampara 141.39 46.62 2173.39 189.408 6 218.079 9 249.828 1 hambantota 75.39 19.39 375.92 95.3617 107.286 6 120.491 1 table 2 simulated supply and demand (minus) amounts for “yala” season of the year 2004. the α and β probabilities are 0.15, 0.05 and 0.01. district std dev variance alpha = 0.15 alpha = 0.05 alpha =0.01 average paddy productio n ‘89-‘03 (‘000 mt) beta = 0.15 beta = 0.05 beta = 0.01 colombo -161.88 24.73 611.66 -187.352 -202.561 -219.402 kalutara -49.1 12.02 144.57 -61.4806 -68.8729 -77.0585 galle -55.65 12.83 164.54 -68.8649 -76.7554 -85.4926 matara -27.93 7.32 53.64 -35.4696 -39.9714 -44.9563 kegalle -34.86 10.51 110.36 -45.6853 -52.149 -59.3063 ratnapura -48.14 10.26 105.37 -58.7078 -65.0177 -72.0048 kandy -73.56 16.17 261.37 -90.2151 -100.16 -111.171 nuwaraeliya -40.74 9.93 98.69 -50.9679 -57.0749 -63.8372 badulla -26.18 6.56 42.99 -32.9368 -36.9712 -41.4386 60 ramanayake and de silva: simulated model for... ruhuna journal of science 2, pp.48-69 (2007) district std dev variance alpha = 0.15 alpha = 0.05 alpha =0.01 puttalam -36.55 2.73 7.46 -39.3619 -41.0409 -42.9 kurunagala -10.89 36.19 1309.81 -48.1657 -70.4226 -95.0679 matale -18.46 3.41 11.65 -21.9723 -24.0695 -26.3917 monaragala -16.17 3.41 11.63 -19.6823 -21.7795 -24.1017 jaffna -19.6823 -21.7795 -24.1017 vauniya -9.29 4.18 17.49 -13.5954 -16.1661 -19.0127 mannar -9.03 1.69 2.85 -10.7707 -11.8101 -12.9609 anuradhapura -11.01 45.82 2099.62 -58.2046 -86.3839 -117.587 polonnaruwa 137.59 54.17 2934.70 193.385 1 226.699 7 263.589 4 trincomalee 4.87 11.16 124.51 16.3648 23.2282 30.8281 6 batticaloa 1.35 10.18 103.63 11.8354 18.0961 25.0286 8 ampara 146.12 40.10 1607.83 187.423 212.084 5 239.392 6 hambantota 50.83 17.00 289.15 68.34 78.795 90.372 in ‘maha season colombo, kalutara, galle, matara, kegalle, ratnapura, kandy, nuwaraelliya puttalam and jaffna districts always be consumer districts in both simulated and actual deterministic transportation strategies. kurunagala is consumer of actual case but it plays as the largest supplier to colombo in the simulation. badulla district does not play as supplier or consumer in both actual and simulated cases of this season. but it has been consumer of the years 1998, 1999 & 2003 and supplier of the year 1989 according to the solutions of transportation problems. when the minimum probability of supply constraint is decreased, the number of suppliers is increased. in this operational study, if the supply minimum probability is reduce from 0.15 to 0.05, the monaragala district becomes a supplier and if that probability further decreases to 0.01, matale and vauniya also become suppliers. in the ‘maha’ season of the year 2004, the actual transportation strategy shows that badulla, matale, and anuradhapura do not play as supplier nor consumer. however, in simulations sometime matale plays as supplier while anuradhapura plays as supplier in all cases. the major consumer, the colombo district obtains all of its requirements from polonnaruwa and trincomalee of year 2004 in ‘maha’ season but in simulation kurunagala, polonnaruwa and batticaloa are the major suppliers for the colombo district. moreover, in the actual case, the highest supplier for colombo is polonnaruwa whereas the simulation shows that the highest supplier is kurunagala. in the ‘yala’ season of year 2004, both actual and simulated cases show that supplier are same and they are polonnaruwa, trincomalee, batticaloa, ampara & ramanayake and de silva: simulated model for... 61 ruhuna journal of science 2, pp.48-69 (2007) hambantota. moreover, when the minimum probability of demand constraint is increased, the number of consumers is decreased. initially, the ‘yala’ season of year 2004 monaragala go away from the consumer list and then matale and vauniya are also removed. however, monaragala and vauniya are not the consumers in actual transportation strategy but matale is a consumer. in this season, suppliers are same in both actual and all simulation cases. as there are few suppliers in ‘yala’ season, most of the consumers take their needs from fictitious supplier which is stored in ‘maha’ season by its suppliers. both actual and simulated cases show that almost all same consumers receive paddy from fictitious supplier but amounts they received are little deviated. in actual case, the kalutara and puttalum receive paddy from fictitious supplier with 0.27 thousand metric tons and 47.39 thousand metric tons respectively but simulation transportation strategy show that these two district requirements are not fulfilled by fictitious supplier; instead they get their requirement from hambantota & polonnaruwa respectively. moreover, puttalum district receive all of its requirements from fictitious supplier in actual case but simulation transportation strategy shows that all its requirements are fulfilled from the polonnaruwa district. 4. discussion and conclusion paddy performance of ’mahavalli h’ region and ‘udawalawe’ region which are recorded on central bank annual reports are added to their respective district to build up transportation problem based on distances of the aaa road map. therefore, paddy production of those regions is added to ‘polonnaruwa’ and ‘hambantota’ districts respectively. moreover, ‘gampaha’, ‘kilinochchi’ and ‘mulativu’ districts are not considered here as the aaa road map not included those districts to get districts-wise minimum distance systematically. transportation costs have been taken to be proportional to the minimum road distance among the districts. however, the other related costs such as loading, unloading, storing and inventory costs are not considered here. therefore, this model assumes the storage of excess production which is assumed to have no significant cost involvement to be used in the next season instead transporting. moreover, in this operational study, it is assumed that the total annual production is sufficient for consumption for the entire population. it has not considered export and import situations. this scenario can be included to improve the model by adding as production of the shipped district if it is imported and deducting as consume amount of the shipped district if it is exported. from this study it is easy to conclude that there is a supply/demand pattern and results are useful in decision making towards cost reduction. further, if supply and demand constraints hold minimum reasonable probabilities, forecasted amounts of supply or demand and transportation strategy are sufficiently close to actual cost beneficial transportation strategy. 62 ramanayake and de silva: simulated model for... ruhuna journal of science 2, pp.48-69 (2007) moreover, it can be concluded that the consumers those who receive their needs from fictitious suppliers in ‘yala’ season could get their regular supplies in ‘maha’ season as shown in transportation strategy. for instance, polonnaruwa supplies large amount of its surplus paddy in ‘maha’ season to fictitious consumer in every year. then the colombo district gets its additional requirement from fictitious supplier in ‘yala’ season in every year. so this requirement could be met from polonnaruwa district which is one of the suppliers of colombo. moreover, additional requirement of colombo and kalutara districts that are taken from fictitious supplier in ‘yala’ is nearly equal or less than the amount which is fictitious consumer getting from polonnaruwa district in ‘maha’ season. in maha season, consumer districts are same in both actual and simulated cases but two more supplier districts are added to simulated cases than the actual case. moreover, the number of supplier districts increased one by one in the second and the third simulated cases. transportation costs are increased 6%, 51% and 34% of simulated cases than the actual case. in contrast, ‘yala’ season supplier districts are same in both actual and simulated cases, but two consumer districts are dropped from simulated cases. however, another three more districts are added as consumers to the first simulation case and then the number of consumer districts decrease by one and two from second and third simulation cases respectively. transportation costs are relatively deviate from actual case however considerable amount of paddy are supplied by fictitious supplier in both actual and simulated case. it is assumed that both supply and demands are normally distributed. but both paddy production and population densities have upward trends. therefore, further research could be done trying with some more probability distributions, and extension of this model is possible for specific practical situation. acknowledgements. we would like to thank dr. m indralingum who gave valuable comments and encouragement of this study. appendix table a.1 initial transportation table of the year 2004 (a) 'maha' season (b) 'yala' season (‘000 mt). (transportation cost, which is proportional to minimum road distance, is indicated after row and column deduction). ramanayake and de silva: simulated model for... 63 ruhuna journal of science 2, pp.48-69 (2007) 20 04 e 1 m ah a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n ’e liy a p ut ta la m ja ffn a s up pl y kurunagala 0 0 130 222 0 40 10 50 55 237 85.62 mannar 66 66 196 288 66 106 52 92 0 0 4.77 anuradhapura 71 70 201 293 79 119 64 105 0 87 80.49 polonnaruwa 15 14 145 237 14 54 0 40 35 109 293.36 trincomalee 15 14 145 237 14 54 0 40 35 109 42.37 baticaloa 15 14 145 237 14 54 0 40 35 109 61.96 ampara 104 45 100 103 39 13 0 0 126 250 189.41 hambantota 100 11 0 0 150 0 168 53 253 435 95.36 demand -177.23 -50.44 -46.46 -26.77 -39.28 -49.93 -74.49 -43.45 -32.71 -62.58 (a) 20 04 m ah a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a p ut ta la m k ur un ag al a ja ffn a s up pl y monaragala 34 0 112 65 95 0 78 0 242 120 335 14.93 vauniya 33 57 196 288 34 130 52 129 0 34 0 7.17 mannar 33 57 196 288 34 130 52 129 0 34 19 9.75 polonnaruwa 0 23 163 255 0 96 18 95 53 0 146 309.1 trincomalee 0 23 163 255 0 96 18 95 53 0 146 61.73 baticaloa 0 23 163 255 0 96 18 95 53 0 146 72.38 ampara 71 36 100 103 7 37 0 37 126 42 269 203.29 hambantota 67 2 0 0 118 24 168 90 253 166 454 68.99 demand -161.25 -50.94 -40.75 -17.41 -37.95 -46.04 -62.4 -44.33 -31.83 -35.82 -31.71 (b) table a.2 simulated initial transportation table of the year 2004 'maha' season (a) α=0.15, 1-β=0.85 (b) α=0.05, 1-β=0.95 (c) α=0.01, 1-β=0.99(‘000 mt) (a) (b) 20 04 y al a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a b ad ul la p ut ta la m k ur un ag al a m at al e a nu ra dh ap ur a s up pl y polonnaruwa 0 42 184 276 1 93 26 66 132 0 0 9 0 151.81 trincomalee 0 42 184 276 1 93 26 66 132 0 0 9 0 0.54 baticaloa 0 42 184 276 1 93 26 66 132 0 0 9 0 10.02 ampara 63 47 113 116 0 26 0 0 0 65 34 0 97 208.45 hambantota 46 0 0 0 98 0 155 40 57 179 145 159 287 61.16 demand -172.38 -61.43 -57.52 -18.68 -39.13 -49.53 -73.69 -49.83 -2.25 -47.39 -40.6 -2.87 -3.6 20 04 e 2 m ah a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n ’e liy a p ut ta la m ja ffn a s up pl y kurunagala 0 0 130 222 0 64 10 87 55 237 85.49 monaragala 67 9 112 65 127 0 78 0 242 316 0.46 mannar 66 66 196 288 66 130 52 129 0 0 9.3 anuradhapura 71 70 201 293 79 143 64 142 0 87 87.7 polonnaruwa 15 14 145 237 14 78 0 77 35 109 312.66 trincomalee 15 14 145 237 14 78 0 77 35 109 52.18 baticaloa 15 14 145 237 14 78 0 77 35 109 77.3 ampara 104 45 100 103 39 37 0 37 126 250 218.08 hambantota 100 11 0 0 150 24 168 90 253 435 107.29 demand -192.78 -57.98 -52.3 -31.91 -46.35 -56.89 -84.17 -49.55 -38.94 -70.18 64 ramanayake and de silva: simulated model for... ruhuna journal of science 2, pp.48-69 (2007) (c) 20 04 e 3 m ah a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a p ut ta la m ja ffn a s up pl y kurunagala 0 0 130 222 0 64 10 87 55 256 85.34 matale 55 54 184 277 38 119 0 77 128 254 1.33 monaragala 67 9 112 65 127 0 78 0 242 335 2.44 vauniya 66 66 196 288 66 130 52 129 0 0 1.44 mannar 66 66 196 288 66 130 52 129 0 19 14.32 anuradhapura 71 70 201 293 79 143 64 142 0 106 95.68 polonnaruwa 15 14 145 237 14 78 0 77 35 128 334.03 trincomalee 15 14 145 237 14 78 0 77 35 128 63.05 baticaloa 15 14 145 237 14 78 0 77 35 128 94.28 ampara 104 45 100 103 39 37 0 37 126 269 249.83 hambantota 100 11 0 0 150 24 168 90 253 454 120.49 demand -210 -66.34 -58.77 -37.59 -54.18 -64.6 -94.89 -56.32 -45.83 -78.6 table a.3 simulated initial transportation table of the year 2004 'yala' season (a) α=0.15, 1-β=0.85 (b) α=0.05, 1-β=0.95 (c) α=0.01, 1-β=0.99(‘000 mt) (a) 20 04 e 1y al a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a b ad ul la p ut ta la m m at al e m on ar ag al a ja ffn a v au ni ya s up pl y polonnaruwa 0 3 145 237 0 54 0 0 36 0 0 247 0 0 193.39 trincomalee 0 3 145 237 0 54 0 0 36 0 0 247 0 0 16.36 baticaloa 0 3 145 237 0 54 0 0 36 0 0 247 0 0 11.84 ampara 182 127 193 196 118 106 93 53 23 184 110 0 234 214 187.42 hambantota 85 0 0 0 136 0 168 13 0 218 189 36 326 325 68.34 demand -187.35 -61.48 -68.86 -35.47 -45.69 -58.71 -90.22 -50.97 -19.42 -39.36 -1.62 -1.33 -19.68 -2.38 (b) 20 04 e 3 y al a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a b ad ul la p ut ta la m m at al e ja ffn a v au ni ya s up pl y polonnaruwa 0 3 145 237 0 54 0 27 93 0 0 0 0 226.7 trincomalee 0 3 145 237 0 54 0 27 93 0 0 0 0 23.23 baticaloa 0 3 145 237 0 54 0 27 93 0 0 0 0 18.1 ampara 102 47 113 116 38 26 13 0 0 104 30 154 134 212.08 hambantota 85 0 0 0 136 0 168 40 57 218 189 326 325 78.8 demand -202.56 -68.87 -76.76 -39.97 -52.15 -65.02 -100.16 -57.07 -19 -41.04 -0.22 -21.78 -0.57 ramanayake and de silva: simulated model for... 65 ruhuna journal of science 2, pp.48-69 (2007) (c) 20 04 e 3 y al a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a b ad ul la p ut ta la m ja ffn a s up pl y polonnaruwa 0 3 145 237 0 54 0 27 93 0 0 263.59 trincomalee 0 3 145 237 0 54 0 27 93 0 0 30.83 baticaloa 0 3 145 237 0 54 0 27 93 0 0 25.03 ampara 102 47 113 116 38 26 13 0 0 104 154 239.39 hambantota 85 0 0 0 136 0 168 40 57 218 326 90.37 demand -219.4 -77.06 -85.49 -44.96 -59.31 -72 -111.17 -63.84 -18.54 -42.9 -24.1 table a.4 the amount of paddy to be transported (optimum table) for the year 2004 'maha' seasons (‘000 mt) 20 04 m ah a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a p ut ta la m k ur un ag al a ja ffn a f ic tit ou s c on su m er monaragala 0 0 0 0 0 14.93 0 0 0 0 0 0 vauniya 0 0 0 0 0 0 0 0 0 0 7.17 0 mannar 0 0 0 0 0 0 0 0 0 0 9.75 0 polonnaruwa 148.6 40.11 0 0 37.95 0 0 0 31.83 35.82 14.79 0 trincomalee 12.65 0 0 0 0 0 0 0 0 0 0 49.08 baticaloa 0 0 0 0 0 0 0 0 0 0 0 72.38 ampara 0 0 0 0 0 31.11 62.4 44.33 0 0 0 65.45 hambantota 0 10.83 40.75 17.41 0 0 0 0 0 0 0 0 table a.5 the amount of paddy to be transported (optimum table) for the year 2004 'yala' seasons (‘000 mt) 20 04 y al a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a b ad ul la p ut ta la m k ur un ag al a m at al e a nu ra dh ap ur a polonnaruwa 107.61 0 0 0 0 0 0 0 0 0 40.6 0 3.6 trincomalee 0.54 0 0 0 0 0 0 0 0 0 0 0 0 baticaloa 10.02 0 0 0 0 0 0 0 0 0 0 0 0 ampara 0 0 0 0 39.13 40.68 73.69 49.83 2.25 0 0 2.87 0 hambantota 0 61.16 0 0 0 0 0 0 0 0 0 0 0 fs 54.21 0.27 57.52 18.68 0 8.85 0 0 0 47.39 0 0 0 66 ramanayake and de silva: simulated model for... ruhuna journal of science 2, pp.48-69 (2007) table a.6 simulated transportation strategy of ‘maha’ season of the year 2004 (paddy ‘000 mt to be transported) (a) α=0.15, 1-β=0.85 (b) α=0.05, 1-β=0.95 (c) α=0.01, 1-β=0.99 (a) 20 04 e 1m ah a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a p ut ta la m ja ffn a f c kurunagala 85.62 0 0 0 0 0 0 0 0 0 0 mannar 0 0 0 0 0 0 0 0 0 4.77 0 anuradhapura 0 0 0 0 0 0 0 0 32.71 47.78 0 polonnaruwa 29.65 47.35 0 0 0 0 0 0 0 10.03 206.33 trincomalee 0 3.09 0 0 39.28 0 0 0 0 0 0 batiticaloa 61.96 0 0 0 0 0 0 0 0 0 0 ampara 0 0 0 0 0 27.8 74.49 43.45 0 0 43.67 hambantota 0 0 46.46 26.77 0 22.13 0 0 0 0 0 (b) 20 04 e 1m ah a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a p ut ta la m ja ffn a f c kurunagala 85.62 0 0 0 0 0 0 0 0 0 0 mannar 0 0 0 0 0 0 0 0 0 4.77 0 anuradhapura 0 0 0 0 0 0 0 0 32.71 47.78 0 polonnaruwa 29.65 47.35 0 0 0 0 0 0 0 10.03 206.33 trincomalee 0 3.09 0 0 39.28 0 0 0 0 0 0 batiticaloa 61.96 0 0 0 0 0 0 0 0 0 0 ampara 0 0 0 0 0 27.8 74.49 43.45 0 0 43.67 hambantota 0 0 46.46 26.77 0 22.13 0 0 0 0 0 (c) 20 04 e 3 m ah a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a p ut ta la m ja ffn a f c kurunagala 85.34 0 0 0 0 0 0 0 0 0 0 matale 0 0 0 0 0 0 0 0 0 0 1.33 monaragala 0 0 0 0 0 2.44 0 0 0 0 0 vauniya 0 0 0 0 0 0 0 0 0 1.44 0 mannar 0 0 0 0 0 0 0 0 0 14.32 0 anuradhapura 0 0 0 0 0 0 0 0 45.83 49.85 0 polonnaruwa 84.56 3.29 0 0 0 0 0 0 0 12.99 233.19 trincomalee 0 63.05 0 0 0 0 0 0 0 0 0 batiticaloa 40.1 0 0 0 54.18 0 0 0 0 0 0 ampara 0 0 0 0 0 38.03 94.89 56.32 0 0 60.59 hambantota 0 0 58.77 37.59 0 24.13 0 0 0 0 0 ramanayake and de silva: simulated model for... 67 ruhuna journal of science 2, pp.48-69 (2007) table a.7 simulated transportation strategy of ‘yala’ season of the year 2004 (paddy ‘000 mt to be transported) (a) α=0.15, 1-β=0.85 (b) α=0.05, 1-β=0.95 (c) α=0.01, 1-β=0.99 (a) 20 04 e 1y al a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a b ad ul la p ut ta la m m at al e m on ar ag al a ja ffn a v au ni ya polonnaruwa 130.35 0 0 0 0 0 0 0 0 39.36 1.62 0 19.68 2.38 trincomalee 16.36 0 0 0 0 0 0 0 0 0 0 0 0 0 batiticaloa 11.84 0 0 0 0 0 0 0 0 0 0 0 0 0 ampara 0 0 0 0 0 25.48 90.22 50.97 19.42 0 0 1.33 0 0 hambantota 0 61.48 6.86 0 0 0 0 0 0 0 0 0 0 0 fs 28.8 0 62 35.47 45.69 33.23 0 0 0 0 0 0 0 0 (b) 20 04 e 3 y al a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a b ad ul la p ut ta la m m at al e ja ffn a v au ni ya polonnaruwa 152.27 0 0 0 10.82 0 0 0 0 41.04 0.22 21.78 0.57 trincomalee 0 0 0 0 23.23 0 0 0 0 0 0 0 0 batiticaloa 0 0 0 0 18.1 0 0 0 0 0 0 0 0 ampara 0 0 0 0 0 35.85 100.16 57.07 19 0 0 0 0 hambantota 0 68.87 9.93 0 0 0 0 0 0 0 0 0 0 fs 50.29 0 66.83 39.97 0 29.17 0 0 0 0 0 0 0 (c) table a.8 data sheet for the year 2004, s: supplier district, d: consumer district 20 04 e 3 y al a c ol om bo k al ut ar a g al le m at ar a k eg al le r at na pu ra k an dy n uw ar ae liy a b ad ul la p ut ta la m ja ffn a polonnaruwa 168.11 0 0 0 28.48 0 0 0 0 42.9 24.1 trincomalee 0 0 0 0 30.83 0 0 0 0 0 0 batiticaloa 25.03 0 0 0 0 0 0 0 0 0 0 ampara 0 0 0 0 0 45.84 111.17 63.84 18.54 0 0 hambantota 0 77.06 13.31 0 0 0 0 0 0 0 0 fs 26.26 0 72.18 44.96 0 26.16 0 0 0 0 0 68 ramanayake and de silva: simulated model for... ruhuna journal of science 2, pp.48-69 (2007) y ea r p ro du ct io n( m t) p op ul atio n (‘0 00 ) g am m a s up pl ie r/ c on su m er a m ou nt ( ‘0 00 ) a dj us te d am ou nt ( ‘0 00 ) m ah a y al a to ta l m ah a y al a m ah a y al a m ah a y al a m ah a y al a 74 .5 82 9 1 74 .5 82 9 1 c ol om bo 13 42 1 23 01 15 72 2 23 42 5. 73 0. 98 d d -1 61 .2 5 17 2. 3 8 16 1. 2 5 17 2. 3 8 k al ut ar a 29 98 3 19 49 4 49 47 7 10 85 27 .6 3 17 .9 7 d d -5 0. 94 61 .4 3 50 .9 4 61 .4 3 g al le 35 31 8 18 55 6 53 87 4 10 20 34 .6 3 18 .1 9 d d -4 0. 75 57 .5 2 40 .7 5 57 .5 2 m at ar a 41 35 9 40 09 7 81 45 6 78 8 52 .4 9 50 .8 8 d d -1 7. 41 18 .6 8 17 .4 1 18 .6 8 k eg al le 21 05 1 19 86 8 40 91 9 79 1 26 .6 1 25 .1 2 d d -3 7. 95 39 .1 3 37 .9 5 39 .1 3 r at na pu ra 32 19 4 28 71 2 60 90 6 10 49 30 .6 9 27 .3 7 d d -4 6. 04 49 .5 3 46 .0 4 49 .5 3 k an dy 36 42 3 25 13 3 61 55 6 13 25 27 .4 9 18 .9 7 d d -6 2. 4 73 .6 9 -6 2. 4 73 .6 9 n uw ar ae liy a 95 20 40 18 13 53 8 72 2 13 .1 9 5. 57 d d -4 4. 33 49 .8 3 44 .3 3 49 .8 3 b ad ul la 77 24 6 41 77 8 11 90 2 4 81 3 95 .0 1 51 .3 9 s d 16 .6 1 18 .8 6 0 -2 .2 5 p ut ta la m 22 46 8 68 98 29 36 6 72 8 30 .8 6 9. 48 d d -3 1. 83 47 .3 9 31 .8 3 47 .3 9 k ur un ag al a 75 01 1 70 22 9 14 52 4 0 14 86 50 .4 8 47 .2 6 d d -3 5. 82 -4 0. 6 35 .8 2 -4 0. 6 m at al e 50 86 3 14 73 2 65 59 5 45 9 11 0. 81 32 .1 s d 16 .6 3 -1 9. 5 0 -2 .8 7 m on ar ag al a 58 36 0 17 73 3 76 09 3 41 0 14 2. 34 43 .2 5 s d 27 .7 8 12 .8 5 14 .9 3 0 ja ffn a 12 73 4 12 73 4 59 6 21 .3 7 d -3 1. 71 31 .7 1 v au ni ya 27 92 6 43 2 28 35 8 14 2 19 6. 66 3. 04 s d 17 .3 3 10 .1 6 7. 17 0 m an na r 23 68 3 53 9 24 22 2 97 24 4. 15 5. 56 s d 16 .4 5 -6 .7 9. 75 0 a nu ra dha pu ra 10 17 8 2 99 17 11 16 9 9 77 3 13 1. 67 12 .8 3 s d 44 .1 3 47 .7 3 0 -3 .6 p ol on na ru w a 33 68 4 4 17 95 5 7 51 64 0 1 37 2 90 5. 49 48 2. 68 s s 30 9. 1 15 1. 8 1 30 9. 1 15 1. 8 1 tr in co m al ee 90 30 0 29 10 7 11 94 0 7 38 3 23 5. 77 76 s s 61 .7 3 0. 54 61 .7 3 0. 54 b at iti ca lo a 11 29 5 1 50 59 2 16 35 4 3 54 4 20 7. 63 93 s s 72 .3 8 10 .0 2 72 .3 8 10 .0 2 a m pa ra 24 90 0 4 25 41 7 0 50 31 7 4 61 3 40 6. 21 41 4. 63 s s 20 3. 29 20 8. 4 5 20 3. 2 9 20 8. 4 5 h a m ba nt ot a 10 91 1 3 10 12 8 7 21 04 0 0 53 8 20 2. 81 18 8. 27 s s 68 .9 9 61 .1 6 68 .9 9 61 .1 6 ramanayake and de silva: simulated model for... 69 ruhuna journal of science 2, pp.48-69 (2007) references de silva g t f, ahlip r a. (dec. 1979). rice transportation strategy for sri lanka. slaas annual session. mohan,c. and nguyen, h. t. (1997). a fuzzifying approach to stochastic programming. opsearch 34, 2: 73-96 hamdy a taha. (1999). operations research an introduction. pp.165-208, 807-809, prentice hall of india. harvey m wagner (1999). principles of operations research with applications to managerial decisions. prentice hall of india, new delhi. central bank of sri lanka (1989-2004). annual report. central bank of sri lanka colombo supply of rice, wheat and sugar for human consumption by source of supply and rate of self suffficiency. http://www.statistics.gov.lk/agriculture/riceproductionandimports.pdf. ruhuna journal of science vol 11 (2): 98-117, december 2020 eissn: 2536-8400 © faculty of science doi: http://doi.org/10.4038/rjs.v11i2.90 university of ruhuna © faculty of science, university of ruhuna 98 sri lanka free radical scavenging potential and antibacterial activity of cola nitida and garcinia kola extracts against bacterial strains isolated from patients with urinary tract infections i.v. anyiam 1* and p.p.e. mounmbegna2 1department of microbiology, faculty of science, federal university otuoke, bayelsa state, nigeria 2department of biochemistry, faculty of science, madonna university nigeria, elele, rivers state, nigeria correspondence: ifetgod@yahoo.co.uk; https://orcid.org/0000-0002-3705-3988 received: 23rd may 2019, revised: 16th july 2020, accepted: 29th november 2020 abstract. cola nitida and garcinia kola are found and widely consumed in west africa. the seeds of these plants have various traditional uses and are reported to exhibit several bioactivities. their phytochemical, antioxidant and antibacterial properties of methanol, ethanol and aqueous extracts were investigated in the present study. phytochemical screening and quantification of total phenolic contents analysis were carried out for phytochemical investigation. preliminary phytochemical screening revealed the presence of flavonoids, alkaloids, tannins, saponins, protein and glycosides in the seed extracts. quantitative phytochemical constituents revealed 0.818 ± 0.021and 0.700 ± 0.017mg of phenolic compounds and total flavonoid content of 25.63 ± 1.60 and 25.10 ± 1.85mg in g. kola and c. nitida respectively. the extracts showed potent antioxidant activities compared to standard antioxidants by significantly inhibiting 2, 2-diphenyl-1picrylhydrazyl (dpph), hydroxyl radical (∙oh), and superoxide anion radicals (o2∙) dose dependently. the methanol extracts of g. kola and c. nitida showed significant inhibitory action (p<0.05) against the bacterial isolates. the minimum inhibitory concentration obtained for methanol extract of the plants and both the mixture was 12mm at 31.25mg/ml for klebsiella pneumoniae while the ethanol and aqueous extract of the plants and both the mixture was 13mm and 12.33mm at 31.25mg/ml and 125mg/ml respectively for e. coli. a direct correlation was observed between total phenolic content of extracts and radical scavenging potential, thus linking the observed bioactivities of these extracts to the presence of the phytochemical. the mixture of these seed extracts showed greater effect against the bacterial isolates, therefore providing a platform for advance studies in the development of drugs against infectious diseases. key words: agar well dilution, antibacterial activity, cola nitida, garcinia kola, free radical scavenging potential 1 introduction the alarming rate of antimicrobial drug resistance by pathogenic microorganisms against synthetic antibiotics (maiyo et al. 2010) is a serious global problem. indeed, the emergence of bacterial resistance to antibacterial drug today has become a common https://creativecommons.org/licenses/by-nc/4.0/ https://orcid.org/0000-0002-3705-3988 i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 99 phenomenon, and consequently, antibiotic resistance has imposed both a biological and economic cost (chabot et al. 2002, chen et al. 2002, chessin et al. 2005). the rate of antimicrobial resistance has prompted the search for new plants with antimicrobial properties and potentials to serve as sources of raw material for the synthesis of new drugs (akoachere et al. 2002). traditional healers use different plant medicines to provide health care to most of the people in a curative rather than a preventive approach in the developing countries for common ailments (gabriel et al. 2007). the availability and economy of these plants as direct therapeutic agents makes it more attractive when compared to modern medicine (agbo and ngogang 2005, agbo et al. 2005). natural plants contain phytochemical properties similar to synthetic antibiotics and have been used in folk medicine to treat infections (ezeigbo 2016). in addition, people with different cultural backgrounds from ancient times to the present day have used herbal medicines (elmahmood et al. 2008) to cure infections. hence, plants continue to be the most preferred exclusive source of drugs for the majority of the world’s population (fabiola et al. 2003, jonathan and fasidi 2003, ajayi et al. 2008). according to who (2000), “medicinal plants when administered to man or animals exert a sort of pharmacological action on them”. for this reason, medicinal plants are used as sources to produce useful drugs utilized by people worldwide for treatment of infectious diseases. infectious diseases are the major causes of death accounting for approximately one half of all deaths in tropical countries (iwu et al. 2009). in recent times, medicinal plants continue to play a major role in primary healthcare as therapeutic remedies in many developing countries (jonathan and fasidi 2003, 2005, jonathan et al. 2007) as some plants have been found to be rich in secondary metabolites, such as tannins, terpenoids, alkaloids, flavonoids, phenols, steroids and volatile oil. these compounds are said to be responsible for their therapeutic activities (rabe and vanstoden 2000, cowan 2009). furthermore, plants can serve as a reservoir of effective chemotherapeutic agent which provides valuable natural drug for effective and efficient management of human and plant diseases (kanomal et al. 2014). in nigeria, studies have been carried out on a variety of these medicinal plants yet a good number of them with putative medicinal and antimicrobial potentials are yet to be studied (amalu et al. 2014). among these plants are garcinia kola and cola nitida whose medicinal uses may have not been fully explored in the treatment of bacterial infections, especially, urinary tract infections. these medicinal properties could occur in different forms varying from biological, synthetic chemotherapeutic, antibiotics, and phytotherapeutic agents (arekemase et al. 2012). the action of these agents could either be ‘bactericidal’ or ‘bacteriostatic’ (arekemase et al. 2012). the importance and quest for these medicinal plants origin that could be of potential benefit as antibacterial agents stimulated the interest in garcinia kola (‘bitter kola’) and cola nitida (kola nut) seeds which are widely consumed as stimulant (atawodi et al. 2005). garcinia kola, also generally known as ‘bitter kola’ is a flowering plant species that belongs to the family of tropical plants known as guttiferae or clusiaceae (adesuyi et al. 2012). in nigerian languages, it is commonly called “namijin goro” in hausa, “orogbo” in yoruba, and “agbilu” in igbo (dalziel 2008). bitter kola is also i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 100 known as african wonder nut because almost every part of it has been found to be of medicinal importance (adegboye et al. 2008). cola nitida (kola nut) (“goro” in hausa; “obi gbanja” in yoruba; “oji” in igbo, keay et al. 2014) is a member of the family steculicca. it is a tree plant found in sierra leone, north ashanti, tropical western africa, west indies, brazil and java (grieve, 2001). cola nitida was originally distributed along the west coast of africa from sierra leone to the republic of benin with the highest frequency and variability occurring in the forest areas of côte d'ivoire and ghana (opeke 2012). in addition, kola nut is a native stimulant which commonly chewed in many west african cultures (opeke 2012). in nigeria, it is often used in traditional occasions, to welcome guest and receive visitors at home. more so, the need for new antimicrobial agents is closely related with the problem of emergence of resistant strains to most antibiotics. hence, this study was conducted to determine the phytochemical constituents, free radical scavenging potential and antibacterial activity of cola nitida and garcinia kola. 2 materials and methods 2.1 isolation and identification sixty urine samples were collected one each from female patients attending federal medical centre, yenagoa, from the period of may to july in 2017. females were used because they were the available patients at the time of the study and are considered to be more predisposed to urinary tract infections. the specimens were cultured on macconkey agar, blood agar and cled (cystine lactose electrolyte deficient) agar plates using the streak method. different agar was used to selectively identify and differentiate the possible bacteria including the fastidious organism that might be present in the culture specimen. plates were inoculated and incubated at 37oc for 24 hours. the isolates were identified using gram staining technique and biochemical tests which include catalase, urease, coagulase, oxidase, and indole. 2.2 collection and authentication of plant material dried seeds of garcinia kola (bitter kola) and cola nitida (kola nut) were procured from a local herb dealer at swali market in yenagoa local government area, bayelsa state, nigeria. they were authenticated with voucher specimen number mp-182 and mp-183 in the pharmacognosy department, madonna university, nigeria, elele, rivers state, nigeria. 2.3 processing and extraction the seeds of garcinia kola and cola nitida were peeled, thoroughly washed and rinsed in distilled water, and both were sliced into tiny pieces with the use of a clean stainless i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 101 steel knife, then air-dried at room temperature for 4 weeks, and pulverized using laboratory mortar and pestle. different organic solvents (methanol, ethanol and aqueous) were used for the extraction of these plants as described by alade and irobi (1993). fifty grams (50 g) of each seed powder was dispensed into a cotton wool stock thimble chamber of the soxhlet apparatus and 500 ml of methanol was dispensed into flat bottom flask. the extraction solvents were heated in the bottom flask, vaporized into the thimble, condensed in the condenser and dripped back. when the liquid content reached the siphon arm, the liquid contents was emptied into the bottom flask again and the process was continued until the absorbent was clear. the extracts obtained were completely evaporated (green 2004) and stored in the refrigerator at 4oc until use. the percentage (%) yields of the dry residue were calculated (pudhom et al. 2007). the same procedure was repeated successfully for ethanol and aqueous extracts. extracts were then dissolved in the appropriate solvent for the phytochemical and antibacterial assay. 2.4 phytochemical screening phytochemical screening was done using qualitative and quantitative phytochemical analysis. qualitative analysis involved tests for flavonoids, tannins, carbohydrates, glycosides, saponins, resins, terpenoids and alkaloids. these were carried out using standard methods (harborne 1984, sofowora 1993, trease and evans 2001). quantitative analysis determined the total phenols, tannin, total flavonoids and total anthocyanin contents. the total phenolics were determined using folin-ciocalteau reagent (fcr) as described by velioglu et al. (1998) with slight modifications. tannin content in each sample was determined using insoluble polyvinyl-polypirrolidone (pvpp), which binds tannins as described by makkar et al. (1993). the flavonoids content was determined according to the method described by kumaran and karunakaran (2006) with slight modifications. this method was based on the formation of a flavonoid-aluminum complex, which absorbs maximally at 415 nm. the total anthocyanin contents of the plant extracts were measured using a spectrophotometric ph differential protocol described by giusti and wrolstad (2001) and wolfe et al. (2003) with slight modifications. 2.5 in vitro antioxidant assays quantitative dpph radical-scavenging assay the hydrogen atoms or electrons donation ability of the corresponding extract was measured from the bleaching of purple coloured methanol solution of dpph. the scavenging activity on dpph free radicals by the extract was assessed according to the method reported by gyamfi et al. (1999) with slight modifications. i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 102 hydroxyl radical (.oh)-scavenging assay the 2-deoxyribose assay was used to determine the scavenging effect of the extract on the hydroxyl (.oh) radical, as reported by halliwell et al. (1987) with minor modifications. superoxide radical (o2.-) scavenging assay this assay was based on the capacity of the extract to inhibit the photochemical reduction of nitro blue tetrazolium (nbt) (beauchamp and fridovich 1971) and the method used by martinez et al. (2001) to determine superoxide dismutase with slight modifications. 2.6 reconstitution of plant extract for preliminary screening of antibacterial activity of the plant extracts against bacteria isolated, all the dried extracts were dissolved in dimethyl sulfoxide (dmso) to a final stock concentration of 2.5% w/v. as dmso has been shown not to have any inhibitory effect on the growth of microorganisms (zgoda and porter 2001, kuete et al. 2008), it was used as the negative control for all the experiments. a two-fold serial dilution was also undertaken to obtain lower concentration ranges in sterile test tubes. 2.7 antibacterial aassay preparation of 0.5 mcfarland turbidity standards was done as described in nccls (nccls 2010). the agar well diffusion method was done following atata (2003). an overnight agar-culture of each bacterial isolate was made, and the suspension of microorganisms was made in sterile normal saline and adjusted to 0.5 mcfarland standards (108 cfu/ml) (nccls 2010). from the stock of 500 mg/ml extract, twofold serial dilutions were made to 250, 125, 62.5, and 31.25 mg/ml. each labeled mueller hinton agar plate was uniformly inoculated with a test organism by using a sterile cotton swab rolled in the suspension to streak the plate surface in a form that lawn growth could be observed. a sterile cork borer of 6mm diameter was used to make 5 wells on the medium in each plate. before boring of the well in agar, the cork borer was sterilized by dipping in alcohol and flaming. 50 µl of the 5 different extract concentrations were dropped into each well using a micropipette. all antibacterial assays were performed on duplicate plates. the underside of each well was appropriately labeled. other solvents used for extraction apart from water were tested for each organism. the inoculated plates were kept in the refrigerator for 1 hour to allow the extracts to diffuse into the agar (atata et al. 2003). the plates were incubated upright at 37°c for 24 hours. after incubation, the diameters of the zones of inhibitions obtained were measured, using a pair of calipers and meter ruler. the measurement was done at the back of the plate. the diameter was measured from one end of the zone to the other. where the zone of inhibition is not perfectly circular, the average of the long and short axis was used. the diameter of the zone of inhibition was obtained for i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 103 the two plates having the same concentration of the extract against a particular microorganism, and the average was used. for positive control, 30µl of 40 mg/ml of gentamycin was used while 50 µl of 2.5% dmso was used as negative control. 2.8 determination of minimum inhibitory concentration the minimal inhibitory concentrations (mics) of the extracts on the bacterial isolates were determined by macro broth dilution techniques following the recommendation of the clinical and laboratory standard institute (clsi 2015). one gram of the extract was dissolved in 1 ml of 20% dmso to get an extract concentration of 250 mg/ml. various serial dilutions were made from this stock solution in tubes of 1 ml sterile mueller hinton broths to get 125 mg/ml, 62.5 mg/ml, and 31.25 mg/ml. an overnight nutrient broth culture of the test bacterial isolate was standardized to 0.5 mcfarland turbidity standards. different dilutions of the suspension were made in a sterile normal saline to obtain a final inoculum concentration of 106 cfu/ml. then 1 ml of this adjusted inoculum was added to each tube of the mueller hinton broth containing different concentration of the crude extract. each tube was mixed and incubated at 37oc for 24 hours (nweze and onyishi 2010). this experiment was conducted in duplicate for all the bacterial isolates. a tube of mueller hinton broth containing only the 1ml suspension of the isolate without extract and the tubes of mueller hinton broth containing different concentrations of the extract without the isolate were used as controls. the tubes were examined after 24 hrs incubation. the mic of the extract was taken as the lowest extract concentration that completely inhibited the growth of the bacterial isolates in the tubes, as indicated by lack of visual turbidity. 2.9 determination of the use of mixtures of the extract this was done by measuring equal volume of each of extract type (ethanol, methanol and aqueous) of the plants seeds and then mixed. 50µl of each mixture was put into each well as in the antibacterial bioassay section to test the sensitivity potentials. this was done in triplicates. 2.10 statistical analysis all experiments were done in triplicate and the data thus obtained were reported as mean ± standard error of mean. statistical analysis was carried out to determine whether there was significant difference among the inhibitory actions of garcinia kola and cola nitida alone and in mixtures of extracts using analysis of variance and bonferroni post-test at 95% confidence level using graph pad prism version 5.01 (chao-hsun et al. 2010). i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 104 3 results total of 48 isolates were obtained from the 60 specimens collected from patients with urinary tract infections (uti) attending federal medical centre. bacterial species isolated include proteus vulgaris, escherichia coli, staphylococcus aureus, pseudomonas aeruginosa and klebsiella pneumoniae. table 1 shows the frequency distribution of bacterial isolates from patients with uti. escherichia coli had the highest occurrence (29.17%) while proteus vulgaris had the lowest occurrence (8.33%). table 1: frequency distribution of bacterial isolates (48) from patients with urinary tract infections. organisms number of isolates percentage occurrence proteus vulgaris escherichia coli staphylococcus aureus pseudomonas aeruginosa klebsiella pneumoniae 4 14 12 8 10 8.33 29.17 25.00 16.67 20.83 table 2 shows the zone of inhibition of the positive control (gentamycin) used against bacterial isolates. the zones of inhibition produced by the positive control were larger than the zones produced by the plant extracts. table 2: zones of inhibition (mm) of gentamycin (positive control) against bacterial isolates. organisms concentration (µl) zone of inhibition (mm) gentamycin a proteus vulgaris escherichia coli staphylococcus aureus pseudomonas aeruginosa klebsiella pneumoniae 30.0 30.0 30.0 30.0 30.0 28.00 26.00 18.00 26.00 29.00 a values are mean inhibition zone (mm) from three replicates the methanol extracts of g. kola, c. nitida and the mixture of extracts showed antibacterial activity against all the bacterial isolates at a concentration of 500 mg/ml, with g. kola having the largest zone of inhibition of 23 mm against e. coli and klebsiella pneumoniae respectively. c. nitida showed activity against klebsiella pneumoniae with largest zone of inhibition of 21.33 mm while the mixture of both extracts showed activity against p. vulgaris and e. coli with the zone of inhibition of 25.66 mm respectively (p< 0.05) as shown in table 3. i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 105 table 3: mean zone of inhibition (mm) of methanol extract of garcinia kola, cola nitida and mixture of extracts against bacterial isolates. values are given as mean ± standard error of three replicate plates ni = no inhibition all values of a particular isolate within the same column with shared alphabet superscript are non-significant (p>0.05). bacterial isolate plant extract concentration 500 mg/ml 250 mg/ml 125 mg/ml 62.5 mg/ml 31.5 mg/ml proteus vulgaris garcinia kola 21.33 ± 0.88a 20.00 ±1.15d 17.33 ± 0.67g 13.00 ± 0.58j ni cola nitida 17.33 ± 0.33b 16.00 ± 0.58e 13.66 ± 0.33h 11.33 ± 0.33j ni mixture of extracts 25.66 ± 0.33c 22.33 ± 0.33f 20.00 ± 0.58i 16.66 ± 0.33k 14.33 ± 0.33l escherichia coli garcinia kola 23.00 ± 0.58a 21.66 ± 0.88d 19.66 ± 0.33g 17.00 ± 0.58j 14.66 ± 0.33m cola nitida 19.00 ± 0.58b 16.00 ± 0.58e 14.00 ± 0.58h 12.33 ± 1.20k 0.00 ± 0.00n mixture of extracts 25.66 ± 0.33c 21.33 ± 0.33d 18.33 ± 0.33g 15.00 ± 0.58j 13.33 ± 0.33m staphylococcus aureus garcinia kola 17.66 ± 0.33a 16.00 ± 0.58cd 14.66 ± 1.20e ni ni cola nitida 17.66 ± 0.88a 15.66 ± 0.33c 0.00 ± 0.00f ni ni mixture of extracts 20.00 ± 0.58b 17.66 ± 0.33d 15.66 ± 0.33e ni ni pseudomonas aeruginosa garcinia kola 15.66 ± 0.33a 13.00 ± 0.58c 0.00 ± 0.00e ni ni cola nitida 16.33 ±0.67a 13.66 ± 0.33c 0.00 ± 0.00e ni ni mixture of extracts 19.00 ± 0.58b 16.00 ± 0.58d 10.33 ± 0.58f ni ni klebsiella pneumoniae garcinia kola 23.00 ± 0.58ab 20.33 ± 0.33c 18.33 ± 0.67e 14.00 ± 0.58g 0.00 ± 0.00h cola nitida 21.33 ± 0.33a 20.00 ± 0.58c 18.66 ± 0.33e 14.00 ± 0.58g 0.00 ± 0.00h mixture of extracts 25.33 ± 0.33ab 22.00 ± 1.15c 20.33 ± 1.20e 16.33 ± 0.33g 12.00 ± 0.58i i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 106 table 4: mean zone of inhibition (mm) of ethanol extract of garcinia kola, cola nitida and mixture of extracts against bacterial isolates. values are given as mean ± standard error of three replicate plates; ni = no inhibition all values of a particular isolate within the same column with shared alphabet superscript are non-significant (p>0.05). bacterial isolate plant extract concentration 500 mg/ml 250 mg/ml 125 mg/ml 62.5 mg/ml 31.5 mg/ml proteus vulgaris garcinia kola 20.00 ± 0.58a 18.00 ± 0. 58d 15.00 ± 0.58g 13.66 ± 0.33j ni cola nitida 16.00 ± 0.58b 14.33 ± 0.33e 12.00 ± 0.58h 0.00 ± 0.00k ni mixture of extracts 25.33 ± 0.33c 22.33 ± 0.33f 18.33 ± 0.88i 15.33 ± 0.88j ni escherichia coli garcinia kola 22.66 ± 0.67a 20.66 ± 0.67df 18.66± 0.33gi 16.66 ±0.88jl 15.66 ± 0.33m cola nitida 18.66 ± 0.67b 16.00 ± 0.58e 14.33 ± 0.33h 0.00 ± 0.00k 0.00 ± 0.00n mixture of extracts 25.66 ± 0.58c 21.33 ± 0.67f 17.33 ± 0.33i 15.00 ± 0.58l 13.00 ± 0.58o staphylococcus aureus garcinia kola 16.50 ± 0.33a 13.66 ± 0.33c 0.00 ± 0.00e ni ni cola nitida 17.00 ± 0.58ab 15.00 ± 0.58cd 13.33 ± 0.33f ni ni mixture of extracts 18.00 ± 0.58b 15.66 ± 0.33d 13.66 ± 0.67f ni ni pseudomonas aeruginosa garcinia kola 15.66 ± 0.33a 12.66 ± 0.33d 0.00 ± 0.00f ni ni cola nitida 14.50 ± 0.33a 13.83 ± 1.00de 0.00 ± 0.00f ni ni mixture of extracts 17.33 ± 0.33b 15.00 ± 0.58e 12.66 ± 0.33g ni ni klebsiella pneumoniae garcinia kola 21.33 ± 0.67ab 17.33 ± 0.33c 15.00 ± 0.58e 0.00 ± 0.00g ni cola nitida 19.66 ± 0.33a 16.66 ± 0.88c 14.00 ± 0.58e 0.00 ± 0.00g ni mixture of extracts 22.33 ± 0.33b 19.66 ± 0.88d 17.66 ± 0.33f 12.33 ±0.33h ni i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 107 ethanol extracts of g. kola, c. nitida and the mixture of extracts showed antibacterial activity against all the bacterial isolates at a concentration of 500 mg/ml. g. kola inhibited the growth of e. coli with 22.66 mm as the largest zone of inhibition while c. nitida showed its highest activity against klebsiella pneumoniae at zone of inhibition of 19.6 mm. furthermore, the mixture of extracts showed its effectiveness with the largest zone of inhibition of 25.66 mm against e. coli with significance of (p< 0.05) (table 4). table 5 shows the aqueous extracts of g. kola, c. nitida and the mixture of extracts against the bacterial isolates. g. kola, c. nitida and the mixture of extracts showed antibacterial activity against all the bacterial isolates at a concentration of 500 mg/ml except against staphylococcus aureus. the effect of g. kola was seen against p. vulgaris with the largest zone of inhibition of 17 mm while c. nitida was able to inhibit the growth of e. coli with the largest zone of inhibition of 16.66 mm. the mixture of both extracts equally showed effectiveness against p. vulgaris with the largest zone of inhibition of 20 mm which shows similar results with omwirhiren et al. (2016). table 5: mean zone of inhibition (mm) of aqueous extract of garcinia kola, cola nitida and mixture of extracts against bacterial isolates. * there was no inhibition detected against any of the bacterial isolates at 62.5 mg/ml 31.5 mg/ml concentrations of the aqueous extracts of single plant species or both plants in mixture, so that those two columns were not shown in the table. values are given as mean ± standard error of three replicate plates; ni = no inhibition; all values of a particular isolate within the same column with shared alphabet superscript are non-significant (p>0.05). table 6 shows the minimum inhibitory concentration (mic) values for the extracts of g. kola, c. nitida and the mixture of extracts against bacteria isolated from uti. the bacterial isolate plant extract concentration* 500 mg/ml 250 mg/ml 125 mg/ml proteus vulgaris garcinia kola 17.00 ± 0.58a 15.33 ± 0.33d 12.66 ± 0.33g cola nitida 12.66 ± 0.88b 10.66 ± 0.33e 0.00 ± 0.00h mixture of extracts 20.00 ± 0.58c 16.00 ± 0.58d 13.33 ± 0.33g escherichia coli garcinia kola 15.00 ± 0.58a 13.66 ± 0.33c 0.00 ± 0.00e cola nitida 16.66 ± 1.20a 14.66 ± 0.33c 12.33 ± 0.33f mixture of extracts 19.00 ± 1.00b 14.33 ± 0.33c 12.66 ± 0.33f staphylococcus aureus garcinia kola ni ni ni cola nitida ni ni ni mixture of extracts ni ni ni pseudomonas aeruginosa garcinia kola 12.33 ± 0.88a ni ni cola nitida 0.00 ± 0.00b ni ni mixture of extracts 13.00 ± 0.58a ni ni klebsiella pneumoniae garcinia kola 15.33 ± 0.67a 13.00 ± 0.58d ni cola nitida 0.00 ± 0.00b 0.00 ± 0.00e ni mixture of extracts 13.33 ± 0.33c 11.66 ± 0.33f ni i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 108 mic values of the extracts against the isolates were obtained from the agar diffusion assay. the lowest mic for g. kola, c. nitida and the mixture of extracts were obtained with methanol and ethanol extracts against all bacterial species respectively. however, there was no mic obtained for s. aureus, p. aeruginosa, and k. pneumonia for aqueous extract, and no mic obtained for s. aureus for aqueous mixture of extracts as well. table 6: minimum inhibitory concentration (mic) values of extract of g. kola, c. nitida and mixture of extracts against bacterial isolates. isolate plant minimum inhibitory concentration (mg/ml) methanol ethanol aqueous proteus vulgaris garcinia kola 62.5 62.5 125 cola nitida 62.5 125 250 mixture of extracts 31.25 62.5 125 escherichia coli garcinia kola 31.25 31.25 250 cola nitida 62.5 125 125 mixture of extracts 31.25 31.25 125 staphylococcus aureus garcinia kola 62.5 250 250 cola nitida 62.5 125 no mic mixture of extracts 62.5 125 250 pseudomonas aeruginosa garcinia kola 250 125 500 cola nitida 125 125 no mic mixture of extracts 125 62.5 500 klebsiella pneumoniae garcinia kola 62.5 62.5 250 cola nitida 31.25 62.5 no mic mixture of extracts 31.25 31.25 250 qualitative analysis on garcinia kola and cola nitida revealed the presence of important phytochemical constituents including phenolic compounds (tannins and flavonoids), saponins and alkaloids as bioactive compounds (table 7). table 7. qualitative phytochemical constituents of garcinia kola and cola nitida extracts. extract flavonoids tannins alkaloids terpenoid glycoside saponins resin protein g. kola + + + + + + + + c. nitida + + + + + + + + (key: + present; + + moderately present; + + + abundant; absent) quantitative analysis on garcinia kola and cola nitida revealed that phenolic compounds were a major class of bioactive components in the extracts table 8). the amount of total phenolics were 0.818 ± 0.021 mg and 0.700 ± 0.017 mg gae/mg of dry plant extracts of g. kola and c. nitida respectively, whereas the total flavonoid contents were as 25.63 ± 1.60 mg and 25.10 ± 1.85 rutin equivalents / g dry weight plant extract of g. kola and c. nitida respectively. i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 109 table 8: phytochemical constituents of garcinia kola and cola nititda (mean ± sd, n=3). extract phenolic contents * † total anthocyanin ‡ total flavonols ‡ total flavonoids total phenols non-tannins tannins g. kola 0.818 ± 0.021 0.507 ± 0.009 0.311 ± 0.001 4.64 ± 0.22 9.41 ± 0.02 25.63 ± 1.60 c. nitida 0.700 ± 0.017 0.376 ± 0.001 0.324 ± 0.007 2.15 ± 0.22 13.81 ± 0.92 25.10 ± 1.85 * expressed as mg gallic acid equivalents / mg dry weight plant extract † expressed as mg cyanidin 3-glucoside equivalents/100g of dry weight extract ‡ expressed as mg rutin equivalents / g dry weight plant extract table 9 shows the concentration of the extracts that inhibited 50% of the free radicals and lipid peroxidation (ic50) which was used to determine the potency of the extracts. the lower the ic50 value the better the extract potency. the plant extracts were efficient inhibitors of different free radicals compared to standard antioxidants. g. kola appears to be more efficient in inhibiting dpph radical (9.6 ± 1.0 μg/ml), superoxide anion (64.6 ± 1.5 μg/ml) and lipid peroxidation (282.9 ± 9.3 μg/ml) while c. nitida extract is a better inhibitor of hydroxyl radical (46.6 ± 2.5 μg/ml). table 9: free radical and lipid peroxidation inhibitory potency (ic50) of garcinia kola and cola nititda (mean ± sem, n=3). ic50 value for inhibitory potential (μg/ml) extract dpph radical hydroxyl radical (.oh) superoxide anion (o2.-) lipid peroxidation c. nitida 24.1 ± 2.1 46.6 ± 2.5 103.7 ± 5.2 575.1 ± 15.4 g. kola 9.6 ± 1.0 99.4 ± 1.7 64.6 ± 1.5 282.9 ± 9.3 standard antioxidant 4.1 ± 0.3 * 38.9 ± 2.8 # 3.3 ± 0.2 β 24.3 ± 1.4 £ * compared to ascorbic acid; # compared to α-tocopherol; β compared to rutin; £ compared to butylated hydroxyltoluene figure 1 shows the graphical representation of garcinia kola and cola nitida extracts which showed significant dose-dependent dpph radical scavenging capacity. garcinia kola appears to be more efficient, inhibiting 92.36 ± 1.31% of dpph at a concentration of 125 μg/ml compared to ascorbic acid which inhibited 94.18 ± 3.22 % at the same concentration. figure 2 shows the graphical representation of garcinia kola and cola nitida extracts scavenged .oh radical in a concentration dependent manner. the two extracts inhibited 2-deoxyribose degradation above 30% with maximal inhibition of 76.7 ± 1.4% at concentration of 500 μg/ml by s. s. the scavenging ability of the extracts was significant at all tested concentrations. c. nitida extract was found to be powerful quencher of .oh radical thereby preventing the propagation of lipid peroxidation. at high concentrations of both extracts lower activities were observed. i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 110 fig. 1: graph comparing dpph antioxidant activity of different concentrations of ascorbic acid and extracts of garcinia kola and cola nitida (values are expressed as mean ± sem, n = 3) fig. 2: graph comparing hydroxyl (.oh) antioxidant activity of different concentrations of αtocopherol and extracts of garcinia kola and cola nitida. values are expressed as mean ± sem, n = 3) figure 3 shows the garcinia kola and cola nitida extracts which inhibited the formation of reduced nbt in a dose-related manner. c. nitida showed the maximal i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 111 o2 .anion inhibitory activity of 86.68±2.10% at the concentration of 250 μg/ml, compared to rutin (96.03±2.2%, at 250 μg/ml). the o2 .scavenging effect of the extracts could culminate in the prevention of .oh radical formation since o2 .and h2o2 are required for .oh radical generation. fig. 3: graph comparing superoxide anion (o2.-) inhibition of different concentrations of rutin and extracts of garcinia kola and cola nitida (values are mean ± sem, n = 3) 4 discussion the involvement of bacteria in urinary tract infection is of great concern. female patients were only participant involved in this study because they are considered to be more predisposed to urinary tract infections. studies have shown escherichia coli and staphylococcus saprophyticus as mostly implicated in urinary tract infection (nicolle 2008). however, this study revealed the presence of escherichia coli as the highest occurring bacteria with 14 (29.17%), while proteus vulgaris as the lowest with 4 (8.33%) which agrees with mansour et al. (2009). the presence of these bacteria could possibly be because of poor sanitary hygiene. the findings from this study revealed the presence of alkaloids, saponins, tannins, flavonoids, glycoside, protein and the absent of terpenoids and resin in the extracts of garcinia kola and cola nitida in the methanol, ethanol and aqueous used as solvent as agreed with work done on c. nitida and g. kola by omwirhiren et al. (2016). studies have shown that the various solvents influence the nature of compounds extracted and their bioactivities (arunkumar and muthuselvam 2009, seanego 2012). however, methanol extracts appear to be most potent and promising as shown by its high i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 112 inhibitory activity against the clinical isolates. this could be attributed to the high presence of some of the polyphenolic compound identified (total flavonoids content of 25.63 ± 1.60 and 25.10 ± 1.85 mg rutin equivalents/g dry weight of g. kola and c. nitida respectively). these results clearly show that the solvent influences the extractability of the phenolic compounds. the phenolic extracts of plants are always a mixture of different classes of phenols, which are selectively soluble in the solvents. this finding is in conformity with previous studies by ukaoma et al. (2013) and alaje et al. (2014). the presence of these secondary metabolites is known to have therapeutic activity against several diseases and therefore could suggest the basis for their traditional use for the treatment of various illness (yousuf et al. 2012) including urinary tract infections. earlier studies have reported that flavonoids have antibacterial property as they have the capability to associate with soluble proteins and bacterial cell walls (doss et al. 2011). the evaluation of the antibacterial properties and the effect of mixture of extracts on bacterial isolates showed that they all possess antibacterial properties. the antibacterial activity was seen at varying concentrations indicating that the plant extract had broad antibacterial spectrum (bankole 1992). presence of alkaloids and flavonoids in g. kola and c. nitida has been observed to be responsible for its antibacterial property. however, the data obtained showed that the inhibitory effects of these plant extracts on the various bacterial isolates were dose dependent. this observation agrees with the findings of agbaje et al. (2006) and akinnibosun et al. (2009). the methanol extract of g. kola was most active against e. coli, k. pneumoniae, p. vulgaris, s. aureus, and p. aeruginosa with zones of inhibition ranging from 23.00 mm to 13.00 mm. the ethanol extract of g. kola extract was active against e. coli, k. pneumoniae, p. vulgaris, s. aureus, and p. aeruginosa with zones of inhibition ranging from 22.66 mm to 12.66 mm. while the aqueous g. kola was active against p. vulgaris, e. coli, k. pneumoniae, p. aeruginosa with zones of inhibition ranging from 17.00 mm to 12.33 mm and showed no zone of inhibition to s. aureus. this result is similar to the work of adegboye et al. (2008), who showed that the crude extract of g. kola exhibited antibacterial activities in vitro against both gram-positive and gram-negative organisms. the antibacterial properties of this plant could be attributed to the presence of tanins and flavonones. studies have shown it to have good antibacterial, antifungal and antiviral properties (terashima et al. 2002, adesuyi et al. 2012). other medicinal properties of the plant include its usage in the treatment of skin infections in liberia and congo democratic republic. the powdered bark of the plant is applied to malignant tumors and cancers, whereas the plants latex is taken internally for gonorrhea and externally to seal new wounds and prevent sepsis (adesuyi et al. 2012). in nigeria, a cold-water extract of the roots and bark with salt are administered to cases of bronchial asthma or cough and vomiting (adesuyi et al. 2012). methanol extract of c. nitida was most active against k. pneumonia, e. coli, s. aureus, p. vulgaris, and p. aeruginosa with zones of inhibition ranging from 21.33 mm to 11.33 mm. ethanol c. nitida extract was active against k. pneumoniae, e. coli, s. aureus, p. vulgaris, p. aeruginosa with zones of inhibition ranging from 19.66 mm to 12.00mm. while the aqueous c. nitida was active against e. coli, p. vulgaris with i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 113 zones of inhibition ranging from 16.66 mm to 10.66 mm and showed no zone of inhibition to s. aureus, p. aeruginosa and k. pneumoniae. the mixture of the extracts produced greater zones of inhibition on the bacterial isolates than the zones of inhibition produced by g. kola and c. nitida when used separately. the effect of mixture of plant extracts on the bacterial isolates was seen with methanol extract and least with aqueous extract. the methanol mixture of extracts was observed to be most active against p. vulgaris, e. coli, k. pneumoniae, s. aureus, and p. aeruginosa with zones of inhibition ranging from 25.66 mm to 12.00 mm. the ethanol mixture of extracts was seen to be active against e. coli, p. vulgaris, k. pneumoniae, s. aureus, and p. aeruginosa with zones of inhibition ranging from 25.66 mm to 12.33 mm. while the aqueous mixture of both plants was also observed to be active against p. vulgaris, e. coli, k. pneumoniae, p. aeruginosa with zones of inhibition ranging from 20.00 mm to 11.66 mm and showed no zone of inhibition to s. aureus. with this result, the effect of the mixture of the extracts showed greater antibacterial activity against the bacterial isolates when favourably compared with the standard antibiotic. although, no studies have shown this, that of solenostemon monostachyus and ocimum gratissimum (chukwura and iheukwumere, 2012) results showed they have greater inhibitory effect as g. kola and c. nitida extract mixture. the mic of the extracts against the bacteria was also determined varying between concentration of 31.25 mg/ml to 500 mg/ml for g. kola extract, c. nitida extract and the mixture of both extracts, respectively. the results of mic showed that the mixture of extracts is more potent against the bacterial isolates even at low concentrations. the broad spectrum of activity displayed by the extracts in this study appears to justify and explain the basis for their uses in traditional medicine, possibly as a remedy to the emergence of drug-resistant strains caused by inappropriate use of orthodox antibiotics. garcinia kola and cola nitida extracts showed significant dose-dependent dpph radical scavenging capacity. garcinia kola appears to be more efficient, inhibiting 92.36±1.31% of dpph at a concentration of 125 μg/ml compared to ascorbic acid which inhibited 94.18 ± 3.22 % at the same concentration as proven by okoko (2009). garcinia kola and cola nitida extracts scavenged .oh radical in a concentration dependent manner. the two extracts inhibited 2-deoxyribose degradation above 30% with maximal inhibition of 76.7±1.4 % at concentration of 500 μg/ml. the scavenging ability of the extracts was significant at all tested concentrations. the high radical scavenging activity of garcinia kola and cola nitida seems to be directly correlated with its total phenolic content as it may play an important role in their antioxidative effect. c. nitida extract was also found to be powerful quencher of .oh radical thereby preventing the propagation of lipid peroxidation. at high concentrations of both extracts lower activities were observed. garcinia kola and cola nitida extracts which inhibited the formation of reduced nbt in a dose-related manner. c. nitida showed the maximal o2 .anion inhibitory activity of 86.68±2.10% at the concentration of 250 μg/ml, compared to rutin (96.03 ± 2.2 %, at 250 μg/ml). the o2 .scavenging effect of the extracts could culminate in the prevention of .oh radical formation since o2 .and h2o2 are required for .oh radical generation. the observed ability of the extracts to scavenge or inhibit ho· radical indicated that the extracts could significantly inhibit i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 114 lipid peroxidation. this corroborates the studies of farshori et al. (2013) and olatunde et al. (2004) who reported that g. kola and c. nitida contains natural antioxidants. the ic50 is the concentration of the extracts that inhibited 50% of the free radicals and lipid peroxidation which was used to determine the potency of the extracts. the lower the ic50 value the higher the extract potency. the plant extracts were efficient inhibitors of different free radicals compared to standard antioxidants. g. kola appears to be more efficient in inhibiting dpph radical (9.6±1.0 μg/ml), superoxide anion (64.6±1.5 μg/ml) and lipid peroxidation (282.9±9.3 μg/ml) while c. nitida extract is a better inhibitor of hydroxyl radical (46.6±2.5 μg/ml). ic50 was calculated as the amount of antioxidant present in the sample necessary to decrease the initial dpph concentration by 50%. the lower the ic50 value the higher is the antioxidant activity. the observed antibacterial property of these seeds could therefore be linked to the presence of the phenolic compounds as they have previously been found to be main contributors of antioxidant activity and are also responsible for anti-inflammatory, antiviral, anticancerous and antimicrobial activities (yang et al. 2013). 5 conclusions the present study revealed the presence of phytochemicals in g. kola and c. nitida which exhibited promising antimicrobial activity against a broad spectrum of bacterial isolates. another striking finding was that the extracts showed free radical scavenging potential properties against the synthetic oxidative molecules and varying degrees of antibacterial activity on the bacteria isolated, with the methanol extract demonstrating the most effective activity against all the isolate at all concentrations. this therefore reaffirms the ethno-pharmacological importance of g. kola and c. nitida and could serve as the basis for advanced studies in the development of drugs against infectious diseases. this would also prove useful especially due to the alarming rate of drug resistance which is posing a threat and a major challenge in treatment of infectious diseases. apart from performing synergistic studies to evaluate the performance of g. kola and c. nitida when combined with orthodox medicine, there is also a need to determine the toxicity of the plant extracts which in our findings will be a prelude to initiating clinical trials in subsequent drug development. acknowledgments two anonymous reviewers are acknowledged for valuable comments on the initial draft of the manuscript. references adegboye mf, akinpelu da, okoh a. 2008. the bioactive and phytochemical properties of garcinia kola (heckel) seed extract on some pathogens. african journal of biotechnology 7 (21): 3934-3938. adesuyi ao, elumm ik, adaramola fb, nwokocha agm. 2012. nutritional and phytochemical screening of garcinia kola. advanced journal of food science and technology 4(1): 9-14. i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 115 agbaje eo, ogunsanya t, aiwerioba oir. 2006. conventional use of honey as antibacterial agent. annals of african medicine 5(2): 78-81. agbo ag, ngogang yj. 2005. toxicity of herbal preparations. cameroon journal of ethnobotanical 1(1): 23-28. agbo ag, oben eo, ngogang yj. 2005. haematinic activity of hibiscus cannbinnus. african journal of biotechnology 2(4): 833-837. ajayi ia, jonathan sg, adewuyi a, oderinde ra. 2008. antimicrobial screening of essential oil of some herbal plants from western nigeria. world applied sciences journal 3(1): 79-81. akinnibosun ha, akinnibosun fi, adieme bc. 2009. biotherapeutic potential of aqueous and ethanolic extracts of solenostemon monostachyus (p. beauv.) brig. leaves on some vegetative gram-negative bacteria. biological and environmental sciences journal for the tropics 6(3): 33-37. akoachere jf, ndip rn, chenwi eb. 2002. antibacterial effect of zingiber officinale and garcinia kola on respiratory tract pathogens. east african medical journal 79: 588-592. alade pi, irobi on. 1993. antimicrobial activities of crude extracts of acalypha wilkesiana. journal of ethnopharmacology 39: 1-9. alaje do, owolabi kt, olakunle tp, oluoti oj, adetuberu ia. 2014. nutritional, minerals and phytochemicals composition of garcinia cola [bitter cola] and aframomum melegueta (alligator pepper) iosr journal of environmental science, toxicology and food technology 8(1):86-91. amalu pc, chukwuezi fo, ugwu opc. 2014. antimicrobial effects of bitter kola (garcinia kola) nut on staphylococcus aureus, escherichia coli and candida albicans. journal of dental and medical sciences (iosr-jdms) 13(4): 29-32. arekemase mo, aliyu mb, kayode rmo, ajiboye ae, ajijolakewu ak. 2012. antimicrobial effects of garcinia kola (bitter kola) on some selected pathogens from university of ilorin teaching hospital ilorin, nigeria” journal of asian scientific research 2(4): 159-169. arunkumar s, muthuselvam m. 2009. analysis of phytochemical constituents and antimicrobial activities of aloe vera against clinical pathogens. world journal of agricultural sciences 5(5): 572576. atata r, sani a, ajewole sm. 2003. effect of stem bark extracts of enantia chlranta on some clinical isolates. biokemistri 15(2): 84-92. atawodi se, mende p, pfundstein b, preussmann r, spiegelhalder b. (2005). nitrosatable amines and nitrosamide formation in natural stimulants: cola acuminate. c. nitida and garcinia kola. food and chemical toxicology 33: 625-630. bankole aa. 1992. the antibacterial activity of the crude seed extracts of three plants used in nigerian ethnomedicine. university of benin, benin city, p. 45. beauchamp c, fridovich i. 1971. superoxide dismutase: improved assay and an assay applicable to acrylamide gels. analytical biochemistry 44: 276-287. chabot s, bel-rhlid r, chenevert r, piche y. 2002. hyphal growth promotion in vitro of the va mycorrhizal fungus, gigaspora margarita becker and hall, by the activity of structurally specific flavonoid compounds under co2-enriched conditions. new phytologist 12: 461-467. chao-hsun, y, yu-chun h, yu-fen c, ming-hsiang, c. 2010. foam properties, detergent abilities and longterm preservative efficacy of the saponins from sapindus mukorossi. journal of food drug analysis 18: 155-160. chen k, shi q, kashiwada y, zhang dc, hu cq, jin jq, nozaki hre, kilkuskie t, cheng yc. 2002. anti-aids agents. 6. salaspermic acid, an anti-hiv principle from tripterygium wilfordii, and the structure-activity correlation with its related compounds. journal of natural products 55: 340-346. chessin m, deborde d, zip fa. 2005. antiviral proteins in higher plants. crc press, inc., boca raton, fla. chukwura ei, iheukwumere ia. 2012. determination of synergistic effect and activity index of solenostemon monostachyus and ocimum gratissimum on selected bacteria. nigerian journal of microbiology 26: 2492-2498. clsi. 2015. clinical and laboratory standards institute, performance standards for antimicrobial susceptibility testing; twenty-fifth informational supplement clsi document m100-s25 35: 3. cowan mm. 2009. plant products as antimicrobial agents. clinical microbiology reviews 12: 564-582. dalziel jm. 2008. the useful plants of west tropical africa. crown agents for the colonies, london 4548. i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 116 doss a, parivuguna v, vijayasanthi m, sruthi s. 2011. antibacterial and preliminary phytochemical analysis of medicago sativa l. against some microbial pathogens. indian journal of science and technology 4(5): 550-552. el-mahmood am, doughari jh, chanji fj. 2008. in vitro antibacterial activities of crude extracts of nauclea latifolia and daniella oliveri. scientific research and essays 3(3): 102-105. ezeigbo or, ejike en, nwachukwu i, ejike bu. 2016. comparative antibacterial activity of methanolic, ethanolic and aqueous extract of garcinia kola (bitter kola) and cola nitida (kola nut). international journal of plant research 6(3): 53-56. fabiola bh, tania u, benedito, pdf, diogenes agc, jose am, celso vn. 2003. effect of essential oil of ocimum gratissimum on the trypanosomatid. herpetomonas samuelpessoai. protozoologica 42: 269-276. farshori nn, al-oquail mm, al-shedde ss, siddequi ma. rauf a. 2013. antimicrobial potentiality of polyalthia longifolia seed oil against multi drug resistant (mdr) strains of bacteria and fungus of clinical origin. african journal of microbiology research 19: 1977-1982. gabriel a, agbor dk, julius eo. 2007. medicinal plants can be good sources of antioxidants: case study in cameroon. pakistan journal of biological sciences 10(4): 537-544. giusti mm, wrolstad re. 2001. characterisation and measurement of anthocyanins by uv-visible spectroscopy. in: r. e. wrolstad (ed.). current protocols in food analytical chemistry. new york: john wiley and sons. pp 45-65. green rj. 2004. antioxidant activity of peanut plant tissues. masters’ thesis. north carolina state university. united states of america. grieve m. 2001. kola nuts (www.botanical.com). gyamfi ma, yonamine m, aniya y. (1999). free-radical scavenging action of medicinal herbs from ghana: thonningia sanguine on experimentally-induced liver injuries. general pharmacology 32: 661–667. halliwell b, gutteridge jmc, aruoma oi. 1987. the deoxyribose method: simple ‘test-tube’ assay for determination of rate constants for reactions of hydroxyl radicals. analytical biochemistry 165: 215219. harborne jb. 1984. phytochemical methods: a guide to modern technique of plant analysis, 2nd eds. london: chapman and hall, london. pp 271. iwu mw, duncan ar, okunji co. 2009. new antimicrobials of plant origin. in: janick j. (ed.) perspective on new crops and new uses. alexandria, va: pp. 459-462. jonathan sg, fasidi io. 2003. antimicrobial activities of lycoperdon pusilum (bat.ex.) and lycoperdon giganteum (pers.) nigerian edible macrofungi. african journal of biomedical research 6: 88-90. jonathan sg, fasidi io. 2005. antimicrobial activities of some selected nigerian mushrooms. african journal of biomedical sciences 8(2): 83-87. jonathan g, ohimain e, kigigha l. 2007. antagonistic effects of extracts of some nigerian higher fungi against selected microorganisms. american-eurasian journal of agricultural environmental sciences 2(4): 364368. kanomal ai, muhammad i, abdullahi s, shehu k, maishanu hm, isah ad. 2014. journal of biology agriculture and healthcare 4(3) issn 2224-3208. keay rwj, onochie cfa, stanfield dp. 2014. nigerian trees. vol. 1 department of forest research, ibadan, nigeria pp 51-54. kuete v, ngameni b, fotso-simo cc, kengap-tankeu r, tchaleu-ngadjui b, meyer jj, lall n, kuiate jr. 2008. antimicrobial activity of the crude extracts and compounds from ficus chlamydocarpa and ficus cordata. journal of ethnopharmacology 120: 17-24. kumaran a, karunakaran j. 2006. in vitro antioxidant activities of methanol extracts of five phyllanthus species from india. lwt-food science technology 40: 344–352. maiyo zc, ngure rm, matasyoh jc, chepkorir r. 2010. phytochemical constituents and antimicrobial activity of leaf extracts of three amaranthus plant species. african journal biotechnology 9(21): 31783182. makkar hp, bluemmel m, borowy nk, becker k. 1993. gravimetric determination of tannins and their correlations with chemical and protein precipitation methods. journal of the science and food agriculture 61: 161-165. http://www.botanical.com/ i.v. anyiam and p.p.e. mounmbegna bioactivities of cola nitida and garcinia kola extracts ruhuna journal of science vol 11 (2): 98-117, december 2020 117 mansour a, manijeh m, zohreh p. 2009. study of bacteria isolated from urinary tract infection and determination of their susceptibility to antibiotics. journal of microbiology 2(3): 118-123. martinez ac, marcelo el, marco ao, moacyr m. 2001. differential responses of superoxide dismutase in freezing resistant solanum curtibolum and freezing sensitive solanum tuberosum subjected to oxidative and water stress. plant science 160: 505-515. nccls. 2010. national committee for clinical laboratory standards, performance standards for antimicrobial disk susceptibility tests: approved standard m2-a7. national committee for clinical laboratory standards. wayne; pennsylvania, united states of america. nicolle le. 2008. uncomplicated urinary tract infection in adults including uncomplicated pyelonephritis. urologic clinics of north america 35(1): 1-12. nweze ei, onyishi mc. 2010. in vitro antimicrobial activity of ethanolic and methanolic extracts of xylopia aethiopica and its combination with disc antibiotics against clinical isolates of bacteria and fungi. journal of rural and tropical public health 9: 1-6. okoko t. 2009. chromatographic characterization, invitro antioxidant and free radical scavenging activities of garcinia kola seeds. african journal of biotechnology 8: 7133-7137. olatunde fe, hansen m, rain-haren p, dragsted lo. (2004). commonly consumed and naturally occurring dietary substances affect bio markers of oxidative stress and dna damage in healthy rats. food and toxicology 42(8): 1315-1322. omwirhiren em, james sa, asefon oa. 2016. the phytochemical properties and antimicrobial potentials of aqueous and methanolic seed extract of cola nitida (vent.) and cola acuminate (beauvoir) grown in south west, nigeria. saudi journal of medical and pharmaceutical sciences 2(12): 354-363. opeke lk. 2012. tropical tree crops. spectrum books ltd. ibadan, nigeria. pudhom n, thakur ar, chaudhuri sr. 2007. characterization of bacterial strains isolated through microbial profiling of urine samples. journal of biological science 7(1): 44-51. rabe t, vanstoden j. 2000. isolation of an antimicrobial quiterpenoid from warbugie salutaris. journal of pharmacology 93: 171-174. seanego cl. 2012. phytochemical analysis and bioactivity of garcinia kola (heckel) seeds on selected bacterial pathogens. unpublished m.sc. (microbiology) thesis, university of forte hare. sofowora a. 1993. medicinal plants and traditional medicine in africa. spectrum books ltd, ibadan, nigeria pp 289-294. terashima k, takaya y, niwa m. 2002. powerful antioxidative agents based on garcinoic acid from garcinia kola. bioorganic and medicinal chemistry 10(5): 1619-1625. trease ge, evans wc. 2001. a textbook of pharmacognosy, 14th edn. w.b. saunders company ltd. london; pp 193-201, 232-233, 249-252. ukaoma aa, ukaoma vo, okechukwu ri, iwuagwu, m. 2013. phytochemical screening and antibacterial properties of garcinia kola. the journal of phytopharmacology 2(3): 34-38. velioglu ys, mazza g, gao l, oomah bd. 1998. antioxidant activity and total phenolics in selected fruits, vegetables, and grain products. journal of agricultural and food chemistry 46: 4113-4117. who. 2000. world health organization, expert committee on medicinal importance of native plants. technical report series. who-geneva. wolfe k, xianzhong w, rui hl. 2003. antioxidant activity of apple peels. journal of agricultural and food chemistry 51: 609-614. yang c, chang h, lin h, chuang l. 2013. evaluation of antioxidant and antimicrobial activities from 28 chinese herbal medicines. journal of pharmacognosy and phytochemistry 2(1): 294-305. yousuf m, aslam k, wani ba, aslam n, dar na, nawchoo ia. 2012. in vitro antibacterial activity and phytochemical studies of methanolic extract of leaves of hypericum perforatum l. growing wild in kashmir himalaya. asian journal of plant science and research 2(4): 414-420. zgoda jr, porter ja. 2001. convenient microdilution methods for screening natural products against bacteria and fungi. pharmaceutical biology 39: 221-225. sv-lncs ruhuna journal of science vol. 5: 2014 http://rjs.ruh.ac.lk/ issn: 1800-279x  faculty of science, university of ruhuna  faculty of science, university of ruhuna list of reviewers, ruhuna journal of science volume 5 (2014) the list includes also the reviewers contributed for the rejected/withdrawn papers submitted for volume 5. prof. pushpa d. abeysinghe, department of botany, faculty of science, university of ruhuna, matara, sri lanka prof. saman abeysinghe, department of botany, faculty of science, university of ruhuna, matara, sri lanka dr. nilanthi dahanayake, department of agricultural biology, faculty of agriculture, university of ruhuna, kamburupitiya, sri lanka prof. w.g.d. dharmaratna, department of physics, faculty of science, university of ruhuna, matara, sri lanka prof. ranil fernando, department of surgery, university of kelaniya, ragama, sri lanka dr. d. l. c. kumari fonseka, department of crop science, faculty of agriculture, university of ruhuna, kamburupitiya, sri lanka dr. priyani hettiarachchi, department of biological sciences, faculty of applied sciences, rajarata university of sri lanka, mihinthale, sri lanka dr. w. g. d. lakmini, department of crop science, faculty of agriculture, university of ruhuna, kamburupitiya, sri lanka dr. sumedha madawala, department of botany, faculty of science, university of peradeniya, peradeniya, sri lanka dr. a. janaki mohotti, department of crop science, faculty of agriculture, university of peradeniya, peradeniya, sri lanka dr. keerthi mohotti, tea research institute of sri lanka, talawakele 22100, sri lanka prof. chandi rajapaksha, department of soil science, university of peradeniya, peradeniya, sri lanka prof. p. ravirajan, department of physics, faculty of science, university of jaffna, jaffna, sri lanka dr. m. b. samarawickrama, department of pathology, faculty of medicine, university of ruhuna, karapitiya, galle, sri lanka dr. g.k.r. senadheera, department of physics, open university of sri lanka, kandy center, kandy, sri lanka dr. t.s. suresh, department of biochemistry, faculty of medical sciences, university of sri jayewardenepura, nugegoda, sri lanka prof. ira thabrew, institute of biochemistry & biotechnology, university of colombo, colombo, sri lanka ruhuna journal of science vol 10(1): 51-64, june 2019 eissn: 2536-8400 faculty of science doi: http://doi.org/10.4038/rjs.v10i1.50 university of ruhuna  faculty of science, university of ruhuna 51 sri lanka development of a novel and benign emulsion paint binder from epoxidized soybean oil and dimethylol urea blend ayodele akinterinwa * , sunday ogwuche, s.a. osemeahon and abdulraheem o. anumah department of chemistry, modibbo adama university of technology, yola, nigeria *correspondence: ayoterinwa@yahoo.com; https://orcid.org/0000-0001-5236-9238 received: 11 th october 2018, revised: 30 th april 2019, accepted: 18 th may 2019. abstract to develop a novel emulsion paint binder, dimethylol urea (dmu) resin synthesized from low formaldehyde to urea stoichiometric ratio (2:1) was blended with epoxidized soybean oil (esbo) with 2.56% oxirane oxygen and 67% iodine value conversion. esbo/dmu blends were prepared by varying the amount of esbo from 10–70%. the blends were analyzed for formaldehyde emission, viscosity, refractive index, turbidity, density, gelation time and solubility in water, while the blends’ films were analyzed for moisture uptake and melting point. homogenous blends were obtained up to 50% esbo/dmu blend, while ftir analyses show reactive esbo/dmu blend at optimum blending. formaldehyde emission and moisture uptake were reduced in the blends’ film, while esbo also act as plasticizer by introducing softness in the blends’ film. comparison of the properties of 40% esbo/dmu optimum blend with other reported binders shows its good potentials and its capability to compete in the coating industry. keywords: binder, dimethylol urea, epoxidation, emulsion paint, soybean oil. 1 introduction health and environmental impacts of painting and coating cannot be overemphasized; as we must stay “beautiful internally” (healthy), to enjoy and appreciate the external beautification. when emulsion or oil paints dries on a substrate, water or organic thinner vapor is released respectively. organic thinners also referred to as volatile organic compounds are hazardous, this is why the production of emulsion/water based paints is progressively being researched to obtain products with competitive properties compared to oil based paint so as to replace them and avoid or limit the hazardous release of volatile organic compounds to safe and acceptable levels in the environment. properties of urea formaldehyde (uf) resin such as solubility in water, fast curing, formation of hard, colorless, glossy, electrical-resistant and water mailto:ayoterinwa@yahoo.com https://orcid.org/0000-0001-5236-9238 a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of science 52 vol 10(1): 51-64, june 2019 insoluble film selects it as a potential emulsion paint binder. however, intrinsic properties of the film such as brittleness, poor moisture resistance which will lead to short term failure, and a hazardous emission of formaldehyde are limiting factors in its application in large surface coating (edoga 2006, osemeahon et al. 2015). a trend of research is focused on various modifications of uf resin to ameliorate its impeding properties as efficient binder in the formulation of emulsion paint. among these modifications are some reported modulations in the resin synthesis (barminas and osemeahon 2007, suurpere et al. 2006, osemeahon et al. 2015). the modulation to synthesize dmu from formaldehyde and urea using reaction stoichiometry ratio 2:1 respectively, was aimed at the reduction of formaldehyde emission from the resin (osemeahon et al. 2015). other modifications include copolymerization and blending with other resins (edoga 2006; osemeahon and barminas 2007, fink 2013, osemeahon et al. 2015, idowu-oyewole et al. 2016). blending is a common procedure that leads to the formation of new substances with outstanding physical, chemical and mechanical properties, with novel applications (pizzi et al. 2002). uf resins have been blended with various substances and the resultant products have shown various improvements in the intrinsic properties limiting the application of the pure resin. however, reports on dmu copolymer blends are scarce. osemeahon et al. (2015) blended dmu with polystyrene; this increased moisture resistance, impact softness/flexibility but could not appreciably reduce formaldehyde emission from the film at optimum blend, hence retaining some challenges. soybean oil (sbo) is available and cheap. it is polyunsaturated with relatively high double bond, and these points of unsaturation present site for such reaction as epoxidation (saithai et al. 2013). epoxidation reaction introduce reactive oxirane ring in the structure of oil (figure 1), thereby making the oil more interactive (saithai et al. 2013, cheng et al. 2015, pantone et al. 2017). epoxidized oils serve as plasticizer/softener in rigid/hard polymers. it has been involved in coatings formulation, one of these is acrylated esbo involved in the production of an emulsion paint binder (habib and bajpai 2011). soybean oil epoxidized soybean oil fig. 1. typical epoxidation reaction. (adopted from pantone et al. 2017) a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of scienc 53 vol 10(1): 51-64, june 2019 in this study, esbo was blended with dmu. a reactive homogenous blend expected from the two substances is anticipated to present a new product (esbo/dmu) with tailored properties acceptable in coating application. esbo/dmu is expected with potentials that will present a competitive alternative among the emulsion paint binder being considered in the coating industry. 2 material and methods 2.1 materials chemicals used include urea, formaldehyde, sodium dihydrogen phosphate, water, sulphuric acid, sodium hydroxide pellets, soybean oil, hydrogen peroxide (30 wt%), glacial formic acid, na2s2o3, naoh solution, lithium aluminate, dams reagent, starch indicator and phenolphthalein, pyridine, bromine, potassium iodide, sodium thiosulphate, chloroform, ethylether, hydrochloric acid and ethanol. all chemicals are analytical grade supplied by the british drug house. refined soybean oil by agarwal industries ltd, india, was purchased from the market. 2.2 epoxidation of soybean oil method by saithai et al. (2013) was adopted with slight modifications as follows. soybean oil (100 g) was poured in to a 500 ml three neck flask equipped with reflux condenser assembled in a water bath with stirrer. formic acid (13.97 g) was added and the mixture was stirred and heated up to 45°c. sulphuric acid (0.5 ml) was included in the mixture, the temperature was raised up to 60°c, and hydrogen peroxide (116.98 g) was finally added drop wisely. molar ratio of the soybean oil (double bonds in oil) to formic acid to hydrogen peroxide was determined to be 1:2.64:8.9. the reaction was allowed to run for 6 hours, cooled and the oil phase was decanted. the epoxidized oil was washed with warm distilled water and dried at 50°c in an oven for about 5 hours. 2.3 determination of iodine value and conversion iodine value of soybean oil before (iv0) and after (ivt) epoxidation was determined according to wijs method (aoac 2000). iodine value conversion (% x) which also measures the relative percentage loss in un-saturation was calculated using equation 1. ............................ (1) a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of science 54 vol 10(1): 51-64, june 2019 2.4 determination of oxirane oxygen content this was carried out using the method by siggia (1963). esbo (0.20g) was quantitatively transferred into a conical flask. a hydro-chlorinated reagent was prepared from 0.2m hcl and ethyl ether (1:1 v/v), and 10 ml was added into the conical flask and the mixture was allowed to stand for 3 hours. blank solution was also prepared and the solutions were titrated with 0.1m sodium hydroxide solution using phenolphthalein as an indicator. percentage oxirane content which also measures the stable epoxy groups on the epoxidized oil was calculated according to equation 2. oxirane oxygen = – ...................................... (2) where, s= volume of naoh used for sample (ml), b= volume of naoh used for blank (ml), m= molarity of the naoh, w= weight of sample (epoxidized oil) used (g) and 16 g/mol = atomic mass of oxygen. 2.5 synthesis of dmu, esbo/dmu blends and determination of physical properties dimethylol urea (dmu) was synthesized according to the one step process (osp) by barminas and osemeahon (2007), as modified in our previous work (osemeahon et al. 2015). esbo/dmu blends were prepared by a starring a mixture of 10, 20, 30, 40, 50, 60 and 70 % esbo in dmu resin, at room temperature (30 °c) till a homogenous mass was obtained. relative viscosity was determined at room temperature (30°c) using a graduated glass macro-syringe standardized with 20% (w/v) sucrose solution (viscosity= 2.0 x 10 -3 nsm -2 at 30ºc), and the gel time was recorded as the time at which a constant viscosity profile was obtained. formaldehyde emission was determined using the standard 2 hour desiccators test. moisture uptake form the atmosphere by the films was gravimetrically determined (barminas and osemeahon 2007). density, turbidity (hanna microprocessor turbidity meter, model h193703), melting points (galenkamp melting point apparatus model mfb600-010f) and refractive index (abbe refractometer) were carried out according to standard methods (aoac 2000). in all analysis, averages of triplicate determinations were recorded. 2.6 ftir analysis dimethylol urea (dmu) and esbo/dmu blend was analyzed using infra-red spectrophotometer (buck scientific inc, ct, usa model m500) between 500 and 4000 cm -1 . a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of scienc 55 vol 10(1): 51-64, june 2019 3 results and discussion iodine value of sbo (iv0) was determined to be 84.8 g of i2/ 100g oil. after epoxidation (ivt) the value dropped to 32.36 g of i2/ 100g oil and the iodine value conversion was calculated as 61.84 %. this indicated that un-saturation (double bonds) has been reduced in the oil. percentage oxirane oxygen of the esbo was obtained as 2.56%, indicating that some point of un-saturation in the oil has been substituted by the oxirane oxygen. 3.1 ftir spectra figure 2 compares the spectra of esbo/dmu blend with dmu. on dmu, peaks between 3700–3200 cm -1 attributes to absorption by o-h and n-h, peaks within 2000–1000 cm -1 attributes to carbonyl (c=o) and c-o-c at the methylene ether linkages. fig 2. ftir spectra of dmu and 40% esbo/dmu blend the sharp peaks at between 3000 2500 cm -1 is due to the combined group (– nh-ch2-oh) on the dmu resin (figure 3). the blend’s spectra show a significant change in intensity, and this may be attributed to interaction between esbo and dmu. the reduction observed for the –nh-ch2-oh peak indicates the involvement of this group in the interaction with esbo as shown in figure 3. there is no oxirane characteristic peak (at 820 cm -1 ) on the blend a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of science 56 vol 10(1): 51-64, june 2019 spectra; showing that the esbo oxirane group could be involved in the interaction with dmu resin. fig. 3. synthesis of dmu resin and a proposed interaction between esbo and dmu in the blend. 3.2 effect of esbo concentration on the physicochemical properties of esbo/dmu blend binder viscosity is a parameter that must be optimized for an optimum paint production and application processes, so as to correct paint leveling, sagging, flow and adhesion (suurpere et al. 2006, idowu-oyewole et al. 2016). figure 4 presents the effects of esbo on the viscosity of the esbo/dmu blend. fig. 4. effect of esbo content on the viscosity of esbo/dmu blends a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of scienc 57 vol 10(1): 51-64, june 2019 the blend’s viscosity increased with the inclusion and increase in the amount of esbo up to 30%, this peak viscosity was maintained till 40% after which a sharp drop in viscosity was observed. the initial effect can be attributed to homogeneity/interactions between esbo and dmu leading to increase in the average molecular weight and dispersity index of the blend (vilas et al. 2001, oladipo et al. 2013). this can be said to reach optimum at the equilibrium concentration of esbo, above which bulk of the blend is heterogeneous. binder density influence many important properties of the paint formulated with it (idowuoyewole et al. 2016). the change in density of esbo/dmu blends with variation in the esbo content is presented in figure 5. increase in esbo gradually decreases the density of the blends. this can be attributed to increase in the specific and free volume in the blended polymer consequent of changes in the morphology of dmu and the blends as esbo increases in the blends (osemeahon et al. 2015). osemeahon et al. (2015) reported similar observation for a blend of dmu/polystyrene resin. fig 5. effect of esbo content on the density of esbo/dmu blends optical properties of binder may influence the paint’s aesthetic properties (oladipo et al. 2013). the effect esbo variations on the optical property of the blends were presented for refractive index and turbidity in figures 6 and 7 respectively. figure 6 showed a sharp rise in the refractive index of the blend at 10% esbo (inclusion of esbo), followed by a gradual increase till 50 %. this may be due to increasing homogeneity/interaction up till 50% esbo leading to increase in light scattering boundaries and discontinuities in the blends (liem et al. 2002). a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of science 58 vol 10(1): 51-64, june 2019 fig. 6. effect of esbo content on the refractive index of esbo/dmu blends in figure 7, the clear dmu resin turned colloidal as the turbidity of dmu increased on inclusion of esbo (at 10%). however, subsequent increase in esbo amount in the blends did not raise the turbidity level further. this effect can be attributed to the haze imparted to the clear dmu resin on addition of esbo which is yellowish in color. similar effect was reported for polystyrene in dmu (osemeahon et al. 2015). fig. 7. effect of esbo content on the turbidity of esbo/dmu blends one of the setbacks of urea formaldehyde resin in coating application is the film’s poor resistance to moisture uptake (salthammer et al. 2010). low moisture resistance in paint binder initiates paint failure such as blistering, a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of scienc 59 vol 10(1): 51-64, june 2019 alligatoring and brooming (akinterinwa et al. 2015). figure 8 shows that there was a sharp decrease in the amount of moisture uptake when esbo was included in dmu (at 10%), followed by a gradual decrease as the amount of esbo increases in the blends. the sharp decrease in moisture uptake can be attributed to the formation or substitution with hydrophobic functionalities form the interaction between esbo and dmu (cakir et al. 2012). this interaction can be said to quickly approach equilibrium with increase in the esbo amount in the blends. fig. 8. effect of esbo content on the moisture uptake of esbo/dmu blends fig. 9. effect of esbo content on the melting point of esbo/dmu films dry film of urea formaldehyde is hard and prone to brittleness, therefore, impacting softness/flexibility is a favorable modification of the resin. melting point of polymers is related to flexibility (yildiz et al. 2007). figure 9 shows a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of science 60 vol 10(1): 51-64, june 2019 the effect of esbo concentration on the melting temperature of esbo/dmu blends’ films. the melting temperature of the blends’ films decreased with increase in the amount of esbo till 50%, after which it drops sharply. this shows that the blends’ films increased in flexibility as the esbo content increases. above 50%, esbo’s soft film forms bulk of the heterogeneous films, this attributes to the low melting temperatures. formaldehyde emission during resin cure is another drawback in urea formaldehyde’s application in coating (salthammer et al. 2010, akinterinwa et al. 2015). figure 10 shows the effect of esbo on formaldehyde emission from blends. formaldehyde emission from the blends decreases as the esbo content was increased till 50%, after which the emission remains relatively constant. reduction in formaldehyde emission may be expected due to the reduction in the amount of dmu present in the blends as the esbo content increases. however, the rapid reduction in emission on inclusion of esbo can be attributed to interactions involving the hydroxyl end of dmu, hindering condensation reactions leading to the emission of formaldehyde (osemeahon et al. 2015). optimum formaldehyde emission (at 40% esbo) is 0.045 ppm, this is appreciably lower than the 0.08 ppm maximum acceptable level in the coating industry (salthammer et al. 2010). fig 10. effect of esbo content on formaldehyde emission of esbo/dmu films. gel or gelation time is the time taken by a substance in a fluid state to dry or set into a rigid structure with mechanical properties. figure 11 shows that the gel time decreases with increase in esbo content in the blend till 50%, after which the dry time increased again. a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of scienc 61 vol 10(1): 51-64, june 2019 fig 11. effect of esbo content on the gelation time of esbo/dmu films this effect can be attributed to an efficient blending/interaction between esbo and dmu resulting in the formation of a substance with unique mechanical properties compare to esbo and dmu. to check the tenability of the blends as possible binder for emulsion paint, its solubility or stable disposability in water was recorded in table 1. esbo/dmu blends uniformly disperse in water up to 40% esbo, above which there was decline in stable disposability in water. this is due to the oil phase which increasingly forms the bulk phase of the blends with esbo more than 40%. table 1: effect of esbo content on the solubility of esbo/dmu blends. concentration (%) solubility 0 soluble 10 soluble 20 soluble 30 soluble 40 soluble 50 sparingly soluble 60 insoluble 70 insoluble a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of science 62 vol 10(1): 51-64, june 2019 table 2. comparison of some properties of esbo/dmu (40%) with other paint binders. type properties density (g/dm 3 ) refractive index melting point (c°) moisture uptake (%) formaldehyde emission (ppm) viscosity (mpa.s) turbidity (ntu) gel time (hours) literature esbo/dmu 1.07 1.46 155.1 0.98 0.045 51.01 880 48 this work dmu/ps 1.1953 1.4295 181.42 0.26 0.0385 35.60 611 24 osemeahon et al. 2015 commercial uf nd nd nd 2 nd 451 nd 54.2 suurpere et al. 2006 tmu/nr 0.5708 1.3501 260 nd 0.058 220 340 nd osemeahon and barminas 2007 xaso 0.912 1.467 nd nd nd 900 nd nd oladipo et al. 2013 innovative uf nd nd nd 0.25 0.07 nd nd nd zorba et al. 2008 aeso nd nd nd nd nd 13,000 nd nd habib and bajpai 2011 rsomar 0.95 nd nd nd nd 3.11 nd 24 aigbodion and pilla 2001 key: tmu: trimethylol urea, ps: polystyrene, uf: urea formaldehyde, nr: natural rubber, xaso: ximenia americana seed oil, aeso: acrylated epoxidized soybean oil, er: epoxy resin, rsomar: rubber seed oil modified alkyd resin, nd: not determined. a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of scienc 63 vol 10(1): 51-64, june 2019 3.3 comparison of esbo/dmu binder with other reported binders being of optimum properties, table 2 compares 40% esbo/dmu blend with those of some reported binders. looking through the data presented on table 2, it can be said that esbo/dmu at optimum blend has good potentials to compete among other emulsion paint binders recommendable to the coating industry. 5 conclusions soybean oil was epoxidized to obtain esbo with 2.56% oxirane oxygen, 67% iodine value conversion and a distinct oxirane oxygen ir peak. esbo blended homogeneously with dmu resin. blends with 10 to 50% esbo in dmu forms a stable continuous phase, and a reaction between esbo and dmu (40 % esbo/dmu) was suspected for the disappearance of the oxirane oxygen peak in the ftir spectra of the blend. increase in the amount of esbo in esbo/dmu blends, decrease formaldehyde emission to safe levels, decrease moisture absorption by the blend film, impact softness/ flexibility as interpreted from the reduction in the melting point of the blend film, but decrease homogeneous disposability/solubility in water. the blend at 40 % esbo/dmu was selected for optimum properties. a comparison of the product with other reported binders shows that it has good potentials to compete in the coating industry as an efficient emulsion paint binder. acknowledgements authors appreciate mr. isaku maizana and the entire laboratory technologists of the department of chemistry modibbo adama university of technology, yola, nigeria. comments from anonymous reviewers are acknowledged. references aigbodion ai, pillai cks. 2001. synthesis and molecular weight characterization of rubber seed oil‐modified alkyd resins. journal of applied polymer science 79(13): 2431-2438. akinterinwa a, osemeahon sa, nkafamiya ii, dass pm. 2015. formulation of emulsion paint from a copolymer composite of dimethylol urea/polystyrene. chemistry and materials research 7(7): 20-26. aoac. 2000. official method of analysis of aoac international. eds: 17 th association of official analytical chemists (aoac), washington dc. barminas jt, osemeahon sa. 2007. development of amino acid resin for paint formulation: copolymerization of methylol urea with polyesther. african journal of biotechnolgy 6(12): 1432–1440. a. akinterinwa et al. novel and benign emulsion paint binder ruhuna journal of science 64 vol 10(1): 51-64, june 2019 cakir m, kartal i, demirer h, samur r. (2012). effect of water absorption on the wear behavior of sol-gel processed epoxy/silica hybrids. scientific research and essays 7(7): 805812. cheng w, liu g, wang x, liu x, jing l. 2015. kinetics of the epoxidation of soybean oil with h2o2 catalyzed by phosphotungsticheteropoly acid in the presence of polyethylene glycol. european journal of lipid science and technology 117(8): 1185-1191. edoga mo. 2006. ethylated urea-ether-modified urea-formaldehyde resins, part i: structural and physicochemical properties. leonardo electronic journal of practices and technologies 9: 121-136. fink jk. 2013. reactive polymers fundamentals and applications: a concise guide to industrial polymers. william andrew.187 pp. habib f, bajpai m. 2011. synthesis and characterization of acrylated epoxidized soybean oil for uv cured coatings. chemistry and chemical technology 5(3): 317-326. idowu-oyewole et, osemeahon, sa, dass pm, akinterinwa a. 2016. synthesis and characterization of an emulsion paint binder from a copolymer composite of trimethylol urea/polyethylene. international journal of materials science and engineering 4(2): 161171. liem h, etchegoin p, whitehead ks, bradley ddc. 2002. raman scattering as a probe of morphology in conjugated polymer thin films. journal of applied physics 92(2): 1154-1161. oladipo go, eromosele ic, folarin om. 2013. formation and characterization of paint based on alkyd resin derivative of ximeniaamericana (wild olive) seed oil. environment and natural resources research 3(3): 52. osemeahon sa, barminas jt. 2007. study of a composite from reactive blending of methylol urea resin with natural rubber. african journal of biotechnology 6(6): 810-817. osemeahon sa, nkafamiya ii,maitera on, akintreinwaa. 2015. synthesis and characterization of emulsion paint binder from a copolymer composite of dimethylol urea/polystyrene. journal of polymer and composites 3(2): 11-21. pantone v, annese c, fusco c, fini p, nacci a, russo a, d’accolti l. 2017. one-pot conversion of epoxidized soybean oil (eso) into soy-based polyurethanes by mocl2o2 catalysis. molecules 22(2): 333. pizzi a, beaujean m, zhao c, properzi m, huang z. 2002. acetal‐induced strength increases and lower resin content of muf and other polycondensation adhesives. journal of applied polymer science 84(13): 2561-2571. saithai p, lecomte j, dubreucq e, tanrattanakul v. 2013. effects of different epoxidation methods of soybean oil on the characteristics of acrylated epoxidized soybean oil-co-poly (methyl methacrylate) copolymer. express polymer letters 7(11): 910-924. salthammer t, mentese s, marutzky r. 2010. formaldehyde in the indoor environment. chemical reviews 110(4): 2536-2572. siggia s. 1963. quantitative organic analysis, eds: 3rd john wiley & sons, new york.103 pp. suurpere a, christjanson p, siimer k. 2006. rotational viscometry for the study of ureaformaldehyde resins. proceedings of the estonian academy of sci engineering 12(2): 134146. vilas jl, laza jm, garay mt, rodriguez m, leon lm. 2001. unsaturated polyester resins cure: kinetic, rheologic, and mechanical‐dynamical analysis. i. cure kinetics by dsc and tsr. journal of applied polymer science 79(3): 447-457. yıldız e, i̇nan t, yıldırım h, kuyulu a., güngör a. 2007. thermal and mechanical properties of poly(arylene ether ketone)s having pendant tertiary butyl groups. journal of polymer research 14: 61-66. zorba t, papadopoulou e, hatjiissaak a, paraskevopoulos k, chrissafis k. 2008. ureaformaldehyde resins characterized by thermal analysis and ftir method. journal of thermal analysis and calorimetry 92(1): 29-33. rjs-vol-1-sept-2006-9.dvi ruhuna journal of science vol. 1, september 2006, pp. 82–95 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. intra-specific morphological plasticity in three puntius species in sri lanka m.p.k.s.k. de silva, n.p.p. liyanage department of zoology, university of ruhuna, matara, sri lanka, kumududs@zoo.ruh.ac.lk s. hettiarachi department of botany, university of ruhuna, matara, sri lanka phenotype of an organism is a result of interaction of genotype and environment. individuals of a same species living in variety of habitats may subject to different environmental conditions. as a result, they may adapt to local conditions in those habitats which include the changes in morphology from the common phenotype. different morphologies in individuals of same genotype is called phenotypic plasticity. this is considered as an important event in evolutionary ecology because these individuals are the first to subject to natural selection. present study focused on phenotypic plasticity in three freshwater fish species namely puntius dorsalis, p. vitatus and p. bimaculatus. fish were sampled from different locations in four different altitude ranges of five major rivers of sri lanka. twenty three length measurements and fifteen meristic characters were recorded from each individual and fourteen physicochemical parameters of each location were also measured. relationship between altitude range and the species morphology was analysed by discriminant analysis and hierachical cluster analysis. correlation of physicochemical parameters with the altitude range was also studied. results showed that individuals living in different altitude ranges in different rivers and also in the same river were morphologically different. results indicated that length characters which determine the shape of the individual mainly contribute to the discrimination of individuals according to their respective altitude range. ratios of maximum body width to standard length, pre ventral length to post ventral length and fork length to standard length in p. dorsalis, p. vitatus and p. bimaculatus respectively were the main discrimination characters that grouped the individuals according to the respective altitudes. variations in length characters were found to be adaptations to their habitat. most physicochemical parameters were significantly correlated (negatively or positively) according to the altitude ranges where three species were collected. phenotypic plasticity in the three puntius species inhabiting different altitudes may have resulted as an adaptation to these variable physicochemical parameters. key words : morphological plasticity, intra-specific, altitude, puntius 1. introduction phenotypic plasticity is the ability of a genotype to respond to alternative environmental conditions to produce array of phenotypes (thompson 1991). in forming a phenotype, the genome and the environment act on the developmental programme (scheiner 1993). genetic variation in a fixed phenotype has been hypothesized to be favored in stable environments (smith 1993) whereas phenotypic plasticity can be an important adaptive strategy for coping with environmental variability (scheiner 82 de silva, liyanage and hettiarachchi: intra-specific morphology ... ruhuna journal of science 1, pp. 82–95, (2006) 83 1993), predation (abrams 2003), for differences in availability of resources such as food (lindsay 1981, magnan 1988) and habitat choice of individuals (smith and sikulason 1996). aquatic environments exhibit great spatial and temporal variability in both abiotic and biotic habitat parameters (lowe-mcconnell 1987, goulding et al. 1988) and intraspecific diversification is well documented in fishes (robinson & wilson 1994, smith & skúlason 1996, taylor 1999, jonsson & jonsson 2001). adaptive phenotypic plasticity in fish morphology has been demonstrated in crucian carp, carassius carassius, in response to the presence or absence of a predator (brönmark & miner 1992), and in pumpkinseed sunfish, lepomis gibbosus (robinson & wilson 1996) and stickleback, gasterosteus spp. (day et al. 1994), in response to differences between benthic and pelagic habitats. in chum salmon (oncorhynchus keta), variation in morphometric and meristic characters according to different temperature conditions in water has been recorded (beacham 1990). body shape in fishes has been demonstrated to be influenced by type of food or feeding mode (day et al.1994, robinson & wilson 1996). barbus neumayeri living in hypoxic habitats (swamps) has shown larger gill size compared to those living in streams which have better oxygen supply (chapman et al. 1999, chapman and liem 1995). body shape of the hatchery reared and wild atlantic salmon also tend to be heavily influenced by rearing conditions (von cramon et al. 2005). divergence of structure, behaviour or habitat could support more living beings to live in the same area. individual phenotypic differences are important in understanding the evolutionary ecology of a population or a species because variation among individuals is the raw material on which natural selections first acts on. phenotypic variation in a species could lead to genotypic variation and this could even lead to origin of a species in long term. puntius is a genus of freshwater fish belonging to family cyprinidae. there are 16 puntius species in sri lanka. they inhabit in a variety of aquatic environments; i.e., rivers, streams, reservoirs, or tanks; waters of stagnant, slow flowing or with strong currents; deep or shallow; relatively higher to lower altitudes and clear or turbid waters (schut et al.1984). three commonly found indigenous puntius species, namely p. bimaculatus, p. dorsalis and p. vitatus were considered in the present study. according to schut et. al. 1984, p. bimaculatus adults were mainly found in hill country in small and large streams while juveniles were found in marshes. in contrast, p. vitatus found in marshes and paddy lands with turbid water in lowlands. p. dorsalis has the widest range of distribution from the rocky hill streams to main rivers below flood level and also found in tanks and reservoirs. these three species show three types of distribution (hill country, lowlands and wide range), where environmental factors (soils, water bodies and altitude ranges) are different. body shape (morphology) of these fish could be adapted to suit these differing environmental conditions. de silva, liyanage and hettiarachchi: intra-specific morphology ... 84 ruhuna journal of science 1, pp. 82–95, (2006) table 1 altitude ranges where sampling done in five rivers. x x x x x x x x x river altitude mahaweli kalu kelani nilwala gin 0–150 ◦ ◦�△∗ ◦�△∗ ◦�△∗ ◦� 150-300 ◦△ ◦ ◦ � 300-600 ◦△ ◦ △ △ 600-1200 △ ◦ △ 1200-1800 △∗ ◦ p. dorsalis � p. vitatus △ p. bimaculatus ∗ not used in the analysis as the samples were not in suitable condition to take measurements. objective of the present study was to determine whether variability of macrohabitats, mainly altitude and the type of water body, has any effect on morphological plasticity of these three puntius spp. in five major rivers namely mahaweli, kelani, kalu, nilwala and gin. 2. materials and methods five major rivers mahaweli, kalu, kelani, gin and nilwala were selected for the study. sampling was carried out in five different altitude ranges of each river. fish samples were collected using hand nets, scoop nets, cast nets and gape nets from streams, streamlets, rivulets and reservoirs belonging to five river basins. figure 1 shows the sampling sites and table 1 shows the altitude range of which the species were caught in each location. total number of fish collected was 31, 53 and 43 for p. vitatus, p. bimaculatus and p. dorsalis, respectively. in each altitude range more than one location was sampled. other fish species caught were released back to the water and three puntius spp. used for the study were preserved in 70% alcohol and brought to the laboratory. species of the fish were identified according to their external morphology and 23 length measurements (measurables) and 15 meristic characters (countables) were recorded from each fish. list 1 shows the characters recorded. fourteen physicochemical parameters of water; temperature, ph, alkalinity, salinity, conductivity, dissolved oxygen, suspended organic matter, suspended inorganic matter, organic suspended solids, chlorophyl content, chemical oxygen demand, biological oxygen demand, phosphate levels and nitrate levels were measured at the sampled sites using standard methods. de silva, liyanage and hettiarachchi: intra-specific morphology ... ruhuna journal of science 1, pp. 82–95, (2006) 85 figure 1 length measurements recorded note. (abbreviations are defined in list 1) list 1 morphometric and meristic characters used (a) morphometric characters tl total length fl fork length mbw maximum body width hl head length pol post orbital length dfl dorsal fin length sl standard length ed eye diameter pdl pre dorsal length podl post dorsal length afl anal fin length pal pre anal length poal post anal length pvl pre ventral length povl post ventral length ppl pre pectoral length popl post pectoral length cfl caudal fin length cspr caudal spread lcpd length of caudal peduncle hcpd least depth of caudal peduncle iow inter orbital width ind inter nostril distance lls no. of lateral line scales de silva, liyanage and hettiarachchi: intra-specific morphology ... 86 ruhuna journal of science 1, pp. 82–95, (2006) (b) meristic characters tr no. of transverse scales cped scales around caudal peduncle dfr dorsal fin rays dfs dorsal fin spines vfr ventral fin rays vfs ventral fin spines pfr pectoral fin rays pfs pectoral fin spines afr anal fin rays afs anal fin spines prds pre dorsal fin scales cfr caudal fin rays psds post dorsal fin scales dfsc dorsal fin scales figure 2 rivers and sampling sites note. ∗ locations in five rivers are shown in five different fonts as the fish were in different sizes, length measurements were converted to ratios to standerdise for the size of the fish. analysis was carried out for each species separately. size adjusted length ratios and meristic characters were subjected to discriminant function analysis using step wise insertion of variables to find out the canonical functions, characters that contribute for the highest percentage variance and principal components that contribute to discriminate the species according to de silva, liyanage and hettiarachchi: intra-specific morphology ... ruhuna journal of science 1, pp. 82–95, (2006) 87 table 2 no. of functions, percentage variance and the characters that contribute for the discrimination of the individuals to their respective altitude range p. dorsalis p. vitatus p. bimaculatus function % character function % character function % character no. variance no. variance no. variance 1 74.2 mbw.tl 1 100 pvl.povl 1 83.8 fl.sl 2 22.5 poal.sl 2 16.2 pol.sl 3 3.4 povl.sl the altitude range that they were found. mean plots were obtained for the major characters that contribute for the highest percentage variance for the altitudinal ranges. principal components obtained were used to cluster the fish according to the altitudes by hierachical cluster analysis. spss 10 statistical package was used for the analysis. physicochemical parameters were log transformed. relationship between physicochemical parameters and altitude ranges were analysed by correlation analysis to determine the most variable parameters in the four altitude ranges in five rivers. 3. results results of the canonical discriminant analysis show that from the 39 morphological descriptors (24 length ratios and 15 meristic), length ratios mainly contribute to the discrimination of individuals (table 2). in p. vitatus ratio of pre ventral length and post ventral length contribute totally for discrimination (100%). fork length mainly contributes to discriminate p. bimaculatus (83.8%). as the percentage variance for post anal length and post ventral length is low, maximum body depth is the major discriminating factor for p. dorsalis. mean plots obtained by one way anova for these characters show that body width (mbw.sl) is significantly higher in the p. dorsalis individuals in lowest altitude of 0-150 m (figure 3a). in p. vitatus mean pvl.povl is highest in 0-150 m altitude range shows that ventral fin is positioned more posteriorly (figure 3b). high mean values obtained for the fork length in p. bimaculatus collected from high altitudes (600-1200 m) show that they have longer caudal fins. for the individuals living in lower altitudes high mean value was obtained for the post orbital length indicate that they have longer heads (figure 3 (c) and (d)). six, five and seven principal components were obtained for p. dorsalis, p. vitatus and p. bimaculatus, respectively. both meristic and length measurements were contributed to group the individuals of p. bimaculatus, p. dorsalis and p. vitatus according to the altitudinal ranges. except for three cases for each species, in p. vitatus and p. bimaculatus all the individuals belonging to the same altitude in different rivers were grouped together. grouping of individuals in p. dorsalis did not strictly follow that pattern. figures 4, 5, and 6 show the clustering of individuals of three species according to the altitude ranges. p. vitatus was found only from samples collected at lower altitudes (0-150 m and 150-300 m) and they form two distinct clusters at 20% level in the dendrogram (1a and 1b) representing two altitudes (figure 4). individuals of de silva, liyanage and hettiarachchi: intra-specific morphology ... 88 ruhuna journal of science 1, pp. 82–95, (2006) (a)p. dorsalis (b)p.vitatus (c)p. bimaculatus (d)p. bimaculatus figure 3 mean plots showing the variation of the main discriminant characters according to the altitude range for 3a. p. dorsalis, 3b. p.vitatus and 3c. p. bimaculatus table 3 key morphometric characters extracted for three puntius species principal components p. dorsalis p. vitatus p. bimaculatus mbw.sl ∗ pvi.povl ∗ fl.sl ∗ poal.sl ∗ povl.sl pol.sl ∗ povl.sl ∗ ppl.popl podl. sl tr cfl.cspr lls cfr dfr afr prds pfr cfr ∗ principal components that give highest percentage variance for the canonical functions p. bimaculatus of nilwala river collected from 300-600 m altitude range separated out first at 80% level in the dendrogram (1a1 and 1a2 of figure 5). from the rest, individuals of altitude 150-300 m clustered separately from individuals of altitude range 600-1200 m (2a1 and 2a2).compared with the other two species p. dorsalis de silva, liyanage and hettiarachchi: intra-specific morphology ... ruhuna journal of science 1, pp. 82–95, (2006) 89 figure 4 clustering of p. vitatus according to the altitudes note. * misclassified cases does not show distinct separation in morphology according to the altitude ranges. individuals belonging to 150-300 m range in nilwala river (1a) show different morphology from the rest of p. dorsalis individuals and individuals belonging to other three altitude ranges do not show much difference in their morphology (figure 6). correlation analysis of altitude range and physicochemical parameters of the localities where three species were collected show that most characters are negatively correlated with the altitude. results of the same analysis for p. bimaculatus show some positively correlated parameters (table 4). chlorophyll content in the water shows no significant correlation with the altitude. de silva, liyanage and hettiarachchi: intra-specific morphology ... 90 ruhuna journal of science 1, pp. 82–95, (2006) figure 5 clustering of p. bimaculatus according to the altitudes note. * misclassified cases 4. discussion according to principal component analysis positioning of fins in the body is important in discriminating the individuals of p. vitatus in relation to the two altitudinal ranges in which they were found. position of the ventral fin, which contributes to 100% variance for canonical functions and the pectoral fin, is more anteriorly placed in individuals collected from the lowest altitude range (0-150 m). p. vitatus is mainly de silva, liyanage and hettiarachchi: intra-specific morphology ... ruhuna journal of science 1, pp. 82–95, (2006) 91 figure 6 clustering of p. dorsalis according to the altitudes. note. * misclassified cases found in marshes and paddy fields. in paddy fields, large numbers of this species gather under spillways and other places with flowing water (schut et al.1984). more anteriorly placed lateral fins are an adaptation for living in streams (brinsmead and fox, 2003). individuals collected from altitude range of 150-300 m have larger caudal spread and this is also an adaptation to their habitats such as spillways with arduous environmental conditions. de silva, liyanage and hettiarachchi: intra-specific morphology ... 92 ruhuna journal of science 1, pp. 82–95, (2006) table 4 significant correlations obtained for altitude ranges and physicochemical parameters of the locations where three species collected p. dorsalis p. vitatus p. bimaculatus temperature conductivity temperature ph total suspended solids dissolved oxygen conductivity organic suspended solids biological oxygen demand total suspended solids inorganic suspended solids + chemical oxygen demand organic suspended solids biological oxygen demand + salinity inorganic suspended solids salinity + phosphate biological oxygen demand nitrate chemical oxygen demand + dissolved oxygen salinity alkalinity nitrate phosphate negative correlation, + positive correlation p. bimaculatus is found in the rocky hill streams (schut et al. 1984). this was the only puntius species found in highest altitude studied (table 1). morphological characters that contributed most to the canonical function analysis in p. bimaculatus are related to caudal fin length and head length. comparison of mean plots of these characters for lower altitudes and higher altitudes showed that individuals in high altitudes have long bodies with longer caudal fins and shorter head lengths. habitats of p. bimaculatus individuals found in higher altitudes are in steep hills with fast flowing waters. as such water in those habitats gets well aerated. longer caudal fins give them more fusiform bodies so that they can withstand fast flowing water. longer post orbital lengths in individuals in lower altitudes is an adaptation to provide more space to increase the size of the gill (lindsay 1981) possibly through increasing the number of gill filaments. garra ceylonensis which is highly adapted to fast flowing water, was the only other cyprinid species found in the highest altitudes, in addition to p. bimacualtus. p. dorsalis was found in most of the locations sampled and had the widest distribution. individuals living in the lowest altitudes have deeper bodies and their lateral fins are placed more anteriorly than those in higher altitudes. webb and wehis (1986) have shown that these characters are the optimal design for maneuvering type of locomotion. p. dorsalis is commonly found in reservoirs and tanks too. deep body is also an adaptation to live in littoral habitats (robinson et al. 1996). principal components obtained for each puntius species were able to classify the individuals according to the altitude ranges they were collected. individuals sampled from the same altitude range of different rivers clustered together indicating that altitude has considerable effect on the morphology of the species. except for a few individuals, phenotypically similar individuals of different rivers belonging to a same altitude could be grouped together. it would be interesting to study this relationship with different fish genera to generalise this outcome. de silva, liyanage and hettiarachchi: intra-specific morphology ... ruhuna journal of science 1, pp. 82–95, (2006) 93 individuals that develop and mature in common environmental conditions may share a similar phenotype. when the movement between riverine populations is limited they develop population specific phenotypes (jerry and cairns 1998). present demonstration of significant similarities found in most physicochemical parameters of the same altitude range in different rivers may have created common environmental influences. this may have caused the individuals to be shaped into similar phenotypes in the same species in those environments. barlow (1961) has shown that latitudinal changes in temperature have affected the expression of morphological characters. similarly the differences found in some physicochemical parameters among altitude ranges may have resulted in differences in phenotype of the same species shown in present study. more experimental work are required to further describe the role that ecological conditions play in maintaining morphological diversity in these groups. p. dorsalis being a macrophagous herbivore and the other two species being algae and diatom feeders their distribution may depend on the availability of their food source. however chlorophyll content of water shows no significant correlation among altitudes. therefore any correlation between chlorophyll content and morphology of puntius cannot be demonstrated. many organisms can modulate their morphology in response to environmental cues. such plasticity is thought to be an important adaptive strategy for populations experiencing variable environmental conditions (scheiner 1993) and it is likely that phenotypic plasticity plays an important role in diversification (west-eberhard 1989). morphology in teleost fish has been shown to be particularly labile in response to multifarious habitat effects (kinsey et al. 1994, corti et al. 1996). present study also shows that the morphology of a species in the same river but different localities could differ. this may lead to form in different populations as the movement of individuals among the most localities is difficult. these populations subject to local selection pressures leading ultimately to increased fitness termed local adaptations (carvalho 1993).this could even result genetic divergence of populations. future studies on genetic component of these phenotypes would result a better picture of phenotypic plasticity of puntius species. references abrams pa. 2003. can adaptive evolution or behaviour lead to diversification of traits determining a trade-off between foraging gain and predation risk? evolutionary ecology research 5:653-670 beacham td. 1990. a genetic analysis of meristic and morphometric variation in chum salmon (oncorhynchus keta) at three different temperatures. canadian journal of zoology 68: 225-229 barlow gw. 1961. causes and significance of morphological variation in fishes. systematic zoology 10: 105-117 de silva, liyanage and hettiarachchi: intra-specific morphology ... 94 ruhuna journal of science 1, pp. 82–95, (2006) brinsmead j and fox mg. 2003. morphological variation between lakeand streamdwelling rock bass and pumpkinseed populations. journal of fish biology 61:16191638 brönmark c. and miner jg. 1992. predator-induced phenotypical change in body morphology in crucian carp. science 258: 1348-1350 carvalho gr. 1993. evolutionary aspects of fish distribution: genetic variability and adaptation. journal of fish biology 43 (suppl a): 53-73 chapman lj., chapman ca., brazeau da., mclaughlin b. and jordan, m. 1999. papyrus swamps, hypoxia and faunal diversification: variation among populations of barbus neumayeri. journal of fish biology 54:310-327 chapman lj. and liem k.f. 1995. papyrus swamps and the respiratory ecology of barbus neumayeri. environment and biology of fishes 44:183-197 corti m., loy a. and cataudella s. 1996. form changes in sea bass, dicentrachus labrax (moronidae: teleostei), after acclimatation to freshwater: an analysis using shape coordinates. enviornmental biology of fishes 47: 165-175 day t. and mcphail jd. 1996. the effect of behavioural and morphological plasticity on foraging efficiency in the three spine stickleback (gasterosteus sp.). oecologia 108:380-388 day t. pritchard j. and schluter d. 1994. a comparison of two sticklebacks. evolution 48:1723-1734 goulding m., carvalho ml. and ferreira eg. 1988. rio negro: rich life in poor water. hague, the netherlands: spb academic publishing. jerry dr. and cairns sc. 1998. morphological variability in the catadromous australian bass, macquaria novemaculeata (perciformes: percichthyidae), from seven geographically distinct riverine drainages. journal of fish biology 52: 829-843 jonsson b. and jonsson n. 2001. polymorphism and speciation in arctic charr. journal of fish biology 58:605-638 kinsey st., orsoy t., bert t.m. and mahmoudi b. 1994. population structure of the spanish sardine sardinella aurita: natural morphological variation in a genetically homogenous population. marine biology 118: 309-317 lindsay cc. 1981. stocks are chameleons: plasticity in gill rakers of corrigonid fishes. canadian journal of fisheries and aquatic sciences 38:1497-1506 lowe-mcconnell rh. 1987. ecological studies in tropical fish communities. cambridge: cambridge university press magnan p. 1988. interactions between brook charr, salvelinus fontinalis, and nonsalmonid species: ecological shift, morphological shift, and their impact on zooplankton communities. canadian journal of fisheries and aquatic science 45:9991009 robinson bw. and wilson ds. 1994. character release and displacement in fishes: a neglected literature. american naturalist 144:596-627 de silva, liyanage and hettiarachchi: intra-specific morphology ... ruhuna journal of science 1, pp. 82–95, (2006) 95 robinson bw. and wilson ds. 1996. genetic variation and phenotypic plasticity in a trophically polymorphic population of pumpkinseed sunfish (lepomis gibbus). evolutionary ecology 10:631-652 robinson bw, wilson ds and shea go. 1996. trade-offs of ecological specialization: an intraspecific comparison of pumpkinseed sunfish phenotypes. ecology 77:170178 scheiner sm. 1993. genetics and evolution of phenotypic plasticity. annual review of ecology and systematics 24:35-68 schut j,, de silva s. and kortmulder k. 1984. habitat associations and competition of eight barbus (=puntius) species (pisces, cyprinidae) indigenous to sri lanka. 34(2):159-181 smith tb. 1993. disruptive selection and the genetic basis of bill size polymorphism in the african finch pyrenestes. nature 363:618-620 smith tb. and sklason s. 1996. evolutionary significance of resource polymorphism in fishes, amphibians and birds. annual review of ecology and systematics 27:111-133 taylor eb. 1999. species pairs of north temperature freshwater fishes: evolution, taxonomy and conservation. reviews in fish biology and fisheries 9: 299-324. thompson jd. 1991. phenotypic plasticity as a component of evolutionary change. trends in ecology and evolution 6:246-249 von cramon t., ling en., cotters d. and wilkins np. 2005. determination of body shape variation in irish hatchery-reared and wild atlantic salmon. journal of fish biology 66: 1471-1482 webb pw. and weihs d. 1986. functional locomotor morphology of early life history stages of fishes. transactions of the american fisheries society 115:115-127 west-eberhard mj. 1989. phenotypic plasticity and the origins of diversity. annual review of ecology and systematics 20:249-278 acknowledgments financial assistance provided by national science foundation of sri lanka under the research grant rg/bt/2003/02 is highly appreciated. sv-lncs ruhuna journal of science vol. 5: 1-6, 2014 http://rjs.ruh.ac.lk/ issn: 1800-279x 1  faculty of science, university of ruhuna a new approach for thoracoscopic posterior mediastinal procedures k. b. galketiya 1 and m. g. v. pinto 2 1 department of surgery, faculty of medicine, university of peradeniya, peradeniya, sri lanka 2 teaching hospital, peradeniya, sri lanka correpondence: 1 kbgalketiya@yahoo.com received: 23 july 2014, revised version accepted: 25 november 2014 abstract.the morbidity of thoracotomy is reduced by thoracoscopy. the space for dissection is obtained by collapsing the lung. during posterior mediastinal procedures the patient is positioned prone. this allows the collapsed lung to fall away from the field of dissection. in face of conversion to lateral thoracotomy re-positioning will take time, which may be dangerous like in a severe bleeding. an alternative is to place the patient semi prone and get in to a near prone position by tilting the table. quick return to lateral position can be achieved by tilting the table in reverse direction. upper thoracic sympathectomy and mobilization of thoracic oesophagus were done in the adopted prone position. bilateral splanchnicectomy was performed in the prone position. in both situations, there was adequate space for instrumentation and dissection. retractors were not required. all the procedures were completed safely with minimal blood loss and an acceptable time. there were no conversions. therefore, the advantages of prone position to provide space for dissection in posterior mediastinal thoracoscopic surgeries were obtained by the adopted semi prone position. keywords. mediastinal, positioning, thoracoscopy 1 introduction surgical procedures performed by thoracotomy causes a significant morbidity. the post-operative pain may cause problems for breathing. therefore the need for ventilatory support and the risk of respiratory tract infections is higher. the hospital stay will be prolonged with delayed return for work (findik et al. 2008). galketiya and pinto positions for thoracoscopic procedures ruhuna journal of science (december 2014) 2 thoracoscopy (figure 1) takes away the morbidity related to the open access incision (jones and hooper 1995; cuschieri and steel 1995; kwong et al. 2005; tomaszewski et al. 2007; shibasaki et al. 2012). it is necessary to create space for surgery at the beginning of the procedure. space is required to identify the anatomy and for instrumentation for dissection. in thoracoscopy, space is obtained by collapsing the lung, which has to be retracted away from the field of dissection. in minimal access surgery gravity is used to get organs fall away from the site. this reduces the need of retractors. use of retractors will need additional ports, assistants and may hinder dissecting instruments. by positioning the patient, collapsed lung can be allowed to retract away from the field of dissection. for surgeries of the posterior mediastinum, patient is placed prone (dapri et al. 2006; tomaszewski et al. 2007; shibasaki et al. 2012), and this allows the lung to fall anteriorly (figure 2). fig 1. images of open thoracotomy (left), and thoracoscopic surgery through mini-incisions (center & right). however prone positioning has problems. positioning the patient in prone takes time and may hinder abdominal breathing if not placed properly with many complications documented in literature (edgcombe et al. 2008). in case of emergency like displaced endotracheal tube or for conversion to thoracotomy, the position needs changing. time taken to change position may be dangerous, for instance in face of a catastrophic bleeding. therefore, we evaluated semi prone position with regards to space for dissection. 2 methods the surgeries included in the study were bilateral splanchnicectomy, thoracic sympathectomy and oesophagectomy. splanchnicectomy was performed for disabling pain of chronic pancreatitis. thoracic sympathectomy was galketiya and pinto positions for thoracoscopic procedures ruhuna journal of science (december 2014) 3 performed for digital ischaemia or hyperhydrosis. oesophagectomy was for carcinoma of oesophagus as the thoracic part of a three stage procedure. all were performed under general anaesthesia and endotracheal intubation. the patients undergoing esophagectomy and thoracic sympathectomy were placed in semiprone position. the patients were positioned in a mid way of prone and lateral (figure 3). during surgery, a near prone position was adopted by tilting the table. getting back to lateral position for conversion to thoracotomy can be simply obtained by tilting the table, while changing to supine position too can be done without delay (figure 3). fig 2. collapsed lung falling away from posterior mediastinum by use of prone position. however, all thoracoscopic splanchnicectomies were performed in the standard full prone position (figure 3); the reason to use the standard prone position is because splanchnic nerve ablation has to be performed on both sides. in a bilateral procedure re-positioning, cleaning and draping is required if semi-prone position is used, as explained next. this will increase the operating time and has a risk of dislodgement of the endotracheal tube. the ipsilateral lung was collapsed to obtain space for dissection. in some patients, the lung collapse was obtained with single lung ventilation. in others, both lungs were ventilated while lung collapse was by a capnothorax of 68mmhg. the respiratory and cardiovascular parameters were monitored. the space for dissection was evaluated. adequate space should provide a clear display of anatomy. there should be space for instrumentation and dissection. the need for retractors was also noted. adequate space is essential for safe completion of the procedure. as measures of successful completion, time taken, blood loss, complications and need for conversion were recorded. galketiya and pinto positions for thoracoscopic procedures ruhuna journal of science (december 2014) 4 fig 3. images of patient placed semi-prone (left), near prone (center) and full prone positions (right) for thoracotomy. 3 results and discussion a total number of 30 procedures were performed. the position adapted is given with time duration, blood loss, use of retractors, and adequacy of space for visualization and dissection and conversions (table 1). table 1. summary of the procedures and related information. procedure position number of patients average duration (min) average blood loss (ml) use of retractors adequacy of space conversions oesophagectomy semi prone 15 150 100 nil good nil thoracic sympathectomy semi prone 8 30 nm nil good nil bilateral splanchnicectomy prone 7 60 nm nil good nil nm = not measurable space for dissection is essential to perform any surgery. in surgeries in the posterior mediastinum, the space was obtained by collapsing the lung and positioning the patient prone. prone position allows the collapsed lung to be retracted away from the field of dissection with the help of gravity (dapri et. al. 2006, tomaszewski et al. 2007, shibasaki et al. 2012). galketiya and pinto positions for thoracoscopic procedures ruhuna journal of science (december 2014) 5 fig 4. thoracoscopic oesophagectomy (left), thoracoscopic sympathectomy (center) and thoracoscopic splanchnicectomy (right) demonstrate the adequacy of space. positioning the patient in prone takes more time and may hinder abdominal breathing, if not placed properly with many other complications (edgcombe et al. 2008). in case of emergency like displaced endotracheal tube or for conversion to thoracotomy, the position needs changing. time taken to change position may be dangerous, for instance in face of a catastrophic bleeding. therefore, we evaluated a new position, an adopted prone position during oesophagectomy and thoracic sympathectomy. the patient was kept in between lateral and prone position. with a table tilt a near prone position was achieved during the procedure. in face of conversion to thoracotomy lateral position can be adopted simply by tilting the table. thoracoscopic splanchnicectomies were performed in the prone position as it’s a bilateral procedure. even though a different procedure, a comparative visual assessment of the space provided was possible. furthermore, in sympathectomy and splanchnicectomy, the anatomical structure dealt is the same, the thoracic sympathetic chain. both prone and semi prone positions provided adequate space for identification of anatomy, instrumentation and dissection. retractors were not used as there was adequate lung retraction by gravity facilitated by the positions used. there were no conversions with minimal blood loss and all procedures were completed safely, for which an adequate space is essential. galketiya and pinto positions for thoracoscopic procedures ruhuna journal of science (december 2014) 6 4 conclusion the advantages of prone position to provide space for dissection in posterior mediastinal thoracoscopic surgeries were provided by the adopted semi prone position. placing the patient semi prone is easier than placing them prone. in case of emergency conversion to lateral thoracotomy, changing the position will be easy with the semi prone position. references cuschieri a, steele rjc. 1995. surgical craft and technology in essential surgical practice, 4 th edition, arnold. p 45. dapri g, himpens j, cadière gb. 2006. robot-assisted thoracoscopic esophagectomy with the patient in the prone position. journal of laparoendoscopic and advanced surgical techniques 16 (3): 278-285. edgcombe h, carter k, yarrow s. 2008. anaesthesia in the prone position british journal of anaesthesia 100 (2): 165-183 doi:10.1093/bja/aem380. findik g, gezer s, sirmali m, turut h, aydogdu k, tastepe i, karaoglanoglu n, kaya s. 2008. thoracotomies in children. pediatric surgery international 24 (6): 721-725. jones mt, hooper tl. 1995. video assisted thoracic surgery. in: introduction to minimal access surgery, bmj publishing group; 62-64. kwong kf, cooper lb, bennett la, burrows w, gamliel z, krasna mj. 2005. clinical experience in 397 consecutive thoracoscopic sympathectomies. annals of thoracic surgery 80 (3): 1063-1066. shibasaki h, kinoshita t, ogata a, miyazaki m. 2012. thoracoscopic esophagectomy in the prone position. hepato-gastroenterology 59 (118): 1840-1843. tomaszewski s, szyca r, jasiński a, leksowski k. 2007. bilateral posterior thoracoscopic splanchnicectomy in a face-down position in the management of chronic pancreatic pain. pol merkur lekarski 22 (131): 399-401. http://www.ncbi.nlm.nih.gov/pubmed?term=findik%20g%5bauthor%5d&cauthor=true&cauthor_uid=18414878 http://www.ncbi.nlm.nih.gov/pubmed?term=gezer%20s%5bauthor%5d&cauthor=true&cauthor_uid=18414878 http://www.ncbi.nlm.nih.gov/pubmed?term=sirmali%20m%5bauthor%5d&cauthor=true&cauthor_uid=18414878 http://www.ncbi.nlm.nih.gov/pubmed?term=turut%20h%5bauthor%5d&cauthor=true&cauthor_uid=18414878 http://www.ncbi.nlm.nih.gov/pubmed?term=aydogdu%20k%5bauthor%5d&cauthor=true&cauthor_uid=18414878 http://www.ncbi.nlm.nih.gov/pubmed?term=tastepe%20i%5bauthor%5d&cauthor=true&cauthor_uid=18414878 http://www.ncbi.nlm.nih.gov/pubmed?term=karaoglanoglu%20n%5bauthor%5d&cauthor=true&cauthor_uid=18414878 http://www.ncbi.nlm.nih.gov/pubmed?term=kaya%20s%5bauthor%5d&cauthor=true&cauthor_uid=18414878 http://www.ncbi.nlm.nih.gov/pubmed/18414878 http://www.ncbi.nlm.nih.gov/pubmed?term=kwong%20kf%5bauthor%5d&cauthor=true&cauthor_uid=16122487 http://www.ncbi.nlm.nih.gov/pubmed?term=cooper%20lb%5bauthor%5d&cauthor=true&cauthor_uid=16122487 http://www.ncbi.nlm.nih.gov/pubmed?term=bennett%20la%5bauthor%5d&cauthor=true&cauthor_uid=16122487 http://www.ncbi.nlm.nih.gov/pubmed?term=burrows%20w%5bauthor%5d&cauthor=true&cauthor_uid=16122487 http://www.ncbi.nlm.nih.gov/pubmed?term=gamliel%20z%5bauthor%5d&cauthor=true&cauthor_uid=16122487 http://www.ncbi.nlm.nih.gov/pubmed?term=krasna%20mj%5bauthor%5d&cauthor=true&cauthor_uid=16122487 http://www.ncbi.nlm.nih.gov/pubmed/16122487 http://www.ncbi.nlm.nih.gov/pubmed/16122487 http://www.ncbi.nlm.nih.gov/pubmed?term=tomaszewski%20s%5bauthor%5d&cauthor=true&cauthor_uid=17679379 http://www.ncbi.nlm.nih.gov/pubmed?term=szyca%20r%5bauthor%5d&cauthor=true&cauthor_uid=17679379 http://www.ncbi.nlm.nih.gov/pubmed?term=jasi%c5%84ski%20a%5bauthor%5d&cauthor=true&cauthor_uid=17679379 http://www.ncbi.nlm.nih.gov/pubmed?term=leksowski%20k%5bauthor%5d&cauthor=true&cauthor_uid=17679379 http://www.ncbi.nlm.nih.gov/pubmed/17679379 microsoft word rjs-vol-ii-p.samarasekara-final1.0.doc.doc © 2007 faculty of science university of ruhuna ruhuna journal of science vol. 2, september 2007, pp. 1-9 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x 1 abstract. the energy of ultra-thin sc(001) ferromagnetic film with three layers will be investigated using classical heisenberg hamiltonian. energy curves show several minimums indicating that the film can be easily oriented in these directions under the influence of certain values of demagnetization factor. when the demagnetization factor is given by ωµ 0 dn =8, the angle corresponding to first minimum is 0.6 radians for sc(001) ferromagnetic lattice. under the influence of demagnetization factor of ωµ 0 dn = 7.5, energy minimum can be observed at 0.6 radians for bcc(001) lattice. the energy curve of bcc(001) ferromagnetic lattice is smoother compared with that of sc(001) lattice. keywords: materials, thin films, heisenberg hamiltonian, demagnetization factor 1. introduction: the properties of ferromagnetic ultra-thin films have been investigated using classical model of heisenberg hamiltonian with limited number of terms (usadel and hutch 2002). because of the difficulties of understanding the behavior of exchange anisotropy and its applications in magnetic sensors and media technology, exchange anisotropy has been extensively investigated in recent past (david et al. 2004). ferromagnetic films are thoroughly studied nowadays, due to their potential applications in magnetic memory devices and microwave devices. bloch spin wave theory has been used to investigate magnetic properties of ferromagnetic thin films (martin and robert 1951). although the magnetization of some thin films is oriented in the plane of the film due to dipole interaction, the out of plane orientation is preferred at the surface due to the broken symmetry of uniaxial anisotropy energy. previously two dimensional heisenberg model has been used to explain the magnetic anisotropy in the presence of dipole interaction (dantinzger et al. 2002). ising model has been used to study magnetic properties of ferromagnetic thin films with alternating super layers (bentaleb et al. 2002). effect of demagnetization factor on total energy of ultra-thin ferromagnetic films with three layers p. samarasekara department of physics, university of ruhuna, matara, sri lanka correpondence: pubudus@phy.ruh.ac.lk 2 samarasekara: effect of demagnetization… r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 1 9 ( 2 0 0 7 ) for the very first time, the variation of energy of ferromagnetic ultra-thin films with demagnetization factor will be described in this report. the energy of non-oriented ultra-thin ferromagnetic films with two and three layers has been calculated using heisenberg hamiltonian with second order perturbation, under the effect of limited number of energy parameters (samarasekara 2006). the properties of perfectly oriented thick ferromagnetic films have been investigated by classical heisenberg model (samarasekara 2006). the variation of energy with angle and number of layers has been studied for thick films up to 10000 layers. the total magnetic energy has been calculated using two different methods depending on discrete and continuous variation of thickness. for bcc(001) lattice, the easy and hard directions calculated using both methods were exactly same. 2. model and discussion: the classical model of heisenberg hamiltonian is given as following (samarasekara 2006), ∑ ∑ ∑∑ −−−+= ≠nm m m z m z m nm mn nmnmnm mn nm nm sdsdr srrs r ss ss j h mm , 4)4(2)2( 53 )()() ).)(.(3. ( 2 . 2 λλ ω rrrrrrrr ∑ ∑−−− nm m msmnd sinkssnh , 0 2)]./([ θµ rrr for the heisenberg hamiltonian given in above equation, total energy can be obtained as following (samarasekara 2006). e(θ)=e0+ εα rr . + εε rr .. 2 1 c =e0 αα rr .. 2 1 +− c the matrix elements of above matrix c are given by 0 2 2cos 4 3 ) 4 ( µ θ ωω d nmnmnmmn n jzc +φ−φ−−= −−− ∑ = −− −−+φ−+ n mmmmn djz 1 )2(22 )cos(sin2)]2cos 4 3 4 ([{ λ λλ θθθ ωω δ }2sin4 4 cossin)sin3(coscos4 0 )4(222 θ µ θθθθθ s d outinm k n hhd +−++−+ θθεα 2sin)()( b rr = are the terms of matrices with θ ω θ λλλλ 2)4( )2( 1 cos2 4 3 )( ddb n m m ++φ−= ∑ = − (1) here (samarasekara 2006) samarasekara: effect of demagnetization… 3 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 1 9 ( 2 0 0 7 ) ])1(2[ 2 100 znnz j e −+−= 10 )1(2{ φ−+φ+ nn )2cos8 3 8 }( θ ωω + )2sincossincos(cos 0 )4(4)2(2 θ µ θθθθ s d outinmm k n hhddn +−+++− e0 is the energy of the oriented thin ferromagnetic film. here j, nmz − , ω, nm−φ , θ, ,,,,,, )4()2( sdoutinmm knhhdd m, n and n are spin exchange interaction, number of nearest spin neighbors, strength of long range dipole interaction, constants for partial summation of dipole interaction, azimuthal angle of spin, second and fourth order anisotropy constants, in plane and out of plane applied magnetic fields, demagnetization factor, stress induced anisotropy constant, spin plane indices and total number of layers in film, respectively. when the stress applies normal to the film plane, the angle between mth spin and the stress is θm. matrix elements for a film with three layers (n=3) can be given as following (samarasekara 2006), 0 1132232112 2 )2cos31( 4 µ θ ω dnjzcccc +−φ+−==== 0 223113 2 )2cos31( 4 µ θ ω dnjzcc +−φ+−== )2( 0 21213311 )2cos2( 2 )2cos31)(( 4 )( m d d n zzjcc θ µ θ ω +−+φ+φ−+== θθθθθθ 2sin4cossin)sin3(coscos4 )4(222 soutinm khhd +++−+ )2( 0 1122 )2cos2( 2 )2cos31( 2 2 m d d n jzc θ µ θ ω +−+φ−= θθθθθθ 2sin4cossin)sin3(coscos4 )4(222 soutinm khhd +++−+ if the second or fourth order anisotropy constants are invariants inside an ultra thin film, then d1(2)=d2(2)=d3(2) and d1(4)=d2(4)=d3(4). under some special conditions (samarasekara 2006), c+ is the standard inverse of a matrix, given by matrix element c cofactorc c nmmn det =+ . for the convenience, the matrix elements c+mn will be given in terms of c11, c22, c32, and c31 only. 33 1131 2 32 2 31221111 2 322211 11 )(2)( ++ = −+− − = c ccccccc ccc c 322321 1131 2 32 2 31221111 11323132 12 )(2)( ++++ === −+− − = ccc ccccccc cccc c 4 samarasekara: effect of demagnetization… r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 1 9 ( 2 0 0 7 ) 31 1131 2 32 2 31221111 3122 2 32 13 )(2)( ++ = −+− − = c ccccccc ccc c )(2)( 1131 2 32 2 31221111 2 31 2 11 22 ccccccc cc c −+− − =+ (2) matrices c and c+ are highly symmetric, and total energy can be given as (samarasekara 2006), e(θ)=e0-0.5[c+11(α12+α32)+c+32(2α1α2+2α2α3)+c+31(2α1α3)+α22c+22] from equation 1, θ ω θθ λλ 2)4()2( 21031 cos2)(4 3 )()( ddbb ++φ+φ+φ−== θ ω θ λλ 2)4()2( 102 cos2)2(4 3 )( ddb ++φ+φ−= because in this case, α1=α3 e(θ)=e0-0.5[2c+11α12+4c+32α1α2+2c+31α12+α22c+22] (3) first simulation will be carried out for 510 )4()2( ====== ωωωωωω msoutinm dand khhdj for sc(001) lattice, z0=4, z1=1, z2=0, φ0=9.0336, φ1= -0.3275 and φ2=0 (usadel and hucht 2002), ωµ θ ωωωω 0 32232112 22cos2456.008.10 d ncccc ++−==== ωµωω 0 3113 2 dncc == θ ωω 2cos2456.2008.103311 +== cc )sin3(coscos20 2 222 0 θθθ ωµ −+− d n θθθ 2sin40cos10sin10 +++ θ ω 2cos49.20164.2022 += c )sin3(coscos20 2 222 0 θθθ ωµ −+− d n θθθ 2sin40cos10sin10 +++ θθ ω α ω α 2sin)cos1047.3( 231 +== θθ ω α 2sin)cos10716.3( 22 += samarasekara: effect of demagnetization… 5 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 1 9 ( 2 0 0 7 ) 78.40 −= ω e θ2cos67.9+ )2sin10cos10sin10cos5cos10(3 0 42 θ ωµ θθθθ +−+++− d n for sc(001), 3-d plot of energy versus angle and ωµ 0 dn is given in figure 1. the graph indicates several energy minimums at different values of angle and demagnetization factor. under the influence of these demagnetization factors, the sample can be easily oriented in certain directions corresponding to energy minimums. for example, one minimum can be observed at ωµ 0 dn =8, and the angle corresponding to this demagnetization factor can be obtained from figure 2. angle corresponding to first minimum is 0.6 radians. figure 1: 3-d plot of energy versus angle and ωµ 0 dn for sc(001) ferromagnetic lattice 6 samarasekara: effect of demagnetization… r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 1 9 ( 2 0 0 7 ) figure 2: energy versus angle at ωµ 0 dn =8 for sc (001) lattice for bcc(001) lattice, z0=0, z1=4, z2=0, φ0=5.8675 and φ1=2.7126 (usadel and hucht 2002), ωµ θ ωωωω 0 32232112 22cos03.232.39 d ncccc +−−==== ωµωω 0 3113 2 dncc == θ ωω 2cos03.232.393311 −== cc )sin3(coscos20 2 222 0 θθθ ωµ −+− d n θθθ 2sin40cos10sin10 +++ θ ω 2cos07.464.7822 −= c )sin3(coscos20 2 222 0 θθθ ωµ −+− d n θθθ 2sin40cos10sin10 +++ θθ ω α ω α 2sin)cos10565.3( 231 +== θθ ω α 2sin)cos1053.1( 22 += samarasekara: effect of demagnetization… 7 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 1 9 ( 2 0 0 7 ) 44.760 −= ω e θ2cos67.10+ )2sin10cos10sin10cos5cos10(3 0 42 θ ωµ θθθθ +−+++− d n 3-d plot of energy versus angle and ωµ 0 dn for bcc(001) ferromagnetic lattice is given in figure 3. this graph also indicates several energy minimums. energy is minimum at ωµ 0 dn = 7.5. the figure 4 has been drawn to find the easy directions corresponding to this energy minimum at ωµ 0 dn = 7.5. energy minimum can be observed at 0.6 radians, and this angle gives the easy direction at ωµ 0 dn = 7.5. this graph is smoother compared with the energy curve given in figure 2. figure 3: 3-d plot of energy versus angle and ωµ 0 dn for bcc(001) ferromagnetic lattice 8 samarasekara: effect of demagnetization… r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 1 9 ( 2 0 0 7 ) figure 4: energy versus angle at ωµ 0 dn =7.5 for bcc (001) lattice 3. conclusion: 3-d plots and 2-d plots of energy indicate several minimums implying that the film can be easily oriented in these directions given by angles corresponding to minimums under the influence of certain values of demagnetization factor. at ωµ 0 dn =8, the angle corresponding to first minimum is 0.6 radians for sc(001) ferromagnetic lattice. under the influence of demagnetization factor given by ωµ 0 dn = 7.5, energy minimum can be observed at 0.6 radians for bcc(001) lattice. this simulation can be carried out for any other values of these energy parameters as well. samarasekara: effect of demagnetization… 9 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 1 9 ( 2 0 0 7 ) references: bentaleb m., el aouad n. and saber m 2002. magnetic properties of the spin -1/2 ising ferromagnetic thin films with alternating superlattice configuration, chinese j. phys., 40(3): 307314 dantziger m., glinsmann b., scheffler s., zimmermann b. and jensen p.j 2002. in-plane dipole coupling anisotropy of a square ferromagnetic heisenberg monolayer, phys. rev., b(66): 094416 1-6 lederman david, ricardo ramirez and miguel kiwi 2004. monte carlo simulations of exchange bias of ferromagnetic thin films on fef2(110), phys. rev. b70: 184422 1-7 martin j. klein and robert s. smith 1951. thin ferromagnetic films, phys. rev. 81: 378380 samarasekara p 2006. second order perturbation of heisenberg hamiltonian for nonoriented ultra-thin ferromagnetic films, elec. j. theo. phys., 3(11): 71-83 samarasekara p 2006. a solution of the heisenberg hamiltonian for oriented thick ferromagnetic films, chinese j. phys., 44(5): 377-386 usadel kd and hucht a 2002. anisotropy of ultrathin ferromagnetic films and the spin reorientation transition, phys. rev. b 66: 024419 1-6 sv-lncs ruhuna journal of science vol 6: issue 1 & 2 (2015) issn: 1800-279x  faculty of science university of ruhuna  faculty of science, university of ruhuna sri lanka list of reviewers: ruhuna journal of science, vol. 6 (issue 1 & 2) 2015 the list includes also the reviewers contributed for the submitted papers which are subsequently rejected. professor upali s. amarasinghe, department of zoology, university of kelaniya, kelaniya, sri lanka professor oscar amarasinghe, department of agriculture economics, university of ruhuna, matara, sri lanka dr. channa n. bambaradeniya, iucn (former), sri lanka professor m.a.k.l. dissanayaka, institute of fundamental studies, kandy, sri lanka dr. p. imali n. fernando, department of management sciences, faculty of management, uwa wellassa university, sri lanka dr. thusitha s.l.w. gunawardana, dept. of marketing, faculty of management and finance, university of ruhuna. professor k.b. suneetha gunawickrama, department of zoology, university of ruhuna, matara, sri lanka dr chamila jayasinghe, department of food science & technology, wayamba university of sri lanka mr. rohan pethiyagoda, national museum, australia. professor r.p.v.j. rajapakshe, department of veterinary pathobiology, university of peradeniya, sri lanka professor e. i. l. silva, chairman/ceo, water resources science and technology (wrst), colombo, sri lanka professor preethi soysa, department of biochemistry & molecular biology, faculty of medicine, university of colombo, kynsey road, colombo 08, sri lanka professor p. vinobaba, department of zoology, eastern university, vantharumoolai, chenkalady, sri lanka professor janak r. wedagedara, simcyp-certara ltd. 2 john street, united kingdom s2 4su dr. deepthi wickramasinghe, department of zoology, university of colombo, colombo 3, sri lanka professor palitha wijesundera, department of physics, university of kelaniya, kelaniya professor wipula b. yapa, department of zoology, university of colombo, colombo 3, sri lanka rjs-2008-shayam-perera.dvi ruhuna journal of science vol. iii, september 2008, pp. 44–52 http://www.ruh.ac.lk/rjs/ i s s n 1800-279x © 2008 faculty of science university of ruhuna. comparison of flow-induced crystallization melt spinning processes s.s.n. perera department of mathematics, university of colombo, colombo 03, sri lanka correspondece: ssnp@maths.cmb.ac.lk abstract. the melt spinning process for artificial fibers has been studied by many research groups throughout the world during the last four decades. however, comparison of flow-induced crystallization melt spinning processes has not yet been treated in the literature. in this study, we analyse the dynamics of the flow induced crystallization melt spinning process. further, we study the sensitivity of the process with respect to fluid shear modulus. non-newtonian and maxwell-oldroyd models are used to describe the rheology of the polymer the fiber is made of. it has been found that the flow-induced crystallization maxwell-oldroyd model has an upper bound for the final velocity. key words : fiber spinning, non-newtonian, maxwell-oldroyd 1. introduction the fiber spinning process is used to make all types of synthetic textile fibers (nylon, polyester, rayon, etc.). in the melt spinning version of the process, molten polymer is extruded a die called a spinneret to create a thin long fiber. far away from the spinneret, the fiber is wrapped around a drum, which pulls it away at a pre–determined take–up speed. the take–up speed is much higher than the extrusion speed; in industrial processes the take–up speed is about 50m/s and the extrusion speed is about 10m/s, see (2, 4). the ratio between the take–up speed vl and the extrusion speed v0 is called draw–ratio and denoted by d = vl/v0 > 1 and hence the filament is stretched considerably in length and therefore decreases in diameter. the ambient atmosphere temperature is below the polymer solidification temperature such that the polymer is cooled and solidifies before the take–up, see figure 1. in industrial processes a whole bundle of hundreds of single filaments is extruded and spun in parallel, however for the analysis we consider a single filament. the dynamics of melt spinning processes has been studied by many research groups throughout the world during the last decades starting with the early works of kase and matsuo (3) and ziabicki (10). despite great scientific progress in the dynamics of fiber formation processes, especially in flow induced crystallization process, there are still some unsettled issues. for example, comparison of flow-induced crystallization melt spinning processes has not yet been treated in the literature. the sensitivity of the maxwell-oldroyd model (both isothermal and non-isothermal) with respect to the characteristic relaxation 44 s.s.n. perera: comparison of flow-induced ... ruhuna journal of science iii, pp. 44–52, (2008) 45 figure 1 sketch of the melt spinning process. time has been discussed in the literature without considering the crystallization process (7, 8). to investigate the same concept with the crystallization process is quite interesting from both the theoretical and industrial points of view because it is closely related to high quality control of products and its theoretical analysis involves the fundamental understanding of the nonlinear dynamics of the process. in this study, we analyse the behaviour of the flow induced crystallization melt spinning process using non-newtonian and maxwell-oldroyd models. further, we study the sensitivity of the flow induced crystallization process with respect to the fluid shear modulus. 2. melt spinning models considering the basic conservation laws for the mass, momentum and energy of the viscous polymer jet, one can obtain the following set of equations, by averaging over the cross– section of the slender fiber, see (4, 5, 6, 7). ρav = w0 . (1a) ρav dv dz = daτ dz − √ a π cd ρair v2 + ρag , (1b) ρcpv dt dz = − 2α √ π √ a (t − t∞) + ρ∆h v dφ dz , (1c) in the mass balance (1a), a denotes the cross–sectional area of the fiber, and v is the velocity of the fiber along the spinline. the density ρ of the polymer is assumed to be constant. in the momentum balance (1b), z denotes the coordinate along the spinline and the axial stress τ is related via the constitutive equations (for the non-newtonian case equation (1d) and the maxwell-oldroyd case equation (1e)) τ = 3η dv dz , (1d) s.s.n. perera: comparison of flow-induced ... 46 ruhuna journal of science iii, pp. 44–52, (2008) τ + λ ( v dτ dz − 2τ dv dz ) = 3η dv dz (1e) to the viscosity η and characteristic relaxation time λ. in the energy equation (1c), t and cp denote the temperature and the heat capacity of the polymer, t∞ is the temperature of the quench air and α denotes the heat transfer coefficient between the fiber and the quench air. according to (4), we assume the following relation for the heat transfer coefficient α = 0.21 r0 κre 1 3 air [ 1 + 64v2c v2 ] 1 6 depending on the reynolds–number of the quench air flow reair = 2vρair ηair √ a π . here r0 is the radius of the spinneret, ρair, ηair and κ represent respectively the density, viscosity and heat conductivity of the air and vc is the velocity of the quench air. the crystallization process generates an enthalpy by change and this is represented by third term of the equation (1c) and ∆h is the specific heat of fusion of a perfect crystal and φ is the degree of crystallinity. according to (4), the model for the evolution of φ is given by v dφ dz = (φ∞ − φ)kmax exp [ −4 ln 2 ( t − tmax d )2 ] . (1f) here φ∞ is the ultimate crystallinity, kmax the maximum crystallization rate, tmax the fluid temperature having the maximum crystallization rate and d denotes the crystallization half width temperature range. the viscosity and characteristic relaxation time are given by η = η0 exp [ ea rg ( 1 t − 1 t0 )] , (1g) λ = λ0 exp [ ea rg ( 1 t − 1 t0 )] . (1h) here η0 > 0 is the zero shear viscosity at the initial temperature t0, ea denotes the activation energy, rg is equal to the gas constant and λ0 = η0 g (g is the fluid shear modulus). the system (1) is subject to the boundary conditions v = v0 t = t0 φ = 0 at z = 0 (1i) v = vl at z = l (1j) where l denotes the length of the spinline. s.s.n. perera: comparison of flow-induced ... ruhuna journal of science iii, pp. 44–52, (2008) 47 3. dimensionless form introducing the dimensionless quantities z∗ = z l , v∗ = v v0 , z∗ = z l , t ∗ = t t0 , a∗ = a a0 , τ∗ = τl η0v0 , φ∗ = φ φ∞ , the system (1) can be formulated in dimensionless form. dropping the star and considering the non-newtonian and maxwell-oldroyd cases the system can be presented as follows dv dz = τ 3η , (2a) dτ dz = re ( 1 3η vτ − fr−1 + c1v 5 2 ) + 1 3η τ2 v , (2b) dt dz = −c2 (t − t∞)√ v + ∆h φ∞ t0cp dφ dz , (2c) dφ dz = kmaxl v0 ( 1 − φ v ) exp [ −4 ln 2 ( t − tmax d )2 ] . (2d) rev dv dz = dτ dz − τ v dv dz + re ( fr−1 −c1v 5 2 ) , (3a) 3η dv dz = τ + de ( v dτ dz − 2τ dv dz ) , (3b) dt dz = −c2 (t − t∞)√ v + ∆h φ∞ t0cp dφ dz , (3c) dφ dz = kmaxl v0 ( 1 − φ v ) exp [ −4 ln 2 ( t − tmax d )2 ] . (3d) in system (2) and (3), re = ρlv0η0 is the reynolds number, fr −1 = gl v20 is the inverse of the froude number, c1 = cd ρair l √ π ρ √ a0 is the scaled drag coefficient and c2 = 2αl √ π ρcpv0 √ a0 denotes the scaled heat transfer coefficient. the deborah number de is given by de = λ0v0 l exp [ ea rgt0 ( 1 t − 1 )] . the systems (2) and (3), are subject to the boundary conditions v(0) = 1 t (0) = 1 and φ(0) = 0, v(1) = d, where d is the draw ratio. s.s.n. perera: comparison of flow-induced ... 48 ruhuna journal of science iii, pp. 44–52, (2008) 4. numerical results 4.1. numerics both systems ((2) and (3)) of ode are solved using the matlab routine ode23tb. this routine uses an implicit method with backward differentiation to solve stiff differential equations. it is an implementation of tr-bdf2 (9), an implicit two stage runge-kutta formula where the first stage is a trapezoidal rule step and the second stage is a backward differentiation formula of order two. since both systems are boundary value problems, the shooting method is used to solve them. 4.2. shooting method now, we present the main steps of the shooting method in general. let y = (v, τ, t, φ). then one can write the system (2) in the following form dy dz = f (y, u) , with y1(0) = 1, y1(1) = d, y3(0) = 1, y4(0) = 0, (4) where f (y, u) =       τ 3η re ( 1 3η vτ − fr −1 + c1v 5 2 ) −c2 (t −t∞)√ v + ∆h φ∞ t0cp ϑ kmaxl v0 ( 1−φ v ) exp [ −4 ln 2 ( t −tmax d )2 ]       , and ϑ is given as follows: ϑ = kmaxl v0 ( 1 − φ v ) exp [ −4 ln 2 ( t − tmax d )2 ] . let us make an initial guess s for y2(0) and denote by y(z; s), the solution of the initial value problem dy dz = f (y, u) , with y1(0) = 1, y2(0) = s, y3(0) = 1, y4(0) = 0 . (5) now we introduce a new dependent variable x(z; s) = ∂y ∂s and define the second system as follows ∂x ∂z = ( ∂ f ∂y ) x with x1(0; s) = 0, x2(0; s) = 1, x3(0; s) = 0, x4(0; s) = 0. (6) the solution y(z; s) of the initial value problem (5) coincides with the solution y(z) of the boundary value state system (4) provided that the value s can be found such that ϕ(s) = y1(1; s) − d = 0. s.s.n. perera: comparison of flow-induced ... ruhuna journal of science iii, pp. 44–52, (2008) 49 using the system (6), ϕ′(s) can be computed as follows ϕ′(s) = x1(1; s). now, using newton–iteration, a sequence (sn)n∈n is generated by sn+1 = sn − ϕ(sn) ϕ′(sn) for a given initial guess s0. if the initial guess s0 is a sufficiently good approximation to the required root of ϕ(s) = 0, the theory of the newton–iteration method ensures that the sequence (sn)n∈n converges to the root s. by rearranging the system (3), the function f (y, u) can be obtained for the maxwell-oldroyd model. 4.3. results figure 2 shows the spinline velocity, temperature profile and crystallinity index of the nonnewtonian and maxwell-oldroyd models. concerning the temperature profile, one sees a jump in the temperature owing to the heat released due to crystallization. further, the behaviour of the temperature and crystallinity profiles are close in both cases. 0 0.2 0.4 0.6 0.8 1 15 20 25 30 35 40 45 50 length m v m /s 0 0.2 0.4 0.6 0.8 1 50 100 150 200 250 300 length m t e m p e ra tu re ° c 0 0.2 0.4 0.6 0.8 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 length m c ry s ta ll in it iy i n d e x 0.572 0.574 0.576 0.578 0.58 156 158 non−newtonian maxwell−oldroyd 0.5 0.55 0.36 0.38 0.4 figure 2 spinline velocity profile(up-left), spinline temperature (up-right) profile, crystallinity index (down-left). figure 3 shows the velocity profile of the maxwell-oldroyd model depending on the fluid shear modulus. from this one sees the the velocity profile of the melt spinning process s.s.n. perera: comparison of flow-induced ... 50 ruhuna journal of science iii, pp. 44–52, (2008) 0 0.2 0.4 0.6 0.8 1 15 20 25 30 35 40 45 50 length m v m /s 0.175 0.18 0.185 43.8 44 44.2 g=1.1⋅ 105 g=9⋅ 104 g=8⋅ 104 g=7⋅ 104 figure 3 velocity profile depending on fluid shear modulus (pa). has no significant variation with respect to the fluid shear modulus. but in the simulation process we experienced difficulties when the fluid shear modulus decreased. we noticed that it is needed to use higher initial guesses for the stress variable for the lower value of the fluid shear modulus. figure 4 visualizes the final velocity vs the initial guess for stress in different fluid shear modulus. one sees from this that for a particular fluid shear modulus value, the final velocity approaches a fixed value. for example, if we consider g = 4 · 104 pa, then the final velocity approaches 38 m/s. in other words, in this case (i.e. g = 4 · 104 pa) if we set the final velocity as 50 m/s then theoretically ode system cannot be solved. the fluid shear modulus is related to the characteristic relaxation time; lower g yields higher λ. we can expect this behaviour since the maxwell-oldroyd model (without the crystallization process) has an upper bound for the final take-up velocity which depends on the characteristic relaxation time (see (7)). this means that the flow induced crystallization maxwell-oldroyd model also has an upper bound for the final take-up velocity. 5. conclusions we compared the velocity, temperature and crystallization index profiles of the flow induced crystallization melt spinning process using non-newtonian and maxwell-oldroyd models. the quantitative behaviour of the maxwell-oldroyd case is similar to the non-newtonian case. but the qualitative behavior of the maxwell-oldroyd model is totally different for s.s.n. perera: comparison of flow-induced ... ruhuna journal of science iii, pp. 44–52, (2008) 51 0 500 1000 1500 2000 20 25 30 35 40 45 50 55 60 65 initial guess for stress f in a l v e lo c it y m /s g=4⋅ 104 g=5⋅ 104 g=6⋅ 104 g=8⋅ 104 figure 4 final velocity depending on the initial guesses for the stress. the lower values of the fluid shear modulus. using numerical simulation, we have seen that the flow induced crystallization maxwell-oldroyd model cannot be solved with any arbitrary final velocity; i.e., the flow induced crystallization maxwell-oldroyd model has an upper bound for the final take-up velocity which depends on the material properties of the polymer. theoretically, setting an arbitrary value for the final velocity may yield the spinning process unstable. instability leads to irregular fibers or induces breakage of the individual filaments of the spinline. clearly, this investigation is important from an industrial point of view. references [1]bird rb, amstrong rc, hassager o (1987) dynamics of polymeric liquids. 2nd edition, volume 1: fluid mechanics, john wiley & sons. [2]brünig h, roland h, blechschmidt d (1997) high filament velocities in the underpressure spunbonding nonwoven process. ifj:129-134, december 1997. [3]kase s, matsuo t (1965) studies on melt spinning, fundamental equations on the dynamics of melt spinning. j. polym. sci. part a 3: 2541-2554. [4]langtangen hp (1997) derivation of a mathematical model for fiber spinning. department of mathematics, mechanics division, university of oslo, december 1997. s.s.n. perera: comparison of flow-induced ... 52 ruhuna journal of science iii, pp. 44–52, (2008) [5]lee js, jung hw, hyun jc, seriven le (2005) simple indicator of draw resonance instability in melt spinning processes. aiche j., vol. 51, no. 10:2869-2874. [6]lee js, shin dm, jung hw, hyun jc (2005) transient solution of the dynamics in low-speed fiber spinning process accompanied by flow-induced crystallization. j. non-newtonian fluid mech. 130:110-116. [7]perera ssn (2008) phase-space analysis of melt spinning processes, nihon reoroji gakkaishi, vol: 36 no. 4:161-166. [8]perera ssn (2009) viscoelastic effect in the non-isothermal melt spinning processes, applied mathematical sciences, vol: 3 no: 4: 177-186. [9]shampine lf, reichelt mw (1997) the matlab ode suite. siam j. sci. comput., vol: 18:1-22. [10]ziabicki a (1976) fundamentals of fiber formation. wiley-interscience, new york. sv-lncs ruhuna journal of science vol. 5: 31-36, 2014 http://rjs.ruh.ac.lk/ issn: 1800-279x short paper 31  faculty of science, university of ruhuna salient characters of weedy rice (oryza sativa f. spontanea) populations in highly infested areas in sri lanka s. somaratne 1 , k. d. k. karunarathna 1 s. r. weerakoon 1* , a. s. k. abeysekera 2 and o. v. d. s. j. weeresena 3 1 department of botany, open university of sri lanka, p.o. box 21, nawala, sri lanka 2 rice research and development institute, batalagoda, ibbagamuwa, sri lanka 3 institute of biochemistry, molecular biology and biotechnology, university of colombo, sri lanka * correspondence: shyamaweerakoon@gmail.com received: 20 may 2014, revised version accepted: 01 october 2014 abstract. weedy rice (wr) was first reported in 1990 and it is occurring with varying population densities in all agro-ecological zones in sri lanka. the identification of wr eco-types using agromorphological characters which vary with time and remains as a major problem among the farmers and the agronomists. the wr population possesses a number of pleisomorphic (primitive) and apomorphic (derived) characters. this study focuses on identification of primitive and derived characters observed in wr. the identification of the salient trends of specialization of characters in wr populations facilitates the understanding of the rate diversification of wr populations. seeds of presumed wr eco-types were collected from five different locations in kurunegala and matara districts. five replicates with three plants of each eco-type were planted in plastic pots with representative paddy soils from each location. replicates were arranged in complete randomized design (crd). agro-morphological characterization (using thirty characters) of wr eco-types, wild rice and cultivated rice varieties were made using a standard characterization catalogue. the collected data were separated into nominal and scalar variables and the nominal data were used to construct classification and regression trees using cart algorithm. the long-fully awned and absence of awn were pleisomorphic characters and short-fully awned and long-partly awned characters were apomorphic in wr eco-type populations in sri lanka. these characters could be hypothesized as derived from mixing of germplasm either of cultivated or wild rice varieties indicating the possibilities of cross-pollination among wild, cultivated and weedy rice eco-types. keywords. agro-morphological, oryza sativa f. spontanea, salient trends, weedy rice. somaratne et al salient characters of weedy rice in sri lanka ruhuna journal of science (december 2014) 32 1 introduction paddy cultivation in the country has been threatened by a number of challenges, and among those, the emergence of rice weeds plays an important role. it has been reported that yield loss due to the infestation of rice fields ranged from 30-40% (khush 1997). however, the appearance of “weedy rice” (wr) became the most prominent weed problem in rice growing areas. wr first reported in 1990 in a small area in ampara district in eastern province of sri lanka and by 1997 it had become a serious problem in the area. wr now occurs, in varying population densities, in all agro-ecological zones of the country (abeysekara et al. 2010). the term “weedy rice” generally includes all the species of the genus oryza which behave as rice and is in rotation with rice weeds. wr populations have been reported in many rice-growing areas in the world where the crop is directly seeded (ferrero and finassi 1995). the phylogenetic origin of the weedy forms is closely related to that of cultivated rice. many weedy plants share most of the features of the two cultivated species oryza sativa and o. glaberrima (khush 1997). agro-morphological and topographical characteristics of the plants have been the criteria in the identification and classification of wr eco-types (qinjin et al. 2006). however, at seedling stage, weedy plants are difficult to distinguish from the crop (hoaghland and paul 1978). the previous studies carried out on the wr populations in the country, especially in matara and kurunegala districts have indicated that higher numbers of wr eco-types were reported from the kurunegala (intermediates zone) district. the main problem faced by the farmers and the agronomist is the identification of wr eco-types using agro-morphological characters. meanwhile, there will be a number of derived characters (apomorphic) and primitive characters (plesiomorphic) within the wr population in the country. the present study focused on the identification of primitive and derived characters observed in the wr populations and the use of those features to identify the primitive eco-types in the wr populations in the country. in addition, the identification of primitive vs. novel characters of wr populations facilitates the understanding of the salient characters and their trends within weedy and cultivated rice populations. somaratne et al salient characters of weedy rice in sri lanka ruhuna journal of science (december 2014) 33 table 1. agro-morphological characters used for the characterization wr and cultivated rice varieties included in the study (pgrc, 1999) c h a ra c te r n u m b e r c h a ra c te r description c h a ra c te r n u m b e r c h a ra c te r description 1 seedling height (cm) recorded at the five leaf stage 16 culm angle 1.erect 3.intermediate 5.open 7.spreading 9.procumbent 2 leaf blade length (cm) measured from top most leaf below the flag leaf on the main culm at late vegetative stage. 17 inter node color after full heading 1.green 2.light gold 3.purple lines 4.purple 3 leaf blade width (mm) measured at the widest portion of the leaf blade 18 culm strength 1.strong 3.moderately strong 5.intermediate 7.weak 9.very week 4 leaf blade pubescent 1.glabrous 2.intermediate 3. pubescent 19 panicle length from the base to the tip of the panicle 5 leaf blade color 1. pale green 2. green 3. dark green 4.purple tips 5.purple margins 6.purple blotch 7. purple 20 panicle type 1.compact 5.intermediate 9.open 6 basal leaf sheath color 1.green 2.purple lines 3.light purple 4.purple 21 secondary branching 0.abscent 1.light 2.heavy 3.clustering 7 leaf angle 1.erect 2.intermediate 3.horizontal 4.descending 22 panicle exsertion 1.well exsertion 3.moderately 5.just exterted 7.partly exserted 9.enclosed 8 flag leaf angle 1.erect 2.intermediate 3.horizontal 4.descending 23 awning after full heading 0. absent 1.short and partly awned 5.short and fully awned 7.long and partly awned 9. long and fully awned 9 ligule length (mm) measured at late vegetative stage 24 apicus color 1.white 2.straw 3.brown 4.red 5.red apex 6.purple 7.purple apex 10 ligule color 0.absent 1.white 2.purple lines 3.purple 25 lemma and palea color 0.straw 2.gold 3.brown spot on straw 4.brown 5.reddish to light purple 6.purple spots on straw 7.purple 8.black 9.white 11 collar colour 1.pale green 2.green 3.purple 26 lemma and palea pubescence 1. glabrous 2.hairs on lemma keel 3. hairs on upper portion 4.short hairs 5.long hairs 12 auricle colour 0.abscent 1.pale green colour 2.purple 27 sterile lemma color 1.straw 2.gold 3.red 4.purple 13 days of heading no. of days from effective seeding to 50% heading 28 sterile lemma length 1.short 3. medium 5.long 7.extra long 9.asymmetrical 14 culm length (cm) from ground level to the base of the panicle 29 100 grain weight a random sample of 100 well developed grains dried 13% moisture content 15 culm number total no. of grain bearing and non bearing tillers 30 seed coat color 1.white 2.light brown 3.speckled brown 4.brown 5.red 6.variable purple 7.purple somaratne et al salient characters of weedy rice in sri lanka ruhuna journal of science (december 2014) 34 2 materials and methodology seeds of presumed different wr eco-types which were identified during field collection, wild rice (o. nivara) and cultivated rice (bg 358, bg 352, bg 359, bg 379-2, bg 307, at 362) varieties were collected from five different locations in matara and kurunagala districts. collected seeds were subjected to dormancy breaking treatments (hot water treatment) and sown in plastic trays (60 cm x 30 cm x 5 cm) in a plant house at the open university of sri lanka, nawala. a total of five replicates with three plants in a replicate of each wr eco-type, wild rice and cultivated rice varieties were planted in plastic pots with representative paddy soils from each location. complete randomized design (crd) was used for this experiment. agromorphological characterization (thirty characters) (table 1) of wr eco-types, wild rice and cultivated rice varieties were made using the standard characterization catalogue (pgrc, 1999). the collected data were separated into nominal and scalar variables and the nominal data were used to construct classification and regression trees using cart algorithm (breiman et al. 1984). the chaid (chi-squared automatic interaction detection) procedure was specified and other parameters were set at default values. the analysis was carried out using spss pc ver. 20. 3 results and discussion the data set included five cultivated rice varieties from kurunagala and matara district and one wild rice variety (o. nivara). the number of wr ecotype varies with the district and the highest number of wr eco-types was reported from the kurunegala district. the result of the classification and regression tree analysis (cart) is shown in fig. 01. according to the figure, there were ten child nodes and three parental nodes. the character, lemma and palea color split the entire sample into two groups. the characters, seed coat color and awning after full heading further split the child nodes into six terminal groups. the used agro-morphological character in the study plays an important role in classifying the wr eco-types, cultivated and wild rice into well-separated entities indicating their primitiveness of the characters they possess. characters such as the lemma and palea color, seed coat color, awning after full heading seem to be derived characters when compared to other 27 characters used for characterization. further, gold and gold furrows on straw background, was the characteristic state of palea color which occurs in o. nivara. meanwhile, all the cultivated rice varieties possess brown furrows on straw. the two character states of lemma and palea color seem to occur in the populations of wild rice and cultivated rice and it can be considered as a primitive character. somaratne et al salient characters of weedy rice in sri lanka ruhuna journal of science (december 2014) 35 fig 1. the classification and regression tree analysis of weedy, wild and cultivated rice varieties. somaratne et al salient characters of weedy rice in sri lanka ruhuna journal of science (december 2014) 36 the character, seed-coat color also indicated a considerable variation between the wild, cultivated and wr eco-types and red seed coat color occur in seeds of o. nivara and brown and light brown color commonly found in wr eco-types and white seed coat color were restricted to the cultivated rice varieties. the white and red seed coat colors seem to be primitive since the characters are shared by o. nivara and cultivated rice varieties. the character such as awning after full heading also indicated the limited occurrence and on this basis o. nivara and cultivated rice varieties could be considered as potential parents of the most weedy rice eco-types. the appearance of the characters, long and fully awned and the absence of the awn can be considered as primitive and the other characters such as short and fully awned and long and partly awned rice seeds can be considered as derived characters. 4 conclusion the limited number of salient agro-morphological characters are found within the wr eco-types and those characters are supposed to be derived from a mixing of germplasm either of cultivated or wild rice varieties. this process is enhanced by the variation in climatic conditions. the character, awning after full heading also indicated the limited occurrence and on this basis, o. nivara and cultivated rice varieties were the potential parents of the most of the wr eco-types. the appearance of long-fully owned and the absence of the awn is pleisomorphic (primitive) characters, while short-fully and long-partly awned characters are apomorphic (derived) characters in weedy rice populations in sri lanka. these characters are hypothesized as derived from mixing of germplasm either of cultivated or wild rice varieties indicating higher possibilities of cross-pollination among wild, cultivated and wr eco-types. however, further studies are necessary to confirm the results. references abeysekara a. s. k., nugaliyadda l., herath h. m. s., wickrame u. b. and iqbal y. b. (2010). weedy rice: a threat to direct seeded rice cultivation in sri lanka. rice congress 2010, pgrc, gannoruwa. pp 17-18. breiman l., friedman j., olshen r., stone c. (1984). classification and regression trees. crc press, boca raton. ferrero a. and finassi, a. (1995). viability and soil distribution of red rice (oryza sativa l. var. sylvatica) seeds. in med. fac. landbouw., rijksunv. gent. pp. 205-211. hoagland r.e. and paul r.v. (1978). a comparative sem study of red rice and several commercial rice (oryza sativa) varieties. weed science 26:619-625. khush g. s. (1997). origin, dispersal, cultivation and variation of rice. plant molecular biology 35: 25-34. pgrc (plant genetic resources centre) (1999). characterization catalogue of rice (oryza sativa) department of agriculture., ministry of agriculture and lands, sri lanka. qinjin c., bao r. l., hui x., jun r., francesco s., alberto s. and fabbrizio g. (2006). genetic diversity and origin of weedy rice (oryza sativa f. spontanea) populations found in northeastern china revealed by simple sequence repeat (ssr) makers. annals of botany 98: 1241-1252. rjs-2007-wijewick-mpaw.dvi ruhuna journal of science vol. ii, september 2007, pp. 10–17 http://www.ruh.ac.lk/rjs/ i s s n 1800-279x © 2007 faculty of science university of ruhuna. a solution for non-stationary, slowly-rotating, cylindrically symmetric, perfect fluid universe wijewickrema p.k.c.m. and wijayasiri m.p.a. department of mathematics, university of ruhuna, matara, sri lanka. correspondece: wijaya@maths.ruh.ac.lk abstract. an analytic solution for the relativistic field equations is obtained for a non-stationary, slowly rotating, cylindrically symmetric distribution of perfect fluid universe. the new metric, is regular with the exception at the point r = 0. there is a gravitational singularity at r = 0. at t = 0 the pressure p and density ρ are maximum and tends to ∞ throughout the radial coordinate r (0 < r < ∞), but the solutions are well behaved for t > 0, and p and ρ are decreasing to zero as t increases through the range 0 < t < ∞. so according to the model, it has the big bang singularity at t = 0, where ρ diverges. key words : einstein’s field equations, slow rotation, cylindrically symmetric, perfect fluid. 1. introduction numerous attempts to obtain exact solutions of einstein’s field equations representing cylindrically symmetric perfect fluid matter distributions have so far been reported. but a considerable number of these solutions have not been diligent in interpreting in a physically meaningful way or even in a less acceptable way. thus within a collection of exact solutions one finds many whose physical meaning is unknown or only partially understood. the main reason for this may be that interpretation is difficult and uncertain. another reason is, the observational verification of general relativity now and in the near future, is likely to depend on a very small number of exact solutions. therefore some workers feel it is a waste of time to try to interpret metrics which have no prospect of observational verification. but it is obvious that we cannot claim to understand general relativity unless we can determine the physics of the exact solutions we know. in the case of some of the solutions there is no physics, or the solutions do not agree with known physics. even if this is the case it is important to try to obtain solutions and interpret them as well as we can. even though a good collection of solutions exist for non-rotating cases (adler, r., bazin, m. and shiffer 1975, davidson 1992, gasperini and de sabbata 1985, stephani 1982), there is a rareness of the solutions for rotating cases. but during the last 40-45 years rotating objects have been studied quite extensively (saha 1981). some of them have studied the structure and stability of rapidly rotating fluid spheres (butterworth and isper 1975) with various amounts of uniform and differential rotation and some have studied uniformly rotating white dwarfs and neutron stars up to second order in angular velocity. other papers related to numerical approach on this subject, have presented the analytic theory of slowly 10 wijewickrema and wijayasiri: a solution for non-stationary ... ruhuna journal of science ii, pp. 10–17, (2007) 11 and uniformly rotating general relativistic bodies and discussed conditions of stability. it is very important to study objects with some kind of rotation, as almost all objects in the sky exhibit some form of rotation, and today there is even the possibility of the universe itself having a slight rotation. 2. field equations and method of obtaining solution the tetrad formalism (chandrasekhar 1983) has been used to obtain the tensor components which we wanted to build-up the system of differential equations, as handling metric coefficients is rather easy in this method than in an approach such as involving the use of euler-lagrange equations. we consider the non-stationary, cylindrically symmetric metric in the following general form: ds2 = e2νdt 2 − e2λdr2 − dz2 − tr2(dφ − ωdt)2, (1) where ν, λ and ω are functions of both time t and spatial coordinate r only. ω(r, t) represents the dragging of inertial frames. for slowlyrotating spacetimes, (i.e. only the first order terms in the angular velocities ω = dφ dt and ω are considered) the following tetrad components of the ricci tensor are obtained. r(00) = −(νrr − λrνr + ν2r + νr r )e−2λ − ( 1 2 νt t + 1 4t 2 − λtt + νt λt − λ2t )e −2ν, r(01) = ( 1 2tr − 1 2 νr t − λt r )e−(ν+λ), r(03) = [ 1 2 (νr +λr)rt 1/2ωr − 3 2 t 1/2ωr − 1 2 ωrr rt 1/2]e−(ν+2λ), r(11) = −(λtt − νt λt + λ2t + 1 2 λt t )e−2ν + (νrr − λr νr + ν2r − λr r )e−2λ, (2) r(13) = [ 1 2 (νt +λt )rt 1/2ωr − 3 4 rωr t 1/2 − 1 2 ωrt rt 1/2]e−(2ν+λ), r(22) = 0, r(33) = −( 1 2 λt t − 1 2 νt t − 1 4t 2 )e−2ν − ( λr r − νr r )e−2λ; where lower indices r and t of each variable denote differentiation with respect to the spatial coordinate r and time t, and the lower indices 0, 1, 2, 3 represent the time coordinate t and the spatial coordinates r, z, φ respectively. here and in what follows the bracketed indices denote that a tetrad frame is being used. the matter distribution is considered as a perfect fluid with fluid pressure p and mass density ρ. t(µν) = (ρ + p)u(µ)u(ν) − pg(µν). (3) wijewickrema and wijayasiri: a solution for non-stationary ... 12 ruhuna journal of science ii, pp. 10–17, (2007) for the case of slow rotation, the tetrad components of the four-velocity are obtained as: u(0) = 1, u(1) = 0, u(2) = 0, (4) u(3) = rt 1/2(ω − ω) eν . with these equations, (3) can be written as: t(00) = ρ, t(03) = (p + ρ)e−νrt 1/2(ω − ω), t(11) = p, (5) t(22) = p, t(33) = p and t becomes, t = ρ − 3 p. (6) according to the einstein field equation, r(µν) = −8π(t(µν) − 1 2 g(µν)t ), (7) again the ricci components are obtained as: r(00) = −4π(ρ + 3 p), r(01) = 0, r(03) = −8πr(p +ρ)e−νt 1/2(ω−ω), r(11) = −4π(ρ − p), (8) r(13) = 0, r(22) = −4π(ρ − p), r(33) = −4π(ρ − p). by using the equations (2) and (8) we get the following system of equations: (νrr −λrνr +ν2r + νr r )e−2λ +( νt 2t −λtt +νt λt −λ2t + 1 4t 2 )e−2ν = 4π(ρ+3 p), ( νr 2t − 1 2tr + λt r )e−(ν+λ) = 0, wijewickrema and wijayasiri: a solution for non-stationary ... ruhuna journal of science ii, pp. 10–17, (2007) 13 [ 3 2 t 1/2ωr + rt 1/2 2 ωrr − (νr + λr) 2 rt 1/2ωr]e−(ν+2λ) = 8πrt 1/2(ρ+ p)(ω−ω)e−ν, (−νrr + λrνr − ν2r + λr r )e−2λ + (λtt − νt λt + λ2t + λt 2t )e−2ν = 4π(ρ − p), (9) [ 3rωr 4t 1/2 + ωrt 2 t 1/2r − (νt + λt ) 2 rt 1/2ωr]e−(2ν+λ) = 0, 4π(ρ − p) = 0, ( λr r − νr r )e−2λ + ( λt 2t − νt 2t − 1 4t 2 )e−2ν = 4π(ρ − p). 3. results assuming that ν is a function of t only and λ is a function of r and t, an analytic solution of these equations is obtained as follows. e2ν = 1 k1 + 4t k2 (10) e2λ = k2t r2 (11) hence the metric as, ds2 = 1 k1 + 4t k2 dt 2 − k2t r2 dr2 − dz2 − tr2(dφ − ωdt)2. (12) e2λ → ∞ at r = 0 ∀t. therefore there is a gravitational singularity at r=0. furthermore, we obtained the following results. p = ρ (13) ρ = k1 32πt 2 (14) ω = 1 (k1 + 4t k2 )1/2t [ r5 (1 + r2)3 ] (15) ω = ω + 2r5(r4 + 5r2 − 20) k1k2(k1 + 4t k2 )1/2(1 + r2)5 (16) where k1 and k2 are arbitrary constants and k1 > 0 and k2 > 0. wijewickrema and wijayasiri: a solution for non-stationary ... 14 ruhuna journal of science ii, pp. 10–17, (2007) 4. discussion here in this attempt, we were interested in obtaining a family of cylindrically symmetric cosmological models for a non-stationary, slowly rotating perfect fluid distribution. in this case, first of all, we had to choose a suitable general metric form. generally, in the case of non-rotating, it is convenient to adopt the metric form appropriate to cylindrical symmetry as: ds2 = d2(r, t)dt 2 − a2(r, t)dr2 − b2(r, t)dz2 −c2(r, t)dφ2. (17) by using this general metric form, it has been found that a one-parameter solution of the einstein field equations for a non-stationary, non-rotating, perfect fluid universe exists. this solution was obtained in (davidson 1992) as: d(r, t) = (1 + r2)−β(β+1)/2(2β+1), a(r, t) = t(3β+1)/(7β+4)(1 + r2)β(3β+1)/2(2β+1), b(r, t) = t β/(7β+4)(1 + r2)β/2, (18) c(r, t) = t(3β+2)/(7β+4)r(1 + r2)β/2. where β is a constant. (to secure physically acceptable models, parameter β is restricted to the range; 0 ≥ β ≥ − 2 5 ). considering one of the above cases, a general metric has been obtained for the case of slowly rotating in the form: ds2 = e2νdt 2 − e2λdr2 − dz2 − tr2(dφ − ωdt)2 (19) on the other hand we can deduce the special case β = 0 of (18), by making ω = 0 and suitably changing the arbitrary variables and constants in the sub sequent work. in this case, the geometric character of the model changes to spatial homogeneity. in this work, we have made several assumptions and conditions (boundary conditions) to develop this metric. one of such main assumptions is that the universe has a slow rotation. it is interesting that there are plenty of physical evidences that almost every object in the sky exhibits some form of rotation, and today there is even the possibility of the universe itself being endowed with a slight rotation. so in order to satisfy the conditions of slow rotation, only the first-order terms in the angular velocities ω and ω have been considered. the following boundary conditions have also been used. since the central axis is nonrotating, ω and ω should satisfy the boundary conditions, ω, ω, ωr, ωr −→ 0 as r −→ 0, in addition to this, it has been assumed that the universe is non-rotating at r −→ ∞. so that ω and ω should satisfy, ω, ω, ωr, ωr −→ 0 as r −→ ∞. furthermore by using equation (14) it is possible to calculate an approximate value for the constant k1. to do this we used the data from the nasa’s wilkinson microwave wijewickrema and wijayasiri: a solution for non-stationary ... ruhuna journal of science ii, pp. 10–17, (2007) 15 2·10 19 4·10 19 6·10 19 8·10 19 1·10 20 time 2·10 -16 4·10 -16 6·10 -16 8·10 -16 1·10 -15 1.2·10 -15 density figure 1 the variation of the density (gkm−3 ) of the universe against time (sec.) 2 4 6 8 10 r 5·10 -42 1·10 -41 1.5·10 -41 2·10 -41 d.i.f. figure 2 the behaviour of the dragging of the initial frame (ω) against the radial coordinate (r). anisotropy probe (wmap) project. it has been estimated the age of the universe to be about 13.7 billion years old with an uncertainty of 200 million years. this measurement was made by locating the first acoustic peak in the microwave background power spectrum to determine the size of the decoupling surface. the light travel to this surface yields a reliable age for the universe. in addition to this, the lower limit of the critical density (5 × 10−15gkm−3) rowe (2001) was assumed as the present density of the universe. by using these two facts, an approximate value for k1 was obtained as 8.4593298 × 10 33gkm−1. hence we were able to plot the graph of the density of the universe against time (in seconds) (see figure 1). by considering the equations (15) and (16), for any fixed value of time, we were able to plot the behaviour of the two angular velocities ω (see figure 2)and ω (see figure 3). here the value of the constant k2 has been assumed as 1. in figure 4 we show the behaviours of the two angular velocities ω (left) and ω (right) against the radial coordinate (r) and time (t). wijewickrema and wijayasiri: a solution for non-stationary ... 16 ruhuna journal of science ii, pp. 10–17, (2007) figure 3 the behaviour of the angular velocity of the perfect fluid (ω) against the radial coordinate (r). figure 4 left: shows the behaviour of the dragging of the initial frame (ω) against the radial coordinate (r) and time (t). right: shows the behaviour of the angular velocity of the perfect fluid (ω) against the radial coordinate (r) and time (t). 5. conclusion according to the metric (12), it is regular with the exception that at the point r = 0, and has a time singularity at t = 0 at which the pressure p and density ρ tend to ∞ throughout the radial coordinate range 0 < r < ∞, but it is subsequently well-behaved if k1 = 0, ρ = 0 ∀t 6= 0 and we may take ρ = 0 ∀t as otherwise ρ → ∞ at t = 0 but becomes zero in an instant!. therefore take k1 6= 0 . further, p and ρ both are decreasing to zero as t increases through the range 0 < t < ∞ and equations (13) and (14) imply that this fluid model has non-negative expressions for the mass density and pressure. even if the solution is not completely concordance with the idea of great big-bang, it makes less disagreements at the critcial points and moreover, the result here seems to agree with the physical interpretation to some extent with the big-bang theory. acknowledgments it is pleasure to thank dr. j. r. wedagedara for his valuable support in the preparation of this paper. references adler, r., bazin, m. and shiffer. 1975. introduction to general relativity. new york: mcgraw-hill. wijewickrema and wijayasiri: a solution for non-stationary ... ruhuna journal of science ii, pp. 10–17, (2007) 17 butterworth, em., jr. isper. 1975. . astrophysical jn. 200. chandrasekhar, s. 1983. the mathematical theory of black holes. new york: oxford university press. davidson, w. 1992. a one-parameter family of cylindrically symmetric perfect fluid cosmologies. phys. rev. d 24. gasperini, m., v. de sabbata. 1985. introduction to gravitation. singapore: world scientific. rowe, e.g.p. 2001. geometrical physics in minkowski spacetime. london: springer. saha, sk. 1981. . phys. rev. d24 . stephani, h. 1982. general relativity. cambridge: cambridge university press.  © 2008 faculty of science university of ruhuna ruhuna journal of science vol. 3, september 2008, pp 34-43 http://www.ruh.ac.lk/rjs/ issn 1800-279x e.p.s.chandana1, n. j. de s. amarasinghe1 and l.a.samayawardhena1 2 1department of zoology, university of ruhuna matara 1department of zoology, faculty of science, university of ruhuna 2present address 20, kuruppu road, boralla abstract. bundala national park, covering an area of 6216 ha, is located about 250 km southeast of colombo, in the hambantota district. the shallow brackish water lagoons located within the park koholankala (390 ha), malala (650 ha), embilikala (430 ha) and bundala (520 ha) form a complex wetland system that harbors a rich bird life, including several species of migratory waterfowl. this led to the declaration of bundala as sri lanka's first ramsar wetland a wetland of international importance especially for migratory waterfowl, in 1990. distribution and the composition of the aquatic bird species inhabiting these lagoons with respect to the habitat characters are poorly understood. present study was conducted (december, 2000-december, 2001) in malala, embillakala and bundala lewaya lagoons with the objective of investigating the relationship between some lagoon parameters (salinity, perimeter, area) and water bird abundance and diversity. data were collected weekly basis. bird abundance and the composition significantly differ among lagoons. this study revealed the most abundant bird groups in each lagoon. highest aquatic bird diversity was recorded in embillakala. this high diversity in embillakala lagoon can be partly attributed to its moderate salinity, water depth and abundance of aquatic macrophytes. lowest aquatic bird diversity was recorded in bundala lewaya. this study also revealed that salinity, aquatic macrophytes and lagoon area were key determinants of aquatic bird abundance. although these lagoons are in the same landscape, they vary each other physically and chemically so that different bird communities might be supported. keywords: bundala national park, lagoons, waterbirds 1 introduction wetlands support and maintain a diverse community of birds (duncan et al 1999). man has been aware of the link between birds and wetlands for thousands of years. coastal lagoons are important bird habitats since they occupy a diverse array of 34 factors affecting the avi-faunal distribution in the three lagoons (malala, embillakala and bundala lewaya) of bundala national park (a ramsar wetland) in sri lanka k. b. suneetha gunawickrama* and h. g. b. n. damayanthi http://www.ruh.ac.lk/rjs/rjs.html c h a n d a n a 1 , a m a r a s i n g h e 1 a n d s a m a ya w a r d h e n a r u h u n a j o u r n a l o f s c i e n c e 3 , p p . 3 4 4 3 ( 2 0 0 8 ) f a c t o r s a f f e c t i n g t h e a v i f a u n a l . . . microhabitats. these habitats are useful for birds for breeding, nesting, and rearing of young (acuna et al, 1994). many lagoonal wetlands are stopovers for migratory birds. declining number of wetland associated birds is partly attributed to the loss of wetlands (duncan et al 1999; mads et al 2002). consumption of the energy at the higher levels of food chain makes birds useful biological indicators (furners et al 1993). wetland bird communities are important biological indicators. comprehension of the aquatic bird diversity and abundance might be helpful in designing conservation strategies. there are many factors affecting the relationship between wetlands characteristics and birds. these include the availability of habitats and healthy water, food, shelter and predators. although the size of a wetland is often a crucial determinant of water bird richness and abundance (akihisa & satoshi 2001), trophic status and or shallowness are also major factors influencing water bird richness and abundance (suter, 1994). studies have shown the direct and indirect responses of water bird populations in response to changes in water quality (owino, 2001). similarly, biotic interactions between primary producers, aquatic invertebrates, fish and birds are gradually becoming better understood and appreciated with reference to the impact of trophic interactions on avian reproductive success (cynthia and william 2000). different waterfowl species adapt to different wetland types, inhabit different geographic areas, and nest at different times. in fact, the relationship between the aquatic birds and their habitats are complex. bundala national park (bnp) is located in deep south of sri lanka (figure 1). bnp was recognized as the first ramsar wetland in sri lanka and best known for the diverse and abundant water birds (bambaradeniya, 2001). five brackish water lagoons viz. maha lewaya, koholankala lewaya, malala lewaya, embillakala kalapuwa and bundala lewaya located in the bnp, occupy a total area of 2,250ha (benthem et al, 1993). this study was concentrated only on malala (650ha), embillakala (430ha) and bundala lewaya (520ha) because usually these three lagoons support diverse community of water birds. the five lagoons of the bundala wetland in sri lanka are known to be a seasonal home to large populations of migratory birds, visiting sri lanka from as far away as siberia (de silva & jakobsson, s., 1996). they find adequate supplies of food in bundala lagoons. the brackish water lagoons (mixed seawater and freshwater) serve as nurseries for shrimp, fish, 36 r u h u n a j o u r n a l o f s c i e n c e 3 , p p . 3 4 4 3 ( 2 0 0 8 ) c h a n d a n a 1 , a m a r a s i n g h e 1 a n d s a m a ya w a r d h e n a f a c t o r s a f f e c t i n g t h e a v i f a u n a l . . . and a variety of other marine life. the delicate ecological balance of at least two of these lagoons has been severely affected by the drainage of water flow from the kirindi oya irrigation and settlement project located upstream to the park (de silva & jakobsson, s., 1996). after the implementation of the project, the salinity of the lagoons has dropped due to inflow of upstream irrigation water (de silva & jakobsson, s., 1996). they also noted that this change in salinity levels has reduced the abundance of visiting water birds, as their food supply has diminished specially the reduced prawn fishery. they have also argued the risk of the conversion of malala and embillakala into freshwater lakes. de silva and jacobson (de silva & jakobsson, s., 1996) have recorded over eighty water bird species around these lagoons. in contrast to fish, the structure and functioning of aquatic bird communities are often poorly documented and understood (cynthia and william, 2000) and those of bundala lagoon systems are no exception. the relationship between water bird species composition and the habitat characters of malala, embillakala and bundala lewaya are poorly documented. therefore the objective of this study was to investigate aquatic birds inhabiting the lagoon systems with a special reference to several environmental parameters. figure 1: bundala national park and its major lagoons. 37 c h a n d a n a 1 , a m a r a s i n g h e 1 a n d s a m a ya w a r d h e n a r u h u n a j o u r n a l o f s c i e n c e 3 , p p . 3 4 4 3 ( 2 0 0 8 ) f a c t o r s a f f e c t i n g t h e a v i f a u n a l . . . 2. methods embillakala, malala, bundala lewaya were selected for this study since they are considered as very important habitat for birds (bambaradeniya, 2001). studies were carried out from december 2000 to december 2001. bird counting was carried out in regular intervals throughout the period. usually bird counts were taken 2-3 times per week during 07.00-11.00 hours each day as described elsewhere (shutler et al, 2000; taku, 2001). four monitoring sites were selected in each lagoon. bird counts were taken separately in each locality (figure 2) on the same day at the same time period. on the spot identifications were made using binoculars with 10x50 field 5.70 100mat 1000m, with the aid of standard picture guides (harrison, 1999). mean bird counts were computed for each month. water salinity was measured by titration method and computed by knudsen equation. primary productivity was measured by in situ by oxygen method (owen, 1974). mean water depth was estimated by measuring the depth of lagoon at thirty sites. for this purpose a graduated pole was used. scaled maps of each lagoon were drawn for the each month during the study period and the lagoon perimeter was estimated by cartometer method and area was estimated by rough calculation method (owen, 1974). figure 2: chosen sites (filled circles) for bird counting in the three lagoons of the bundala national park. 38 r u h u n a j o u r n a l o f s c i e n c e 3 , p p . 3 4 4 3 ( 2 0 0 8 ) c h a n d a n a 1 , a m a r a s i n g h e 1 a n d s a m a ya w a r d h e n a f a c t o r s a f f e c t i n g t h e a v i f a u n a l . . . number of birds counted for particular species for each sampling occasion was pooled and divided by the number of sampling occasions to obtain the monthly mean count (december 2000 to december 2001). shannon's diversity index (freitas & petrere, 2001)) and evenness were calculated for each lagoon. pearson correlation coefficient was calculated between monthly mean bird density and either the lagoon area, lagoon perimeter, water depth or salinity. overall abundance of bird species among lagoons was statistically compared. bird abundance at each identified localities (where abundant aquatic bird community was observed) was also statistically compared. statistical analysis was performed using spss 98 statistics software. 3. results dominant groups of birds recorded in the three lagoons are given in the table 1. based on the shannon's diversity index, overall data indicated that the bird diversity was highest in embillakala lagoon during the period of study. malala lagoon accommodated a moderate diversity of birds while bundala lewaya accommodated the least diverse bird community during the study period (table 2). the statistical comparisons of bird species densities of each lagoon revealed that pelicans, egrets, herons, painted storks preferred embillakala lagoon while cormorants, sand pipers and plovers preferred bundala lewaya lagoon. in contrast, the malala lagoon accommodated above bird groups in moderate densities (table 1). table 1: dominant groups of birds recorded in the three lagoons malala embillakala bundala lewaya pelicans egrets herons cormorants painted storks terns pelicans, egrets herons painted storks terns in moderate numbers cormorants sand pipers plovers terns this study has identified the sites of each lagoon where aquatic birds abundantly inhabit (figure 3). interestingly, aquatic weeds were abundant at the sites where aquatic 39 c h a n d a n a 1 , a m a r a s i n g h e 1 a n d s a m a ya w a r d h e n a r u h u n a j o u r n a l o f s c i e n c e 3 , p p . 3 4 4 3 ( 2 0 0 8 ) f a c t o r s a f f e c t i n g t h e a v i f a u n a l . . . birds were recorded abundantly. during the study period, however, mean gross primary productivity of malala lagoon and embillakala lagoon were 1.48 ± 0.3 mg/h/l and 1.2 ± 0.2 mg/h/l respectively while it was 0.08 ± 0.02 mg/h/l in bundala lagoon. low primary productivity of bundala lewaya might not have supported a rich bird community. figure 3: localities in the embillakala, malala and bundala lewaya where aquatic birds inhabit abundantly. lagoon water levels were drastically reduced in mid february 2001 (table 3) and eventually dried out at the beginning of august 2001. nevertheless, embillakala lagoon was completely dried out in september 2001. in march 2001 all three lagoons accommodated a low density of birds. nevertheless, all the three lagoons accommodated a high number of birds when water level was around 20 -40cm. highest and lowest salinity levels recorded in embillakala laggon, malala lagoon and bundala lewaya during the study period were 0.4-28 ppt, 0.71-48.3 ppt and 13.6-42.6 ppt respectively. embillakala lagoon accommodated more birds than other two lagoons at moderate depths (40-80 cm) with moderate salinity levels (1.8 ppt). this could be attributed to the enhancement of food material produced by aquatic weeds under moderate salinity levels (leck and brock, 2000). statistical analysis showed that (table 4) there was no significant correlation of bird density and lagoon perimeter, water depth. this may be due to the fact that birds show a specific distribution pattern rather than a uniform distribution. nevertheless, the bird density was significantly correlated with the salinity levels and lagoon area (table 4). 40 r u h u n a j o u r n a l o f s c i e n c e 3 , p p . 3 4 4 3 ( 2 0 0 8 ) c h a n d a n a 1 , a m a r a s i n g h e 1 a n d s a m a ya w a r d h e n a f a c t o r s a f f e c t i n g t h e a v i f a u n a l . . . table 3: monthly bird density, water depth, lagoon area and lagoonperimeter recorded in each of the three lagoons during study period m-malala lagoon, e-embillakala lagoon, bl-bundala lewaya, *-significant occurrences of birds were not detected since lagoons were almost dried out. table 4: significant levels of correlation coefficients computed for bird density vs lagoon perimeter, water depth and lagoon area. bird density of each month vs malala embillakala bundala lewaya water depth 0.256/0.128 0.486/0.243 0.074/0.037 area 0.04 0.03 0.02 lagoon perimeter 0.286/0.143 0.261/0.131 0.253/0.127 salinity 0.01 0.01 0.01 4. discussion birds are ideal to be used as an index of biotic integrity since their presence or absence tends to signal the health of several conditions that are keys to the proper functioning of an ecosystem. furthermore, this relationship is often associated with levels of human disturbance (cintia et al, 2001). in addition to providing an overall signal of ecosystem health, birds are also ideal because they are relatively easy to sample and their natural history is well described relative to other taxonomic groups in wetland ecosystems. 41 lagoon mean depth (cm) m e bl m e bl m e bl m e bl 600 450 450 152± 2.3 107±2.1 36±2.6 6 3.6 4 343 132 366 january 2001 600 450 450 152±2.6 108±2.6 37±2.9 6 3.6 4 320 361 213 february 350 300 400 48±3.1 102±2.3 38±3.6 3.2 2.1 4.2 60 83 89 march 350 250 300 46±2.3 76±2.1 30±3.2 3.2 2 3.6 10 4 21 april 350 200 200 46±2.6 48±2.6 30±2.5 3 2 3 183 378 164 may 300 200 100 42±2.8 48±2.3 29±2.5 3 2 3 254 440 141 june 250 150 100 38±3.0 48±2.3 20±2.6 3 2 3 160 477 110 july 100 100 50 28±2.3 50±3.2 2.5 1.6 3 53 1057 august* september* october* november 100 75 50 14±1.6 10±2.3 10±1.6 0.4 0.3 0.6 30 85 2 december 100 75 50 12±1.2 9±2.3 8±2.3 1 0.6 0.3 740 137 342 approximate lagoon area (ha) approximate lagoon perimeter (km) bird density (total count per month per visit x 10) december, 2000 c h a n d a n a 1 , a m a r a s i n g h e 1 a n d s a m a ya w a r d h e n a r u h u n a j o u r n a l o f s c i e n c e 3 , p p . 3 4 4 3 ( 2 0 0 8 ) f a c t o r s a f f e c t i n g t h e a v i f a u n a l . . . habitat heterogeneity usually increases with area (elmberg et al. 1994). the degree of community persistence is likely to depend on the physical and temporal stability of habitats and on the interactions between the species in the community (tejo et al, 2001; benston et al, 1997) based on, for example, foraging guilds, territory size, and winter range. a positive relationship between island area and species richness has been so widely documented that it comes close to being a universal ecological law. the relationship has played an important role in the development of ideas in population biology, community ecology and island biogeography. a previous study has shown that sandpipers were strictly sensitive to the size of the habitat for their breeding success (peter and malcolm, 1994). our results confirm the importance of habitat size in explaining the richness of aquatic birds within the wetland complex analyzed. our findings are also consistent with results of other studies in a variety of environments (sillén and solbreck 1977; brown and dinsmore 1986; opdam 1991; andrén 1994a; turner 1996). embillakala lagoon maintained the moderate water depth with more aquatic macrophytes and more associated grasslands when compared to malala lagoon and bundala lewaya. rich aquatic bird diversity in the embillakala lagoon can be partly explained by high habitat heterogeneity. bird populations are influenced by a variety of factors at both small and large scales that range from the presence of suitable nesting habitat, predators, and food supplies to climate conditions and land-use patterns (forcey et al, 2007). present study revealed that there is no uniformity in the bird distribution pattern in each lagoon. moreover, our data indicate that aquatic birds preferred specific localities in malala, embillakala and bundala lagoons for foraging and resting (figure 2). this observation indicates the existence of specific ecological relationship between birds and the lagoon environment. these sites retain a moderate water depth and abundant aquatic weed community. lagoon area in bundala water bodies might also influence species richness indirectly via its correlation with other factors that affect diversity directly. among the most plausible of such potentially confounding variables is habitat diversity, which is often presumed to increase in direct relation to island area (kohn & walsh, 1994). if larger islands supported greater habitat diversity as a result of greater topographic and geological heterogeneity, this increased habitat diversity might promote increased species richness, particularly if the species involved tended to be habitat specialists. habitat area influenced the habitat diversity in many instances (douglas & lake, 1994). several studies indicated the relationship between habitat parameters and aquatic bird density (davis, and smith, 1998; mark, 2006). our study indicates that salinity was negatively correlated with bird density. this phenomenon can be partly attributed to the changes of abundance of food items such as fishes and invertebrate tax in the lagoons with varying salinity levels. the bundala lagoon systems change continuously. habitat parameters and inhabiting aquatic fauna and flora also change. this might also influence the avi-fauna diversity and distribution. continuous monitoring of aquatic bird populations with habitat parameters is highly recommended. 42 r u h u n a j o u r n a l o f s c i e n c e 3 , p p . 3 4 4 3 ( 2 0 0 8 ) c h a n d a n a 1 , a m a r a s i n g h e 1 a n d s a m a ya w a r d h e n a f a c t o r s a f f e c t i n g t h e a v i f a u n a l . . . acknowledgements we would like to thank university grants commission for funding and department of zoology, faculty of science for providing the infra structure facilities. we also thank for the department of wild life conservation of sri lanka for granting the necesssary permission to carry out this study. references acuna, r., contras, f., and kerekes, j. (1994) aquatic bird densities in two coastal lagoon systems in chiapas state, mexico, a preliminary assessment. hydrobiologia: 279/280, 101-106 akihisa, h. & satoshi, m. (2001) habitat use and diversity of water birds in a coastal lagoon around lake biwa, japan. ecological research: 16, 543-553 andrén, h. 1994a. effects of habitat fragmentation on birds and mammals in landscapes withdifferent proportions of suitable habitat: a review. oikos 71: 355-366. bambaradeniya, c. n. b. (compiler) (2001). guide to bundala: a guide to the biodiversity of bundala national park a ramsar wetland in sri lanka. iucn, sri lanka. xiv +54pp. bengtsson, j., baillie s.r., & lawton, j. (1997) community variability increases with time. oikos: 78, 249-256 benthem, w., jeanes, k.w. & van lavieren, l.p. (1993) wetland site report and conservation and management plan-bundala national park. colombo. central environmental authority/euro consult. 84p brown, m. and j. j. dinsmore. 1986. implications of marsh size and isolation for marsh birdmanagement. journal of wildlife management 50: 392-397. cintia, c., sergio, a.n. & pablo, a.m. (2001) effects of human activity on the structure of coastal marine bird assemblages in central chile. conservation biology: 15 (5) 1396-1404 cynthia, a. & william, m.t. (2000) community concordance between the fish and aquatic birds of lakes in northern alberta, canada: the relative importance of environmental and biotic factors. fresh water biology: 43, 421-437 davis, c. a. & l. m. smith, 1998. ecology and management of migrant shorebirds in playa lakes region of texas. wildlife monographs 140: 1–45. de silva, r.i. & jakobsson, s. (1996). water bird survey in the bundala national park. in:ekarathne, s. et. al. (eds). workshop on research and management of coastal resources in the tropics. colombo. nara. (abstract) douglas, m. & lake, p.s. (1994) species richness of stream stones: an investigation of the mechanisms generating the species-area relationship. oikos: 69, 387-396 duncan, p., hewison, a. j. m., houte, s., rosoux, r., tournebize, t., dubs, f., burel, f. & bretagnolle,v. (1999) long-term changes in agricultural practices and wildfowling in an internationally important wetland, and their effects on the guild of wintering ducks. journal of applied ecology: 36 (1) 11-23 elmberg, j., p. nummi, h. pöysä and k. sjöberg. 1994. relationships between species number, lake size and resource diversity in assemblages of breeding waterfowl. journal of biogeography. 21: 75-84. freitas, c.e. c. & petrere, m. (2001) influence of artificial reefs on fish assemblage of the barra bonita reservoir (são paulo, brazil). lakes & reservoirs: research and management: 6 (4): 273-278 forcey, g.m., g.m. linz, w.e. thogmartin, and w.j. bleier (2007) influence of land use and climate on wetland breeding birds in the prairie pothole region of canada. can. j . zool. 85: 421-436 furners, r.w., greenwood, j.j.d. & jarvis p.j. (1993) in: i.n.r.w. furners and j.j.d. greenwood (eds) can birds be used to monitor the environment? birds as monitors f environmental change. chapman and hall, london. pp1-41. harrison, j. (1999) a field guide to the birds of sri lanka. oxford university press, new york.340p kohn, d.d. & walsh, d.m. (1994) plant species richness the effect of island size and habitat diversity. journal of ecology: 82, 367 377. leck, m.a. & brock, m.a. (2000) ecological and evolutionary trends in wetlands: evidence from seeds and seed banks in new south wales, australia and new jersey, usa. plant species biology: 15 (2), 97-112. mads, c., forchhammer, e.p. & nils, c.h.r.s. (2002) north atlantic oscillation timing of longand short-distance migration. journal of animal ecology: 71(6), 1002-1014 mark v. hoyer, sky k. notestein, thomas k. frazer & daniel e. canfield jr (2006) a comparison between aquatic birds of lakes and coastal rivers in florida. hydrobiologia 567:5–18 opdam, p. 1991. metapopulation theory and habitat fragmentation: a review of holartic breeding bird studies. landscape ecology 5: 93-106. 43 c h a n d a n a 1 , a m a r a s i n g h e 1 a n d s a m a ya w a r d h e n a r u h u n a j o u r n a l o f s c i e n c e 3 , p p . 3 4 4 3 ( 2 0 0 8 ) f a c t o r s a f f e c t i n g t h e a v i f a u n a l . . . owen, t. l. (1974) hand book of common methods in limnology. the c.v. mosby company, u.s.a.31p owino, a.o. (2001) pattern of variation in water birds numbers on four rift valley lakes in kenya, 1991-1999. hydrobiologia: 458, 45-53 peter, v.d., malcolm, l.h. & scott, m.m. (1994) effects of habitat area on the distribution of grassland birds in maine. conservation biology: 8 (4), 1087-1097 shutler, d., adele, m. & robert, g. c. (2000) bird communities of prairie uplands and wetlands in relation to farming practices in saskatchewan. conservation biology: 14 (5), 1441-1451. sillén, b. and b. solbreck. 1977. effects of area and habitat diversity on bird species richness in lakes. ornis scandinavica 8: 185-192. suter w. (1994) over wintering waterfowl on swiss lakes: how are abundance and species richness influenced by trophic status and lake morphology? hydrobiologia: 279/280, 1-14 taku, m. (2001) patterns of bird abundance and habitat use in rice fields of the kanto plain, central japan. ecological research: 16 (3), 569-585 teijo, v., veijo, k., antti, l., pirkko, s. & jukka, s. (2001) geographical patterns of species turnover in aquatic plant communities. freshwater biology: 46 (11), 1471-1478 turner, i. m. 1996. species loss in fragments of tropical rain forest: a review of the evidence. journal of applied ecology 33: 200-209. 44 ruhuna journal of science vol 10 (2): 108-119, dec 2019 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v10i2.77 university of ruhuna  faculty of science, university of ruhuna 108 sri lanka metal bioaccumulation and translocation studies of spinacea oleraceae and celosia argentea cultivated on contaminated soil amos l. ogunyebi1, ojuolape e. olojuola1, koleayo o. omoyajowo1, 2 and gbemi e. shodunmola1 1department of cell biology and genetics, university of lagos, akoka, nigeria 2department of science policy and innovation studies, national centre for technology management, victoria island, lagos, nigeria correspondence: logunyebi@unilag.edu.ng https://orcid.org/0000-0003-2315-470x received: 18th december 2018; revised: 16th august 2019; accepted: 07th november 2019 abstract the quest for tolerant plants with excellent phytoremediation potential is a stark reality. additionally, the use of dumpsite soil for growing vegetables and ornamental plants is relatively a common practice by farmers in urban cities across nigeria. hinged on these concerns, this study was poised to evaluate and compare the bioaccumulation factor (baf) and translocation factor (tf) of zn, cu, cr, pb and ni in two common vegetable species (spinacea oleraceae and celosia argentea) grown on the experimented olusosun dumpsite soil and the undisturbed sandy loam top soils of the university of lagos biological garden. the latter represents the control group. this study observed a considerable increase in metallic concentrations in the vegetable species grown on olusosun dumpsite soil in comparison to the control. the level of zn, cu and ni (except for pb and cr) were within the fao/who permissible limit. both vegetable species experimented on olusosun dumpsite soil have bafs <1 for zn, cu, cr, pb and ni implying that they never accumulated those metals in their tissues. likewise, both vegetable species have tf<1 in the order of zn>pb>ni>cu>cr for celosia argentia and zn>cu>cr>pb>ni for spinacea oleraceae. the use of dumpsite soil for growing vegetables have increased the levels of pb, cu, zn, ni, cd, and cr in their different parts and may further pose a serious threat to human health in the future if such practice continues. keywords: bioaccumulation factor, celosia argentea, dumpsite soil, metal accumulation, phytoremediation. 1 introduction the fate of heavy metals and other pollutants in the environment is considered not only as a critical aspect of ecotoxicology but also a global environmental http://doi.org/10.4038/rjs.v10i2.77 mailto:logunyebi@unilag.edu.ng https://orcid.org/0000-0003-2315-470x a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 109 issue seeking consensus attention with ‘bioaccumulation’ being discoursed within this context. indeed, anthropogenic activities have drastically altered nature’s biogeochemical cycles and the natural balance of earth’s fundamental elements making them more abundant in the ecosphere. heavy metals are natural elements that are cosmopolitan in distribution throughout the earth’s crust. these metals in their natural state are harmless and nontoxic until anthropogenic influence redistributes them thus making their proportion uneven and unwanted in their new environment. these anthropogenic activities include mining, metal processing, coal burning, refining, sewage disposal from industrial plant cooling process, metal recycling processes, domestic and agricultural use of metal containing compounds such as pesticides and fertilizers, mechanised farming and the indiscriminate disposal of wastes generated during these activities. in addition, when some of these activities interact with natural phenomenon such as rainfall, wind-action and so on, the upshot of these events could lead to metal corrosion and erosion, atmospheric deposition and metal evaporation which either increases the level of heavy metals or introduces their unwanted presence in the biosphere. heavy metals such as pb, cd, cr, ni, and cu can be toxic and naturally persistent. their uncontrolled release and interactions end up contaminating the food chain, posing a major concern to biodiversity and survival of people (morais et al. 2012). a case study of the minamata disease tragedy has been reported in japan, where mercury poisoning occurred in a small fishing town causing its inhabitant to experience neurological and congenital disorders which lead to their untimely death as a result of ingesting aquatic food contaminated with methyl-mercury (mehg) (otitoloju 2016). another notable case of metal pollution is the zamfara lead poisoning incident in northern nigeria, where lead-contaminated ore caused death causalities in human and livestock (otitoloju 2016). interestingly, some of the elements we consider as pollutants were once forms of useful substance that have either exceeded their lifespan or are found in an undesired environment and have since become household and commercial wastes. the injudicious disposal of these wastes and the ill-management of dumpsites have been recently considered as a public health concern in major towns and cities in nigeria as these wastes could elevate the level of metals in the soil (singh et al. 2004, mapanda et al. 2005). the introduction of heavy metals into the terrestrial ecosystem includes fertilizer and sewage discharge, metallic waste disposal, open air incineration and dumpsites (morais et al. 2012, adedokun et al. 2017, omoyajowo et al. 2017). the presence of open and unsafe dumpsites in most cities of nigeria has raised several public health concerns even as its urban population is also on the rise (bukar et al. 2012). the use of these dumpsites or their soil in agricultural purpose is also a common practice in lagos and other cities in nigeria perhaps because of the long-standing notion that a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 110 decomposed wastes increases soil fertility and enhance plant growth (ogunyemi et al. 2003). plant species have the natural proclivity to absorb nature’s elements into its tissues in a process termed bioaccumulation and breaks them down into simpler forms via a process called bio-utilization. however, the rate at which some of these species absorb these metals or pollutants is quite faster than the rate at which they catabolize them. moreover, studies have established that certain plant species accumulates contaminants at higher magnitudes than others. these plants are of immense importance to soil remediation projects and have aroused the interest of many researchers globally (hu et al. 2014, yashim et al. 2014). streit (1992) reported that bioaccumulation may encourage the presence and persistence of heavy metals in the ecosystem; due to its absorption by producers (plants) and ingestion by consumers (human and animals). further studies have shown that heavy metal pollution may have a long-term cumulative health effects when they are ingested, stored and transmitted along the food chain (opaluwa 2012, singh et al. 2010, abdulqadir et al. 2015). singh and kumar (2006) in another study posited that bioaccumulation efficiency and retention capacity vary among plant species and soil types. therefore, the metal retention capacity of soils could be reduced due to the incessant release of pollutants causing changes in ph, thus discharging toxic metals from the topsoil into the ground water or soil and making it available for phyto-absorption (ortiz and alcaniz 2006). however, a series of pollution studies have helped to establish the relationship between dumpsites, heavy metals, phyo-accumulation and associated impacts. a study carried out by bukar et al. (2012) observed a high concentration of cu in dumpsite soil when compared to the soil from a control site. another study suggested a positive relationship between the levels of cr and crude fat in fresh fruits implying that cr may probably influence the nutritive value of fresh fruits (omoyajowo et al. 2017). cd and pb are generally toxic to plant productivity; they reduce chlorophyll content and inhibit the growth of leaves, shoot and root and alter enzymatic activities (zeng et al. 2008, farooqi et al. 2009, lai et al. 2012). cd also inhibits respiration, photosynthesis, water, and nutrient uptake (kuoet al. 2006). in nigeria, bioaccumulation studies with respect to plants on dumpsite soil are relatively few. in addition, many urban farmers use this soil to grow edible and aesthetic plants perhaps because of the limited arable farmlands and poor soil condition dealt by urbanisation and climate change, or the perception that dumpsite soil are rich in nutrients and organic matter deposited by decayed and composted wastes which enhances soil fertility. the olusosun dumpsite is particularly one of the largest repositories of waste in sub-saharan africa. it is a beehive of economic activities for some poor persons and scrap scavengers dwelling in lagos. it is about a 100acre dumpsite located (6.591111ºn 3.381389ºe) in ojota, lagos, nigeria, and this ill managed site receives up to 10,000 tons of refuse daily. coincidentally, a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 111 this dumpsite is surrounded by commercial and residential areas with parks and boulevard that makes up edible, medicinal, and aesthetic plants with bioaccumulation tendencies for heavy metals and which when consumed can cause potential health detriments to humans. studies have consistently considered bioaccumulation efficiency/ phytoremediation potential of nonedible plants with less attention on edible plants, but it is important to understand the heavy metal accumulation potentials of edible plants as well. therefore, this study is poised to determine the bioaccumulation of heavy metals in two broad leafy vegetable crops spinacea oleraceae and celosia argentea grown on soil samples collected from olusosun dumpsite. 2 material and methods 2.1 study area this experiment was conducted in the biological garden of the faculty of science, university of lagos, nigeria (between 6.0310ºn, 3.02310ºe and 6.51667ºn, 3.38611ºe). the university biological garden is dedicated to the collection, field study, cultivation and aesthetic exhibition of wide range of plants and animals labeled with their scientific names. 2.2 sample location and sampling the olusosun dumpsite located in ojota, lagos, nigeria was used as a case study for this experiment. soil samples were collected from two spots or sites on the dumpsite. the sites where the samples were chosen were selected at random. for the control soil, undisturbed sandy loam topsoil was identified at the site of experiment. the soils were collected with a shovel at a depth of 5.0-8.0 cm and thoroughly homogenized. like the soil samples for the control, the collection was done by dividing the experimental site each into four quadrants; five soil samples were collected from each quadrant on a diagonal basis following the methods of nuonom et al. (2000). these soil samples were thoroughly sieved in order to make them fit for planting. 2.3 experimental procedure first, a small quantity of the dumpsite soil was analyzed to determine its metal content against an undisturbed soil representing the control. subsequently, the collected soil was used in planting the selected vegetables. spinacia oleraceae and celosia argentea were taken as representative plants for broad leafy vegetables. after seven (7) weeks of planting, whole plants from each replicate were harvested (uprooted) for metal analysis. a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 112 2.4 planting procedure triplicates of polythene bags filled with 1 kg of the dumpsite and control soil samples were collected and labeled appropriately. the seeds were sown directly into the soil 1 inch deep without the use of a nursery bed. this was done to ensure that the plants accumulate as much of the metals as possible. the seeds of both plants were treated the same way. this experiment was carried out in a screen house that shielded the plants from rainfall and pests. in other to simulate a natural condition with respect to good agricultural practice, soils were stirred per week with clean fingers to enhance aeration and to handle carefully the fragile roots of the vegetables. no chemical was used for controlling insects and weeds. this was done to avoid heavy metal inputs from such chemicals. insects were controlled by hand picking. watering was done five (5) times per week while distilled water was used to irrigate the plants. 2.5 digestion of soil sample soil samples were digested with a mixture of acids by open wet digestion method. well-mixed soil sample (20 g) in a digestion vessel was heated on a hot plate at 150-180 ºc. first, hnoз was added simply to remove all organic matter then hf and hclo4 were subsequently added. afterwards, hcl and distilled water were added to it in the ratio (1:3) to further dissolve the residues. the solution was heated on a bunsen burner severally until all reddish-yellow flames were expelled. the solution was brought down, allowed to cool and filtered into a 10ml standard flask and filled up to the mark with water and the digested sample was ready for analysis. 2.6. digestion of plant sample the collected plants were washed and rinsed with distilled water to ensure that they are thoroughly cleaned, and that all external contamination has been removed. plant samples were dried at room temperature for 1 week, pulverized and passed through a 2 mm stainless sieve. for digestion of plant tissue samples, a microwave digestion system was applied. 20g of plant tissue samples were precisely measured then hno3 and hydrogen peroxide were added at a ratio of 1:3. after ensuring a complete dissolution of plant tissues at 180ºc, the digests were quantitatively transferred into volumetric flasks and were ready for analysis. a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 113 2.7 analysis of metals the metallic concentrations zn, cu, cr, pb, ni and cd in the soil and plant samples were analyzed in triplicate using flame atomic absorption spectrometry (aas). 2.8 estimation of bioaccumulation factor, translocation factor and enrichment factor the bioaccumulation factor (baf) and the translocation factor (tf) were calculated to determine the degree of metal accumulation in the plants grown on soil samples collected from olusosun dumpsite. baf = concentration of metal in plant concentration of metal in soil baf < 1 indicates that the studied plants only absorbed metals but did not accumulate them (chopra and pathak 2012). tf = concentration of metal in plant shoot concentration of metal in plant root tf > 1 represent that translocation of metals was made effectively to the shoot from root (rezvani and zaefarian 2011). enrichment factor (ef) = concentration of metals in contaminated soil concentration of metals in uncontaminated soils ef was used to assess the degree of metal contamination in the understudied soil, using the following criteria (sutherland et al. 2000 as cited in likuku et al. 2013). ef<2 = minimal enrichment 2≤ ef<5 = moderate enrichment 5≤ef<20 = significant enrichment 20≤ ef<40 = very high enrichment and ef≥40= extremely high enrichment 2.9 data analysis the data obtained for the metallic concentration of soil and plants’ samples were analyzed using student t-test and analysis of variance on spss 22. a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 114 3 results and discussion this study assessed the metallic soil profile of the olusosun dumpsite, and the heavy metals accumulated by plants grown on this soil. the two different locations on the same dumpsite showed varying concentrations of pb, cd, cr, ni, and cu. the results of data analysis (paired two sample t-test) revealed no significant difference (p>0.05) in the metallic concentration of the five candidate metals for each vegetable grown on olusosun dumpsite soil. however, their concentrations were significantly higher in both species of vegetable samples grown on olusosun dumpsite soil in comparison to the control (table 1). cd was below detection limit in both sites (a and b) of olusosun dumpsite and even that of the control. it has been posited that since cd is structurally similar to zn, then plants may unlikely distinguish between the two ions (chaney et al. 1994 cited in yashim et al. 2017). table 1. heavy metal concentrations in vegetables planted (bdl: below detection limit; values are expressed as mean ± sem; fao/who 2002, adah et al. 2013) however, there is a significant range difference between the concentrations of zn observed in spinacea oleraceae grown on control soil site and the one grown on the dumpsite soil. the concentration of zinc ranged between 2.20 ± 0.001 mg/kg and 3.60 ± 0.002 mg/kg for all samples. the maximum allowed concentration of zn in edible plant is 5 mg/kg, therefore zn concentration present in both vegetables was observed to fall below the recommended values of fao/who. the permissible limit of cu in plants is 10 mg/kg as recommended by the who. in all the collected plant samples from both control and dumpsite soil cu concentration ranged between 0.98 ± 0.002 mg/kg and 2.08 ± 0.002 site vegetables zn (mg/kg) cu (mg/kg) cr (mg/kg) pb (mg/kg) ni (mg/kg) site a celosia argentia 3.01±0.003 1.94±0.003 1.92±0.002 2.89±0.002 0.40±0.002 spinacea oleraceae 3.60±0.002 2.08±0.002 1.98±0.003 3.00±0.002 0.90±0.002 site b celosia argentia 2.42±0.003 1.08±0.003 0.85±0.001 2.20±0.000 0.25±0.000 spinacea oleraceae 2.29±0.003 1.00±0.002 0.62±0.002 1.97±0.001 0.03±0.000 control celosia argentia 2.31±0.002 1.02±0.002 bdl 1.02±0.003 bdl spinacea oleraceae 2.20±0.001 0.98±0.002 bdl 1.01±0.002 bdl safe limits who (mg/kg) 60 30 0.3 2.0 - a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 115 mg/kg. cu concentration in both species of vegetables (spinacea oleraceae and celosia argentia) grown on both soils were found to be below the recommended values of fao/who. the permissible limit of cr for plants is 1.30 mg/kg, a value recommended by the who. comparing the mean concentration value of cr detected in spinacea oleraceae cr concentration ranged between 0.62 ± 0.002 mg/kg and 1.98 ± 0.003 mg/kg. although cr was detected in spinacea oleraceae grown on the control soil, its concentration was below the permissible limit while cr concentration spinacea oleraceae grown on olusosun dumpsite was higher than fao/who limit. the permissible limit of pb in plants recommended by the who is 2mg/kg. pb concentration detected in both spinacea oleraceae and celosia argentia grown on both control and dumpsite soil ranged between 1.01 ± 0.002 mg/kg and 3.00 ± 0.002 mg/kg. pb concentrations in both species of vegetables were found to be above the fao/who recommended limit. nickel has been considered an essential trace element for both humans and animals health. however, the maximum permissible limit of ni in plants as recommended by the who is 10 mg/kg. the concentration of ni in spinacea oleraceae grown on both soils ranged between 0.003 ± 0.000 and 0.90 ± 0.002 mg/kg. the concentration of ni detected was found to be below the fao/who recommended limit. the metallic concentrations recorded for celosia argentea in this study were lower than previous studies. adedokun et al. (2017) observed higher values for zn (18.8), cu (8.14) and ni (3.50) but pb (0.38), cr (0.38) were lower than the values observed in this present study. the differences in metallic concentrations in vegetables may be due to differences in their ability to absorb and accumulate metals. table 2: metal content of dumpsite soil and its control (bdl: below detection limit; mean values are expressed for metal content). the ef for olusosun dumpsite soil was found in the order cu>pb>zn for site a and zn/pb>cu for site b (table 2). the ef was maximum for cu (3.95) and minimum for zn (2.31) for site a. for site b, the ef was maximum for zn and pb (1.50) and minimum for cu (1.29) (table 2). ef values greater metal site a (mg/kg) site b (mg/kg) control (mg/kg) enrichment factor (ef) site a site b zn 24.25 15.75 10.50 2.31 1.50 cu 21.75 13.50 5.50 3.95 1.29 cr 75.00 25.00 bdl -- pb 125.00 75.00 50.00 2.5 1.50 ni 25.00 22.50 bdl -- cd bdl bdl bdl -- a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 116 than 1 indicate minimal enrichment while ef values greater than 40 indicate extremely high enrichment which connotes environmental pollution (singh et al., 2010, sutherland et al. 2000 as cited in likuku et al. 2013). according to sutherland classification, it was clear that site a of olusosun dumpsite soil was in moderate enrichment category for zn (2.31), cu (3.95), pb (2.50) while site b of olusosun dumpsite soil was in minimal enrichment category for zn (1.50), cu (1.29) and pb (1.50) as stated in table 2. soil properties such as ph, organic matter, cation exchange capacity (cec), redox potential, soil texture, and clay content may also be influenced the metal uptake as also purported by overesch et al. (2007). in both vegetables grown on soil samples collected from both sites (a and b) of olusosun dumpsite, baf values for zn (0.124-0.154), cu (0.07-0.096), cr (0.025-0.034), pb (0.023-0.029) and ni (0.001-0.036) were less than one (1) indicating that these plants only absorb metals but did not accumulate them. the baf values calculated for these vegetables were in order of zn>cu>cr>pb>ni for celosia argentia on both sites (a and b); zn>cu>ni>cr>pb for spinacea oleraceae on site a and zn>cu>pb>cr>ni on site b (table 3). table 3: bioaccumulation factor (baf) of vegetables planted on dumpsite soil metal bioaccumulation in the food chain can be highly dangerous to human health due to the possibility of metals being accumulated and transferred from lower organisms to higher organisms including humans (abdul-qadir et al. 2015). the tf is an important index that shows the mobility of metals in plants (rezvani and zaefarian 2011). tf follow an order of zn>pb>ni>cu>cr in celosia argentia and zn>cu>cr>pb>ni for spinacea oleraceae (table 4). table 4: translocation factor (tf) of metals in vegetables site vegetables zn cu cr pb ni a celosia argentia 0.124 0.089 0.026 0.023 0.016 spinacea oleraceae 0.148 0.096 0.026 0.024 0.036 b celosia argentia 0.154 0.080 0.034 0.029 0.011 spinacea oleraceae 0.145 0.074 0.025 0.026 0.001 metal celosia argentea spinacea oleraceae zn 0.163 0.167 cu 0.011 0.113 cr 0.003 0.025 pb 0.024 0.023 ni 0.015 0.021 a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 117 zn (0.167), cr (0.025) and ni (0.021) in spinacea oleraceaewas higher in comparison with celosia argentia. the values of the tf were very lesser than 1 for zn, cu, cr, pb and ni in both plants showing that translocation of metals was not effective (from root to shoot) in both plant species and as such cannot be considered as a phyto-remediating agent (mganga et al. 2011; rezvani and zaefarian 2011). these findings do not agree with a similar study assessing the tolerance level or metal accumulation efficiency of lycopersicon esculentum, rumex acetosa, and solanum melongena on soil collected near a metal-scrap dumpsite (yashim et al. 2014). nonetheless, the absorption and accumulation of metals in plant tissue depends may be influenced by temperature, nutrient availability among others (chopra and pathak 2012). 5 conclusions understanding the bioaccumulation potential of edible plants is very crucial from the food safety perspective as this may inform farmers, food safety and environmental agencies on how and where to grow edible plants with high bioaccumulation potential for zn, cu, cr, pb, ni and cd. furthermore, it may enrich our understanding about phytoremediation studies. though it is evident that levels of zn, cu, pb, cr and ni in the plant and soil samples from the experimented dumpsite were generally higher than those of their control counterparts but they still show low bioaccumulation potential. the study concluded that the use of dumpsite soil for growing crop vegetables increased the concentration of heavy metals (pb, cu, zn, ni and cr) in their tissues and such habit must not be encouraged. hence, farmers and the public must be sensitized on the potential environmental health implication of using dumpsite soil to grow vegetables and other crops. acknowledgements comments from three anonymous rjs reviewers on the initial draft of the submitted manuscript are acknowledged. references abdul-qadir a, qudsia k, ruqia b, balal y. 2015. accessing potential bioaccumulation of heavy metals in selective vegetables from gujranwala district. pakistan journal of environment and earth science 8: 58-70. adah ca, abah j, ubwa st, ekele s. 2013. soil availability and uptake of some heavy metals by three staple vegetables commonly cultivated along the south bank of river benue, makurdi, nigeria. international journal of environment and bioenergy 8(2): 56-67. adedokun ah, njoku kl, akinola mo, adesuyi aa, jolaoso ao. 2017. heavy metal content and the potential health risk assessment of some leafy vegetables cultivated in some a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 118 floodplains and farmlands in lagos, nigeria funai journal of science & technology 3(1): 30-47. alia n, sardar k, said m, salma k, sadia a, sadaf s, toqeer a, miklas s. 2015.toxicity and bioaccumulation of heavy metals on spinacia oleraceae grown in a controlled environment. international journal of environmental research and public health 12: 7400-7416. amir s, hafidi m, merlina g, revel jc. 2005. sequential extraction of heavy metals during composting of sewage sludge. chemosphere 59: 801-810. bukar l, hati s, dimari g, tijjani m. 2012.study of vertical migration of heavy metals in dumpsites soil. journal of science and technology 2(2): 50-55. chaney rl, green ce, filcheva e, brown sl. 1994. in: yashim zi, israel ok, hannatu m. 2014.a study of the uptake of heavy metals by plants near metal-scrap dumpsite in zaria, nigeria. journal of applied chemistry, article id 394650: 1-5. chopra ak, pathak c. 2012. bioaccumulation and translocation efficiency of heavy metals in vegetables grown on long-term wastewater irrigated soil near bindal river, dehradun. agric res. 1(2):157–164. fao/who 2002. list of maximum levels recommended for contaminants by the joint fao/who codex alimentarias commission. cac/fal.3: 18. farooq m, farooq a, umer r. 2008. appraisal of heavy metal contents in different vegetables grown in the vicinity of an industrial area. pakistan journal of botany 40(5): 99-106. farooqi zr, iqbal mz, kabir m, shafiq m. 2009. toxic effects of lead and cadmium on germination and seedling growth of albizia lebbeck benth. pakistan journal of botany 41:27–33. hook se, fisher ns. 2001. sublethal effects of silver in zooplankton: importance of exposure pathways and implication for toxicity testing. environmental toxicology and chemistry 20:568-574. hu y, wang d, wei l, zhang x, song b. 2014. bioaccumulation of heavy metals in plant leaves from yan׳an city of the loess plateau, china. ecotoxicology and environmental safety 110: 82-88. kuo cy, wong rh, lin jy, lai jc, lee h. 2006. accumulation of chromium and nickel metals in lung tumors from lung cancer patients in taiwan. journal of toxicology and environmental health 69:1337–1344. lai y, xu b, he l, lin m, cao l, wu y, mou s, he s. 2012. cadmium uptake by and translocation within rice (oryza sativa l.) seedlings as affected by iron plaque and fe2o3. pakistan journal of botany 44:1557–1561. magalhães dd, marques mr, baptista df, buss fd. 2015. metal bioavailability and toxicity in freshwaters. environmental chemistry letters 13(1):69-87. mapanda f, mangwayana en, nyamangara j, giller ke. 2005. the effect of long-term irrigation using wastewater on heavy metal contents of soils under vegetables in harare, zimbabwe. agric ecosyst environ 107:151–165 mganga n, manoko ml, rulangaranga zk. 2011. classification of plants according to their heavy metal content around north mara gold mine, tanzania: implication for phytoremediation. tanzanian journal of science 37: 109-119. morais s, costa fg, pereira ml. 2012. heavy metals and human health,environmental health emerging issues and practice, intech, retrievedfrom http://www.intechopen.com/ books/environmental on 18/09/2016. nuonom l, yemefack m, techienkwa m, njongang r. 2000. impact of natural fallow duration on cameron. nigerian journal of soil research 3: 52–57 ogunyemi s, awodoyin ro, opadeji t. 2003. urban agricultural production: heavy metal contamination of amaranthuscruenties l. grown on domestic refuse landfill soil in ibadan, nigeria. journal of agricultural science 15(2): 87-94. omoyajowo ko, njoku kl, babalola oo, adenekan o. 2017. nutritional composition and heavy metal content of selected fruits in nigeria. journal of agriculture and environment for international development 111 (1): 123-139. http://www.intechopen.com/%20books/environmental%20on%2018/09/2016 http://www.intechopen.com/%20books/environmental%20on%2018/09/2016 a.l. ogunyebi et al. metal accumulation in vegetables grown on dumpsite soil ruhuna journal of science vol 10(2): 108-119, december 2019 119 opaluwa od. 2012. assessment of heavy metals in water, fish and sediments from uke stream nassarawa state, nigeria. current world environment 7(2): 210-220. ortiz o, alcaniz jm. 2006. bioaccumulation of heavy metals in dactylis glomerate l. growing in a calcareous soil amended with sewage sludge. bioresource technology 97: 545–552. otitoloju a. a., 2016. today’s apple: perspective of an environmental toxicologist. 12th inaugural lecture, university of lagos. university of lagos press. 1-77pp. overesch m, rinklebe j, broll g, neue hu. 2007. metals and arsenic in soils and corresponding vegetation at central elbe river floodplains (germany). environmental pollution 145:800–812. rezvani m, zaefarian f. 2011. bioaccumulation and translocation factors of cadmium and lead in aeluropuslittoralis. australian journal of agricultural engineering, 2(4):114-119 sharma rk, agrawal m, marshall f. 2006. heavy metals contamination in vegetable grown in wastewater irrigated areas of varanasi, india. environmental contamination and toxicology 77: 312-318. singh kp, mohan d, sinha s, dalwani r. 2004 impact assessment of treated/untreated wastewater toxicants discharged by sewage treatment plants on health, agricultural, and environmental quality in the wastewater disposal area. chemosphere 55:227–255. singh r, kumar n, bhargava sk, barman sc.2010. accumulation and translocation of heavy metals in soil and plants from fly ash contaminated area. journal of environmental biology 31: 421-430. singh s, kumar m. 2006. heavy metal load of soil, water and vegetables in peri urban dehli. environmental monitoring and assessment 120: 79-91. streit b. 1992. bioaccumulation processes in ecosystems. experientia 48(10):955–970. sutherland ra, tolosa ca, tack fm, verloo mg. 2000. in: likuku as, mmolawa kb, gaboutloeloe gk. 2013. assessment of heavy metal enrichment and degree of contamination around the copper-nickel mine in the selebiphikwe region, eastern botswana. environment and ecology research 1(2): 32-40 yashim zi, israel ok, hannatu m. 2014. a study of the uptake of heavy metals by plants near metal-scrap dumpsite in zaria, nigeria. journal of applied chemistry, article id 394650 (1-5). zeng x, li l, mei x. 2008. heavy metals content in chinese vegetable plantation land soils and related source analysis. agricultural sciences in china 7:1115–1126. https://link.springer.com/journal/18 ruhuna journal of science vol 10 (2): 149-160, december 2019 eissn: 2536-8400 ©faculty of science doi: http://doi.org/10.4038/rjs.v10i2.80 university of ruhuna © faculty of science, university of ruhuna 149 sri lanka short paper determinants of watermelon production and its impact on the farmers in ifelodun l.g.a, kwara state, nigeria afolabi monisola tunde department of geography & environmental management, university of ilorin, nigeria. correspondence: folamoni70@yahoo.com; https://orcid.org/ 0000-0003-3326-4064 received: 8th july 2019; revised: 12th november 2019; accepted: 11th december 2019 abstract this paper assesses the factors that encouraged watermelon production and their impact on the growers in a rural part of ifelodun l.g.a, kwara state, nigeria. it further analyzes factors affecting the production and evaluates the determinants of rural farmers for diversification into watermelon production, impacts of growing watermelon on the livelihood of the producers and the constraints involved in its production. purposive sampling technique was used and a total of one hundred and six (106) farmers involved in producing watermelon in the study area were selected. focus group discussions (fgds) were used to complement the questionnaire. the study employs likert rating scale, descriptive and inferential statistics to analyze collected data. the results revealed that nature of soil and amount of water/ rainfall with a coefficient of determination of 0.779 and 0.859 respectively, are the most important factors affecting the production of watermelon. it was further discovered that accessibility to fertile land (x=3.9) and income generation (x=3.7) are the most important determinants for diversification into watermelon production by farmers. it revealed that watermelon production has significantly contributed to the income, health and livelihood of the people involved in it. hence, the practice of horticulture should be encouraged more in other rural local governments of kwara state for sustainable rural development. proper institutional support towards the development and growth of watermelon should be put forward by policy makers. keywords: horticulture, income generation, rural areas, watermelon. 1 introduction agriculture is an indispensable sector in nigeria despite its heavy reliance on oil. studies have shown that 70% of nigerians live in rural areas (egbe 2014), and their primary occupation is farming. majority of the farmers in the rural mailto:folamoni70@yahoo.com https://orcid.org/%200000-0003-3326-4064 a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 150 areas are small holders who depend largely on simple implements for crop production (dan-azumi 2011), and they grow different kinds of vegetables and fruits, including watermelon. the important roles played by fruits in the health and general well-being of man cannot be over emphasized. fruits provide important vitamins to the human body while contributing significantly to the livelihood of farmers involved in producing fruits. watermelon belongs to the family cucurbitaceae, which includes cucumbers, muskmelons, squash, pumpkins and gourds (schippers 2000). it occupies the third position among the world’s highly cultivated crops, and china is known to be the current highest producer of watermelon worldwide (dhaliwal 2017). the fruit thrives very well on fertile sandy soil particularly on sandy riverbanks and requires little maintenance once it sprouts. according to zafour (2007), watermelons are grown both in the tropics and subtropics, although they do best in the hot drier areas with abundance of sunshine. consumers have realized that watermelon is not just a summer fruit due to its health benefits (baxley 2018). for instance, agricultural marketing resource center (2018) reported that one cup of diced, fresh watermelon provides 21% of the daily requirement of vitamin c, 18% of vitamin a, significant levels of vitamin b6, lycopene, antioxidants and amino acids. in nigeria as well as in other parts of the world, watermelon is highly relished as a fresh fruit preferably for thirst-quenching attribute in addition to the many other identified nutritional values. in a study by oguntola (2006), among five other exotic vegetables examined in ibadan city of oyo state, nigeria, watermelon was reported as the most preferred. it consisted of 93% water and little amounts of fat, protein, minerals and vitamins (namdari et al. 2011). it is a major fruit in ifelodun local government area of kwara state, nigeria. mohammed (2011) confirmed this by reporting that an average yield of 57.70 kg/hectare for sole melon was obtained on a field in ifelodun local government area kwara state. the fruit and the seeds are used as a domestic remedy for urinary tract infection, control of blood pressure and possibly prevent stroke, and also can treat eye problems, dry skin, eczema, cancer, hepatic congestion, catarrh and psoriasis (adekunle et al. 2007, enukainure et al. 2010, inuwa et al. 2011). dauda et al. (2008) reported that despite all these nutritional benefits from watermelon, the production of the fruit is still low in nigeria. it is against this background that some farmers are diversifying into watermelon production in some parts of the country. this study therefore assesses the factors that encouraged watermelon production and its impact on the growers in a rural part of ifelodun l.g.a, kwara state. according to david livingstone (a missionary explorer), watermelon originated in the kalahari desert and semi-tropical regions of africa in the 1850s. presently, watermelons are cultivated as an important source of water during dry periods in these areas. the food and agriculture organization (2003) reported that watermelon is one of the most widely cultivated crops in the world with a global production of about 89.9 million metric tons. a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 151 according to the national agricultural statistics service, the united states cultivated 50,990.4 hectares of watermelons with a value of $460 million in 2009. as a traditional food plant (gyulai et al. 2011), watermelon can improve nutrition, supplement food security, promote rural development and support sustainable land cares (peet 1995; national research council 2008). according to mohammed (2011), seedless watermelon is grown by small holder rural farmers and usually inter-planted with crops like maize, yam, cassava, pepper to make maximum use of the land resources and increase returns from the production systems. it has an average temperature requirement of between 65°f and 95°f and requires a constant supply of moisture during the growing season. ufoegbune et al. (2014) compared crop parameters between wet and dry season in abeokuta, nigeria, and has recommended that watermelon should be planted in dry season with application of irrigation and in the wet season with supplemental irrigation. adekunle et al. (2007) noted that watermelon thrives well in the northern part of nigeria where suitable drier savanna agro ecology is found, and hence, it is the highest producing area. essentially, based on the benefits derived from watermelon, the fruit is currently grown on all continents throughout the warm regions of the globe including kwara state, nigeria. recently, it was discovered that a group of farmers migrated to oro-ago and ago-olomo communities to farm and specifically grow watermelon. the question now is what could be responsible for this action? it was against this development that the present study aimed to examine perception on the determinants of watermelon production and their impact on the growers in a rural ifelodun l.g.a, kwara state. specifically, the study examines the demographic characteristics of the farmers growing watermelon, identifies factors affecting the production of watermelon, evaluates the determinants of rural farmers’ diversification to watermelon production, determines the impact of growth of watermelon on the livelihood of the producers and assesses the constraints faced by farmers in watermelon production. 2 material and methods 2.1 study area ifelodun local government area (figure 1) of kwara state is the study area. it is the largest local government area in kwara state and is located within latitudes 8024”41 and 8025”44n of the equator and between longitudes 04035”35 and 4035”02e of the east of the greenwich meridian. it has nine districts and had a population of 206,042 as at 2006 census (npc, 2006) with an area of 3,435 km2. the annual rainfall ranges between 1000 and 1500 mm. the maximum average temperature ranges between 300 ºc and 350 ºc and a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 152 humidity ranges from 35 to 60% (mohammed 2011). the major source of livelihood and occupation of the people in the area is farming. farming is traditional in nature with emphasis on the cultivation of crops such as sorghum, cassava, yam, maize and melon (mohammed 2008). cultivation of watermelon growth is very recent in ifelodun, kwara state. productivity differences over time and farming types can result from a variety of factors including variations in scale or level of production. fig. 1: kwara state showing ifelodun local government area (inset is nigeria showing kwara state) (source: kwara state ministry of lands, 2017) 2.2 methods a multistage sampling technique was employed in the study. the first stage involved purposive selection of the local government area (ifelodun) where watermelon is grown on commercial basis. the second stage was the purposive selection of the three villages in the local government area based on accessibility and the quantity of watermelon that are grown. hence, all the one hundred and six (106) farmers involved in the production of watermelon in the study area were sampled by using copies of questionnaire. focus group discussions (fgds) were also organized for the farmers to complement the administered questionnaire. simple parametric and non-parametric analyses were used to analyze the data gathered. to identify the factors affecting watermelon production, multiple regression analysis was employed. likert a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 153 rating scale and matrix ranking were used to evaluate the determinants of rural farmers’ diversification into watermelon production, the impact incomes generated from watermelon has on the livelihood of the farmers, and to assess the constraints facing farmers in watermelon farming respectively. 3 results and discussion 3.1 demographic characteristics of the farmers majority of the farmers sampled were males. this is due to the traditional pattern that the male is dominant in the farming system in that area and because migrant farmers are mostly of male gender. out of 106 farmers sampled, majority (95%) were within the ages of 21-50 years, married and non-educated. the age signifies that most of the farmers are in the economically active age and therefore considered to be active and productive. this corroborates the findings of yusuf et al. (2013) that most farmers are within the age bracket of less than thirty to fifty years. most of them are married, meaning they have more mouths to feed and more hands to assist on farm. average household size is 6; this could also mean an additional hand to help on farm. all of them are full-time farmers with over 5 years farming experience. average annual income realized was ₦454, 307.69 in the study area. the finding agrees with ebiwoei (2013) who reported that marketing of watermelon in the niger delta area of nigeria is highly profitable. most of them (90%) are migrant farmers who migrated from different parts of the country to embark on farming at various locations within ifelodun local government area. this is as a result of the presence of fertile land for growing of the fruit. 3.2 factors affecting production of watermelon to determine perception of sampled farmers on the most significant factor affecting production of watermelon, seven variables were subjected to stepwise multiple regressions. the dependent variable is measured by the annual yield of watermelon. the independent variables are x1-x9. the formula is written as: y = a +b1x1 + b2x2 + b3x3 ……….. bn xn + e where, y = yield of watermelon a = intercept e = error terms b1x1 to bnxn = regression coefficients of (1-n) independent variables x1 = nature of soil x2 = solar radiation a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 154 x3 = amount of water/rainfall x4 = accessibility to fertile land x5 = availability of fertile land x6 = intensity of rainfall x7 =duration of rainfall x8 = number of rainy days x9 = type of seedling these variables were chosen based on farmers’ listing and reconnaissance survey carried out in the study area. the surrogate measure for each is percentage ranking by respondents in order of importance. table 1. regression analysis on factors affecting watermelon production variables r r2 % contribution adjusted r std error of the estimate nature of soil (x1) 0.883 0.779 77.9 0.538 5.91 amount of water/ rainfall (x3) 0.927 0.859 8 0.673 52.18 source: researcher’s fieldwork/computer output, 2018 table 1 shows the stepwise multiple regression analysis performed for factors affecting watermelon production in the study area. out of the nine variables that were fed into the model, only two variables x1 (nature of soil) and x3 (amount of water/ rainfall) were found to be significant at the specified tolerance level of 5% for entry into the model. x1 is the best predictor among the factors affecting watermelon production in the study area with a coefficient of determination of 0.779. this indicates that 78% of the variance of nature of soil is associated with watermelon production in the study area. the implication of this is that watermelon thrives well in well-drained loamy soil which is peculiar to the study area. x3 (amount of water/ rainfall) appeared to be the next most important factor that determines the growth of watermelon. it has a coefficient of determination of 0.859, meaning about 9% additional explanation of its variance is associated with variation in watermelon production in the study area. this means that amount of water required for the growth of watermelon should be regulated as this determines how well watermelon grows. it suggests that 85% of the joint variance in watermelon production is explained by variables x1 and x3. this is in accordance with the study by adojutelegan et al. (2015) that rainfall, transportation, storage of watermelon, market price of watermelon and prevalence of pest and watermelon diseases were the factors that affect production of watermelon in ekiti state, nigeria. the remaining seven variables x2, x4, x5, x6, x7, x8 and x9 were not meaningful in explaining their variation because their correlation coefficients are too low. to complement the information obtained through the questionnaire, two sessions a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 155 of focus group discussions (fgds) were conducted to explore the opinion of the farmers on their level of understanding of factors affecting production of watermelon in the study area. according to them, the type of soil found in the study area is a significant factor that assisted the yield of watermelon. they considered all other factors to be secondary. 3.3 determinants of rural farmers’ diversification into watermelon production to determine the factors responsible for rural farmers’ involvement in watermelon production, the use of likert rating scale was employed. table 2 reveals the factors necessitating farmers’ diversification to watermelon production. table 2. determinants of rural farmers’ involvement in watermelon production (% within parentheses). determinants/ reasons strongly agree agree disagree strongly disagree undecided weighted mean (x) rank income generated 91 (85.8) 9 (8.5) 1 (0.9) 0 5 (4.8) 3.7 2 nutritional benefits / food security 61 (57.5) 20 (18.9) 10 (9.4) 2 (1.9) 13 (12.3) 3.1 5 unemployment 66 (62.3) 10 (9.4) 20 (18.9) 10 (9.4) 0 3.2 4 accessibility to fertile land 96 (90.6) 8 (7.5) 0 0 2 (1.9) 3.9 1 accessibility to credit facilities 6 (5.7) 10 (9.4) 30 (28.3) 60 (56.6) 0 2.1 9 culture 51 (48.1) 20 (18.9) 25 (23.5) 10 (9.5) 0 3.0 6 educational status 26 (24.5) 20 (18.9) 30 (28.3) 25 (23.5) 5 (4.8) 2.3 7 increase in prices of food items 11 (10.4) 15 (14.3) 40 (38.1) 30 (28.6) 10 (9.5) 1.9 10 type of farm tasks performed 3 (2.8) 10 (9.4) 38 (35.9) 55 (51.9) 0 1.6 11 interest/ hobby 26 (24.5) 18 (17) 22 (20.8) 40 (35.8) 0 2.2 8 skills/ experience 66 (62.3) 35 (33) 3 (2.8) 2 (1.9) 0 3.6 3 source: researcher’s own fieldwork, 2018 from table 2, accessibility to fertile land (x=3.9) to produce watermelon is the most important determinant of rural farmers’ diversification to a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 156 watermelon production in the study area. this result is similar to that of adojutelegan et al. (2015) who reported that improved soil fertility and income generation are the reasons for farmers’ cultivation of watermelon in ekiti state, nigeria. income generation (mean of 3.7) was ranked second motivating farmers’ involvement in watermelon production. income generated from the sales of harvested watermelon can be used to cater for other expenses such as purchase of other food items, household items, acquire assets, expand the farm and pay school fees among others. this result is in support of ibeawuchi et al. (2015) that fruits and vegetables contribute to the income of both the rural and urban dwellers. skills/experience is the third ranked determinant (mean of 3.6), where the technical knowledge accrued to these farmers in growing watermelon has encouraged them into the business. unemployment (mean of 3.2) is the fourth ranked reason for farmers’ involvement in watermelon production. other reasons including nutritional benefits and culture were also important among others. table 3. perception of impact of income from watermelon on the livelihood of the producers (% within parentheses). impact strongly agree agree disagree strongly disagree undecided weighted mean (x) rank assist in the expenditure on household items 87 (82.1) 13 (12.3) 6 (5.6) 0 0 3.8 2 acquire assets 70 (66) 15 (14.2) 15 (14.2) 5 (4.7) 1 (0.9) 3.4 3 expenditure on food items 90 (84.9) 16 (15.1) 0 0 0 3.9 1 pay children’ school fees 30 (28.3) 5 (4.7) 50 (47.1) 15 (14.2) 6 (5.6) 2.4 5 savings/thrift 6 (5.6) 10 (9.4) 36 (34) 0 54 (51) 1.2 9 loan/debt servicing 2 (1.9) 5 (4.7) 45 (42.4) 54 (51) 0 1.6 8 increase farm size 2 (1.9) 8 (7.5) 70 (66) 26 (24.6) 0 1.9 6 accessibility to fresh nutritious fruit 56 (52.8) 30 (28.3) 20 (18.9) 0 0 3.3 4 purchase farm input 25 (23.6) 10 (9.4) 25 (23.6) 16 (15.1) 30 (28.3) 1.8 7 source: researcher’s own fieldwork, 2018 3.4 impact of watermelon on the livelihood of the producers a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 157 in order to determine the perception of the impact of watermelon production on the livelihood of its producers, some variables were considered with the use of likert rating scale as revealed in table 3. expenditure on food items (mean of 3.9) was ranked first as most important impact that sales of watermelon have on the growers. this means income realized from watermelon is being used to purchase food items for the growers’ households majorly. the second important impact that the income realized from the sales of watermelon has on farmers is that it assists in the area of expenditure on household items (x=3.8). this indicates that income realized from the sales of watermelon is also used to purchase household items. the third important impact through income realized from sales of watermelon is to acquire assets with a mean value of 3.4 (such as radio, television, land, grinding machine, and bike among others) and accessibility to fresh nutritious food has a mean value of 3.3. table 4. respondents’ ranking of challenges confronting watermelon production. problems respondents / ranks a 1 2 3 4 5 6 7 8 9 10 11 12 13 1. inadequate capital/credit facilities 50 42 13 1 2. lack of access to fertile land 37 21 38 4 1 4 1 3. pest & diseases 1 4 7 18 26 16 13 9 12 4. lack of technology 17 35 27 13 9 5 5. lack of extension services 4 1 10 4 4 9 43 31 6. marketing problem 9 15 12 13 13 26 6 4 8 7. poor road condition/ inadequate transportation 4 40 39 10 9 4 8. lack of farm input/seed procurement 8 12 39 29 9 9 9. poor seasonal rainfall 9 4 19 32 19 23 10. lack of education 1 27 13 36 12 17 11. middlemen problem 13 18 30 26 5 5 9 12. perishability/lack of storage facilities 15 42 6 14 9 16 4 13. pilfering 13 2 18 16 25 12 14 6 source: researcher’s own fieldwork, 2018 a ranks are given from 1 to 13; rank 1 is considered the most severe constraint to watermelon farmers while rank 13 is the least severe constraint. a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 158 3.4 constraints faced by farmers in watermelon production when asked about the challenges faced by the watermelon farmers in the study area, (table 4) a total of 92 sampled farmers ranked inadequate capital/credit facilities for the production of watermelon as the first or the second most pressing problem constraining their increased production. this agrees with the findings of adeoye et al. (2011) and ajewole (2015) that inadequate credit is a major constraint to watermelon production in nigeria. this has a big influence on the increase in production since capital is needed to expand the farm and increase production. a total of 58 sampled farmers ranked lack of access to adequate fertile land as first or second most pressing problem constraining their production in the study area. the aforementioned could be linked with the land tenure system available in the study area since most of these watermelon farmers are outsiders; they must rent the land occupied. the fear that they can be chased away at any time by the owner of the land has really constrained their expansion. fifty-seven sampled farmers ranked perishability or lack of storage facilities as the first and second most pressing problems. it corroborates the findings of ekerete and asa (2014) that the most severe constraints to watermelon marketing in uyo metropolis, akwa ibom state, nigeria include losses resulting from fruits spoilage, lack of preservation facilities, high cost of transportation, and lack of credit facilities. furthermore, 44 sampled farmers ranked poor road condition/inadequate transportation as the first, second or third most pressing problems. by nature, watermelon fruits are heavy, and if the roads leading to where they are being produced are bad, transporting them to the destination will be a very serious problem. according to the sampled farmers, they said the roads that lead to their different farms are very bad and as a result, it has affected the cost of transportation. they also reported that they do not have access to extension services at all in the area. other problems confronting watermelon farmers in the study area but were ranked differently include lack of technical knowhow, lack of extension services, pests and diseases, pilfering, poor seasonal rainfall, marketing problems, middlemen problem, lack of farm input and lack of education. this corroborates the report by adojutelegan et al. (2015) that lack of capital, poor markets, perishable nature of watermelon and lack of technology for large scale production were the serious constraints faced by farmers during the cause of production of watermelon. 4 conclusions from the study, it was discovered that most farmers diversify into watermelon production in the study area because of the type of soil found in the area a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 159 which is loamy, and the income being realized from their sales. it can therefore be concluded that watermelon production has really contributed significantly to the income, health and livelihood of the people involved in it. hence, the practice of horticulture should be encouraged more in rural areas of nigeria for sustainable rural development. what happened in ifelodun local government, kwara state could be a minor thing if farmers are given adequate attention, as this could be replicated in other parts of the country for sustainable agricultural development. proper institutional framework towards the development and growth of watermelon should therefore be put forward by policy makers. farmers should equally be encouraged more in the production of watermelon by providing them with necessary farm input to realize increased production. acknowledgements comments from two anonymous rjs reviewers on the initial draft of the submitted manuscript are acknowledged. references adekunle aa, fatunbi ao, adisa s, adeyemi oa. 2007. growing watermelon commercially in nigeria: an illustrated guide. usaid ics-nigeria and iita. retrieved october 5, 2012,from http://www.fao.org/sd/erp/toolkit/books/watermelon_illust_guidebook.pdf adeoye ib, olajide-taiwo fb, adebisi-adelani o, usman, jm. badmus ma2011. economic analysis of watermelon-based production system in oyo state, nigeria. arpn journal of agricultural and biological science 6(7): 53-59. adojutelegan ot, adereti fo, makanju ts, olorunfemi od 2015. analysis of factors affecting watermelon in ekiti state, nigeria. science, technology and arts research journal 4(2):324-329. agricultural marketing resource center 2018. watermelon. usda office. washington d.c. https://www.agmrc.org. ajewole oc 2015. income and factor analysis of watermelon production in ekiti state, nigeria. journal of economics and sustainable development 6(2):67-71. baxley r. 2018. farming 2018: watermelon no longer just a summer fruit due to its health benefits. the times and democrat. 1010 broughton street orangeburg, sc. https://www.agmrc.org>vegetables>watermelon dan-azumi jj. 2011. agricultural sustainability of smallholder floodplain agricultural systems: a case study of fadama areas in north-central nigeria. a ph.d. thesis submitted to development planning unit/bartlett faculty of the built environment, university college, london. http://discovery.ucl.ac.uk/id/eprint/138073. dhaliwal ms. 2017. cucurbits. handbook of vegetable crops 3rd edition. published by kalyani publishers. https://wwwresearchgate/publication/313098886. dauda sn, ajayi fa, ndor e, 2008. growth and yield of watermelon as affected by poultry manure application. electronic journal of environment, agriculture and food chemistry 7 (7):3108-3114. egbe ej 2014. rural community and development in nigeria. an assessment. osr journal of humanities and social science (iosr-jhss)19 (3):104-112. http://www.fao.org/sd/erp/toolkit/books/watermelon_illust_guidebook.pdf a.m. tunde watermelon production in kwara state, nigeria ruhuna journal of science vol 10(2): 149-160, december 2019 160 ebiwoei kp 2013. marketing margin and determinants of net return of watermelon marketing in yenagoa metropolis of bayelsa state, nigeria. journal of experimental biology and agricultural science 1 (4): 241-247. ekerete bi, asa ua 2014. constraints to watermelon marketing in uyo metropolis of akwa ibom state, nigeria. journal of agriculture and environmental sciences 3 (4): 63-69. doi: 10.15640/jaes.v3n4a5 url: http://dx.doi.org/10.15640/jaes.v3n4a5. enukainure ol, oke ov, daramola ao, adenekan so, umanhonlem ee. 2010. improvement of biochemical properties of watermelon rinds subjected to s. cerevisiae solid media fermentation. pakistan journal of nutrition 9(8):806–809. doi:10.3923/pjn.2010.806.809. fao 2003. world agriculture: towards 2015/2030. summary report, rome. eds bruinsma, journal. earthscan publications limited. https://www.fao.org. gyulai g, toth z, bittsanszky a. 2011. medieval citrullus dnas-unlocking domestication events (13th and 15th cent). in: plant archaeogenetics. edited by g gyulai. chapter 7. nova science publisher inc., new york, usa. ibeawuchi ii, okoli na, alagba ra, ofor mo, emma-okafor lc, peter-onoh ca, obiefuna jc. 2015. fruit and vegetable crop production in nigeria: the gains, challenges and the way forward. journal of biology, agriculture and healthcare 5(2): 194-209. inuwa hm, aina vo, gabi b, aimola i, thompson v. 2011. determination of differences in nutrient composition of citrullus vulgaries (watermelon) fruits after plucking. british journal of dairy science 2(2):27–30. mohammed bt. 2008. economics analysis of melon (citritillus lanatus) production in kwara state. proceeding of 2nd annual conference agric society of nigeria (asn) held on 19th– 23rd october at abakaliki. 57-58pp. mohammed bt. 2011. socio-economic analysis of melon production in ifelodun local government area, kwara state, nigeria. journal of development and agricultural economics 3(8):362-367. namdari m, mohammedi a, mobtaker hg. 2011. assessment of energy requirements and sensitivity analysis of inputs for watermelon production in iran. international journal of plant, animal and environmental sciences 1(1). national population commission (npc) 2006. population census report. national population commission census 2006. nigeria. national research council 2008. lost crops of africa. volume iii: fruits, washington, d.c. the national academies press. oguntola s. 2006. watermelon: hidden gem yet to be discovered. nigerian tribune thursday 13th july 2006. in a. ya‘u, b. a. mahmoud, & h. kubura (ed.), profitability of water melon production and marketing in kirfi local government area of bauchi state, nigeria. proceedings of the 25th farm management association of nigeria (faman) conference, 5th 8th september 2011, akure, nigeria. peet m. 1995. sustainable practice for vegetable production in the south, ncsu. schippers rr. 2000. african indigenous vegetable, an overview of the cultivated species, n.r/aco, eu. chatthan, uk. 56-60pp ufoegbune gc, fadipe oa, bello nj, eruola, ao, makinde aa. 2014. growth and development of watermelon in response to seasonal variation of rainfall. journal of climatology & weather forecasting 2:117. doi:10.4172/2332-2594.1000117. yusuf sfg, lategan fs, ayinde ia 2013 profitability and adoption of watermelon technologies by farmers in moro local government of kwara state, nigeria. journal of agricultural science 5(5):91-99. doi:10.5539/jas.v5n5p91. zafour 2007. report on production and marketing system of watermelon in selected areas of patuakhali district. department of agricultural economics and rural sociology, faculty of business administration and management, patuakhali science and technology university, bangladesh. http://dx.doi.org/10.15640/jaes.v3n4a5 http://www.cabdirect.org/abstracts/20033056815.html;jsessionid=56701f6bb1af3ba9c68fe9d4f3ffff39 https://www.fao.org/ ruhuna journal of science vol 11 (2): 118-130, december 2020 eissn: 2536-8400 © faculty of science doi: http://doi.org/10.4038/rjs.v11i2.91 university of ruhuna © faculty of science, university of ruhuna 118 sri lanka modelling non-life insurance in sri lanka using cox hazard model and classification of risky customers w.a.r. de mel and w.a.p.a. chathurangani department of mathematics, university of ruhuna, wallamadama, matara 81000, sri lanka correspondence: piumiayodyachathurangani@gmail.com; https://orcid.org/0000-0001-8296-9218 received: 20th june 2020, revised: 25th december 2020, accepted: 30th december 2020 abstract. some of the major factors that help the decision-making process of an insurance company include time of the first claim (tfc), claim size and the frequency of claims. however, in most situations researchers focus mainly on the second and third factors mentioned above. we hypothesize the importance of the tfc of an insurance contract in the decision-making process. empirical evidence of motor vehicle insurance data in sri lanka suggests that nine covariates are responsible for the claim sizes. in the current study, our main objective is to find the key factors of those nine that are responsible for the tfc of the insurance contract. this study is based on the claim data in the whole year of 2016 of non-life insurance policies of a particular insurance company in sri lanka. considering the tfc as rightcensored data, selected nonparametric methods, i.e., kaplan-meier, nelsonaalen estimators, and cox proportional hazard model are used to analyze the data. we identified the five most influential covariates namely, vehicle type, log of premium value and that of assured sum, the lease type and the age range via fitting the cox model to tfc data. after a thorough residual analysis, the logistic regression model has been used to identify the important covariates to classify future customers as risky or not. key words: classification, cox proportional hazard model, kaplan-meier, right-censored 1 introduction in medical follow-up studies, life testing, insurance, and other fields, it is impossible to observe the lifetime or the time of the first insurance claim of all subjects in the study. the reason is time itself. in most cases, it is highly likely that all the events have not been observed by the time one wants to analyze these lifetimes. for example, in a non-life insurance, not every insurance contract has a claim during its contract period. the individuals in the study who have no claims by the end of the study (or contract) period are labeled as right-censored. the only information the researcher has is the time between the contract initiation and the end of the study time, which is naturally less than the time to the first claim. the simplest kind of censoring is single censoring https://creativecommons.org/licenses/by-nc/4.0/ mailto:piumiayodyachathurangani@gmail.com https://orcid.org/0000-0001-8296-9218 war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 119 which occurs when all subjects are censored at the same time. there are two types of censoring, namely type i and type ii censoring. in type i, the censoring time is predetermined whereas type ii occurs when a predetermined number of failures are observed, and the remaining subjects are then right censored. in many studies, subjects are not censored at the same time. this is called random censoring. these types of data are commonly referred to as survival data. the analysis of such data is important in many fields including reliability, engineering, biology, insurance, and medicine (de mel 2014). under the random censoring model, we assume that 𝑋1, 𝑋2, … , 𝑋𝑛 are independent nonnegative random variables with an absolutely continuous distribution function 𝐹(𝑡) = 𝑃(𝑋 ≤ 𝑡). the censoring variables 𝑌1, 𝑌2, … , 𝑌𝑛 are also independent nonnegative random variables with an absolutely continuous distribution function 𝐺(𝑦) = 𝑃(𝑌 ≤ 𝑦). we further assume that both random variables 𝑋 and 𝑌 are independent from each other. in this model, the observable random variables are 𝑍𝑖 = min (𝑋𝑖 , 𝑌𝑖 ) and 𝛿𝑖 = 𝐼(𝑋𝑖 ≤ 𝑌𝑖 ), where 𝛿𝑖 indicates whether 𝑍𝑖 is an uncensored observation or not. the 𝑌𝑖 ’s right censor the 𝑋𝑖 ’s. in survival analysis, a variety of parametric and nonparametric significant tests can be used to identify the observed differences among the empirical survival curves. the most commonly used nonparametric test is based on logrank statistic (oulidi et al. 2010). in survival literature, the survival function is usually estimated from the observed data by using the kaplan-meier estimator (kaplan et al. 1958). nelson-aalen proposed a nonparametric estimator (nelson 1972) for the cumulative hazard rate function 𝛬(𝑡). in non-life insurance studies, one of the main interests is to investigate the association between the claim sizes or claim times of insurance contracts and related covariates. these covariates are sometimes referred to as risk factors. examples of commonly encountered risk factors include age, sex and the health condition of the contract holder, vehicle type, premium value, hiring status, assured sum, lease type, and the brand name of a vehicle. identifying and measuring this association helps insurance companies to understand how these factors are associated with the occurrence or nonoccurrence of an insurance claim. this in turn helps insurance companies to calculate the premium of a contract according to the customer’s risk. since the survival data is not normally distributed and has partial information, the standard statistical techniques like multiple linear regressions cannot directly be applied to analyze such data. cox proposed a semi-parametric multiple regression models for survival data called cox’s proportional hazard model (cox 1972). this model can be used to identify the important factors in the study and to compare the hazard rate functions among different groups. furthermore, contrast to parametric models, this method makes no assumptions about the baseline hazard function (cox 1972). in the final part of this research, we use the logistic regression model to identify a future customer as a risky or not. this is a classification problem with binary response data variable having categories, 1 for a risky customer and 0 if not risky. in machine learning literature the most commonly used model for this purpose is the logistic regression model, a type of multiple linear regression model. in insurance, we war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 120 collect the data in only the insurance policy period. therefore, we can classify each customer as a risky customer or not without any difficulty. 2 materials and methods in this section, we introduce some basic concepts and their definitions in survival analysis 2.1 survival function let 𝑇 be an arbitrary continuous nonnegative random variable with distribution function 𝐹(𝑡) = 𝑃(𝑇 ≤ 𝑡) and the density function 𝑓(𝑡) = 𝑑𝐹(𝑡) 𝑑𝑡 . survival function of 𝑇 is defined as 𝑆(𝑡) = pr(𝑇 > 𝑡) = 1 − 𝐹(𝑡). (1) this measures the probability that a subject survives beyond some specific time 𝑡. the hazard rate function or instantaneous event rate is usually denoted by ℎ(𝑡) and it is the probability that an individual who is under observation has an event at time 𝑡. define by ℎ(𝑡) = lim 𝑑𝑡→0 pr(𝑡 ≤ 𝑇 < 𝑡 + 𝑑𝑡| 𝑇 ≥ 𝑡) 𝑑𝑡 = 𝑓(𝑡) 𝑆(𝑡) . (2) the function 𝛬(𝑡) = ∫ ℎ(𝑢)𝑑𝑢 𝑡 0 is the cumulative hazard function of 𝑇 and 𝑆(𝑡) = exp {−𝛬(𝑡)}. 2.2 kaplan-meier survival estimator the kaplan-meier (km) (kaplan et. al, 1958) is a nonparametric estimator of a survival function 𝑆(𝑡) given by �̂�(𝑡) = ∏ exp (− 𝑑𝑘 𝑛𝑘 ) =𝑘:𝑇𝑘≤𝑡 ∏ (1 − 𝑑𝑘 𝑛𝑘 ) ,𝑘:𝑇𝑘≤𝑡 (3) where 𝑛𝑘 is the number of subjects at risk (alive) just before 𝑇𝑘 and 𝑑𝑘 is the number of failures at time 𝑇𝑘. the times 𝑇1 < 𝑇2 < ⋯ < 𝑇𝐿 denote the l distinct ordered observed failure times in the study with n subjects. the nelson-aalen estimator (nelson, 1972) for λ(t) is given by �̂�(𝑡) = ∏ 𝑑𝑘 𝑛𝑘 𝑘:𝑇𝑘≤𝑡 . with tied failure data, the above formulas need to be modified. war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 121 2.3 cox proportional hazards model the cox proportional hazards model (cox 1972) is a semi-parametric model because it does not assume any conditions on ℎ0(𝑡). the cox’s model (cox 1972) is given by ℎ(𝑡| 𝑋) = ℎ0(𝑡) exp (∑ 𝛽𝑖 𝑋𝑖 ) 𝑝 𝑖=1 , (4) where ℎ(𝑡| 𝑋) which is usually written as just ℎ(𝑡) is the hazard rate function at time 𝑡, ℎ0(𝑡) the baseline hazard rate function, and 𝑋 = (𝑋1, 𝑋2, … , 𝑋𝑝) the covariate vector of size 𝑝. this model can be used to analyze survival data by regression model such as multiple linear regression and generalized linear models because equation (4) can be reduced to log( ℎ(𝑡| 𝑋) ℎ0(𝑡) ) = ∑ 𝛽𝑖 𝑋𝑖 𝑝 𝑖=1 . the dependent variable in the cox’s model is the hazard rate function, ℎ(𝑡). therefore, one can use this model to compare the survival curves in different groups by taking into account other related covariates. we assume that ℎ0(𝑡) is unknown and common to all subjects in the study. most of the time, the coefficients 𝛽1, 𝛽2, … , 𝛽𝑝 are estimated by using the partial likelihood method (cox, 1972). the partial likelihood function is given by the following equation. 𝑃𝐿(𝛽) = ∏ ∏ [ 𝑌𝑖(𝑡)𝑒 𝑋𝑖(𝑡)𝛽 ∑ 𝑌𝑖(𝑡)𝑒 𝑋𝑖(𝑡)𝛽𝑛 𝑖=1 ] 𝑑𝑁𝑖(𝑡) 𝑡≥0 𝑛 𝑖=1 (5) where 𝑁𝑖 (𝑡) = 𝐼(𝑍𝑖 ≤ 𝑡, 𝛿𝑖 = 1) and 𝑌𝑖 (𝑡) = 𝐼(𝑋𝑖 ≥ 𝑡) are the event process and atrisk process respectively for 𝑖𝑡ℎ subject (fleming et al., 1991). in most of the times, one cannot find the exact solutions to equation (5) but can obtain the numerical solutions for the coefficient vector 𝛽 using numerical method like newton-raphson. 2.4 logistic regression model to model a binary response variable 𝑌, one can use the logistic multiple regression model. this is given as log( 𝑝(𝑋) 1−𝑝(𝑋) ) = 𝛽0 + 𝛽1𝑋1 + 𝛽2𝑋2 + ⋯ + 𝛽𝑝𝑋𝑝 (6) where 𝑋 = (𝑋1, 𝑋2, … , 𝑋𝑝) is vector of covariate, 𝛽 = (𝛽0, 𝛽1, 𝛽2, … , 𝛽𝑝) vector of regression coefficients, and 𝑝(𝑋) = 𝑃(𝑌 = 1 /𝑋). with a little algebra, one can rewrite the equation (6) as 𝑝(𝑋) = exp (𝛽0+𝛽1𝑋1+𝛽2𝑋2+⋯+𝛽𝑝𝑋𝑝) 1+exp (𝛽0+𝛽1𝑋1+𝛽2𝑋2+⋯+𝛽𝑝𝑋𝑝) (7) war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 122 in logistic regression, parameters are estimated by using maximum likelihood method. future observations can be classified into two groups by using 𝑝(𝑋) with some probability threshold, for example 𝑝(𝑋) = 0.5. 3 results and discussion in this study, we consider a dataset consisting of motor vehicle insurance contracts from a leading insurance company in sri lanka. customers were enrolled for a oneyear period from january 01, 2016 to december 31, 2016 therefore; the end of the study period was december 31, 2016. our dependent variable is the time of the first claim of the contract and we also consider nine covariates, namely, vehicle type, premium value, hiring status, gender, age, assured sum, lease type, brand name of a vehicle, and class type of a vehicle. our insurance dataset contains more than 300,000 non-life insurance (motor vehicle insurance) contracts. after removing contracts with missing values we use simple random sampling technique to obtain a random sample with 599 contracts. any contract with no claims in the study period is treated as right censored data. in figure 1, we depict a part of our dataset to give some idea about the survival data. study began on january 01, 2016 and was terminated on december 31, 2016. insurance contracts can initiate at any time during the contract year. in figure 1, the horizontal lines depict the claim time of the first claim of an insurance contract. the blue dots on these lines represent the actual time of the first claim and this is known as the calendar time of the first claim. fig 1: calendar times of subjects in the study. war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 123 if we cannot observe the first claim during the year, these times go beyond the termination time of the study and they are called right censored data. for our dataset, we first compute the kaplan-meier survival estimate by using survival package in r software. it is depicted along with the 95% confidence band in figure 2. fig 2: kaplan-meier estimate for claim times for motor vehicle insurance claim data table 1 displays the summary statistics obtained using logrank test to compare significant differences among the survival curves for categories of categorical variables in the study. all these factors are significant at the 5% significance threshold. table 1: summary statistics for categorical variables based on logrank. figure 3 displays the estimated survival curves for two age ranges. this clearly displays a significant difference between two estimated curves for age range categories and verifies the logrank result. here, the two groups for age range are above and below 35 years. but one can use more than two groups. it is clear from the figure 3 that the youngers are riskier than the elders. identifying theses risky groups helps insurance companies to calculate the premiums for their insurance contracts. variable name p-value age range 0.0421 lease type 0.00441 vehicle type 3.69e-11 war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 124 fig 3: estimated survival functions by age range we next fit the cox proportional hazard model (cox 1972) including all above mentioned nine covariates. results from r software are displayed below in table 2. according to the output, only five variables, namely vehicle type, premium value, assured sum, lease type, and age range are significant at the 5% confidence threshold. gender is not significant. after including other six predictors in the model only age range is significant but not gender at the significance level 0.05. we can remove the gender from the cox proportional hazard model (cox 1972) since the p-value is larger than the significance level 0.05 but p-value for age range is less that above significance level. table 2: parameter estimates for cox proportional hazard model with all the variables. variable name β (estimated coefficients) se (est. coef.) wald test p-value vehicle type 9.515e-02 3.341e-02 2.848 0.004395 premium value 3.731e-05 7.627e-06 4.892 1.00e-06 hiring status 4.213e-02 2.342e-01 0.180 0.857234 assured-sum 5.978e-07 1.646e-07 3.632 0.000281 lease type 6.972e-01 1.782e-01 3.912 9.16e-05 gender 7.823e-02 1.965e-01 0.398 0.690582 age range 5.256e-01 1.813e-01 2.900 0.003736 brand name 3.291e-03 7.609e-03 0.433 0.665334 class type of vehicle 5.872e-03 6.672e-02 0.088 0.929873 war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 125 after removing insignificant predictors, we fit the cox model again with predictors, namely vehicle type, premium value, assured sum, lease type, and age range. the results from r software for the reduced cox proportional hazard model are displayed below in table 3. table 3: parameter estimates for the reduced cox proportional hazard model. variable name β (estimated coefficients) se (est. coef.) wald p-value vehicle type -9.91e-02 2.16e-02 -4.59 4.5e-06 premium value 3.78e-05 7.49e-06 5.05 4.5e-07 assured_sum -6.02e-07 1.61e-07 -3.75 0.00018 lease type 6.82e-01 1.75e-01 3.89 9.9e-05 age range -5.08e-01 1.79e-01 -2.84 0.00447 according to the results in table 3, all variables in the reduced model are significant at the 5% confidence threshold. therefore, we could treat this reduced cox proportional hazard model with variables, namely, vehicle type, premium value, assured sum, lease type, an age range as our final model for the motor vehicle insurance data set. before using it for prediction purposes it is necessary to carry out the model validation. in model validation, we check the proportional hazard assumption, influential observations assumption and nonlinearity assumption. 3.1 model validation since cox proportional hazard model fits under several assumptions it is better to check whether the fitted cox model adequately explains the data. in order to check the three main assumptions namely, violation of the assumption of proportional hazards, influential data and, nonlinearity in the relationship between the log hazard and the covariate, the residuals method is used. schoenfeld and martingale residuals (fleming et al. 1991) are the most frequently used residuals to check the assumptions of the cox model. testing the proportional hazards assumption to check the proportion hazard assumption of cox model, we use the scaled schoenfeld residuals of the final fitted model. we plot these schoenfeld residuals with all the covariates in the final model, namely vehicle type, premium value, assured sum, lease type, and age range. since there are no apparent patterns in any of the plots in figure 4, we can assume that the proportional hazard assumption holds. war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 126 fig 4: schoenfeld residual graphs for vehicle type, premium value, assured sum, lease type, and age range in figure 4, the solid line represents a smooth spline fit to the plot and the dashed line represents a ±2 standard-error band around the fit. since there is no pattern with time, the assumption of proportional hazards appears to be satisfied by the above mentioned covariates. checking for influential observations we use the dfbeta values to identify influential observations or outliers. here dfbeta measures the difference in each parameter estimate with and without the influential point (montgomery et al. 2006). figure 5 displays these dfbeta values from the final fitted model. we use the usual cutoff value,|𝐷𝐹𝐵𝐸𝑇𝐴| > 2/√𝑛 to identify an influential observation. here 𝑛 is the sample data value in the training dataset which is used for training the cox model. this cutoff is 0.08 for our study. there are only a few dfbeta values that correspond to large premium values and assured sum. but we can neglect these points in the dataset. war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 127 fig 5: index plots of dfbeta values for the fitted cox regression versus age range, premium, lease type, assured sum, and vehicle type. detecting nonlinearity nonlinearity, an incorrectly specified functional form in the parametric part of the model, is a potential problem in cox regression like in linear and generalized linear models. the martingale residuals may be plotted against covariates to detect nonlinearity. for nonlinear assumption for cox’s model, continuous type covariates should be linear. in our study, this assumption is checked only for the predictors, namely, assured sum and premium value. figure 6 displays these two residual plots which are clearly not linear. war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 128 fig 6: martingale residual graph for covariates, assured sum and premium. in figure 6, we plot the martingale residuals against the two covariates, assured-sum and premium. we also fit the lowess smooth based on a span 0.2. here, null cox model means the model with all continuous type variables. to correct this nonlinear violation, we use logarithm transformations of both above variables in the fit. the results are displayed in table 4. all predictors in this model fit are significant at the 5% significant threshold and all residual plots seem to be good. but we do not report these plots here because they are very similar to the above residual plots. therefore, we can treat this fitted model as our final model for the motor vehicle insurance data. so we can use this model for future predictions. table 4: final fitted cox model variable name β (estimated coefficients) se (coef) wald p-value vehicle type 0.07102 0.02604 2.728 0.006378 log (premium value) 1.28256 0.30479 4.208 2.58e-05 log (assured sum) 0.69764 0.24308 2.870 0.004105 lease type 0.62199 0.17238 3.608 0.000308 age range 0.47701 0.18084 2.638 0.008344 3.2 classification of the risky customers using logistic regression model in this subsection, we fit the logistic regression model to the response variable with the above nine predictors in the motor vehicle claim dataset. here, we choose two levels, war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 129 namely, a risky customer and a not risky customer as our response variable and it is a qualitative variable with two levels. in this study, we treat a customer as a risk if his or her first claim size is above 30,000 during the contract duration; otherwise we consider him or her as not risky. to fit this model, we randomly select 500 data points from our original sample for training data and we treat the remaining 99 data points as the test data. table 5 shows the coefficient estimates for a logistic regression model that uses all of the above predictor variables, but we present only significant variables at the 5% significant level. they are vehicle type, premium value, lease type, and age range. table 5: parameter estimates in fitted logistic regression model. estimate std. error z value (intercept) 1.94e+00 5.55e-01 -3.498 0.000468 premium value 1.41e-05 5.91e-06 2.386 0.017011 lease 2.46e-01 3.91e-01 2.63 0.005287 vehicle -1.44e-01 5.31e-02 -2.718 0.006576 age range -8.04e-01 3.89e-01 -2.064 0.039042 table 6: confusion matrix for the test dataset next, we use the fitted logistic regression model with five significant predictors, namely, vehicle type, premium value, lease type, and age range to predict the customers in the test dataset as a risky or not risky customer. the results are displayed in the following confusion matrix for the test dataset in table 6. for the test dataset, the logistic regression model makes 83 correct classifications out of 99 sample data and the percentage accuracy is 0.8384. this achieves high accuracy. but logistic regression model makes only 16 misclassifications, and the percentage misclassifications is 0.1616. the final goal of any insurance is to create profitable portfolio by assessing risk factors. as we mentioned, many of the research work conducted earlier related to motor insurance contracts analyzed the claim sizes with cox proportional hazard models without conducting a proper residual analysis. in our study, we analyzed the time of the first claim of motor vehicle insurance contracts by using cox’s proportional hazard model. we identified five predictors as the most influential to the time of the first claim. these findings will help the company to take the decisions regarding the premium size of the contract for the different age groups, for example, according to their risk. the survival estimates of the time of the first claim obtained by using true default status no yes total predicted risk value no 82 3 85 yes 13 1 14 total 95 4 99 war de mel and wapa chathurangani modelling non-life insurance in sri lanka ruhuna journal of science vol 11 (2): 118-130, december 2020 130 kaplan-meier estimator will give the idea of the distribution of these claim time data. since one contract may have more than one claim in its contract period, in future studies, we can analyze this type of data by treating them as recurrent event data. by conducting a thorough residual analysis, we found the optimal transformation of some selected covariates. finally, we obtained a simple way to identify the risky customers by applying the logistic regression model. in future studies, one can apply more sophisticated methods like linear or quadratic discriminant analysis and bayesian classifier methods to high accuracy. supplementary material supplementary material for this paper is available separately. acknowledgments two anonymous reviewers are acknowledged for valuable comments on the initial draft of the manuscript. references oulidi a, marion j-m, ganachaud h. 2010. survival analysis methods in insurance applications in car insurance contracts. (accessed at https://studylib.net/doc/8687251/survival-analysis-methods-ininsurance-applications-in-car) (published online). cox dr. 1972. regression models and life-tables. journal of royal statistical society b 34: 187-200. de mel withanage ajith raveendra. 2014. on some inferential problems with recurrent event models. doctoral dissertation, missouri university of science and technology (accessed at http://scholarmine.mst.edu/doctoraldissertations/2340. anna e. 2017. variable selection for the cox proportional hazards model. master thesis. umea university (accessed at http://www.divaportal.org/smash/get/diva2:1067479/fulltext01.pdf). fleming, thomas r, and harrington, david f. 1991. counting processes and survival analysis. john wiley & sons inc, new york. kaplan el, meier p. 1958. nonparametric estimation from incomplete observations. journal of the american statistical association 153 (282): 457-481. montgomery dc, peck ea, vining gg. 2006. introduction to linear regression analysis, fourth edition. john wiley & sons inc, new york. nelson w. 1972. theory and application of hazard plotting for censored failure data. technometrics 14: 945-966. terry mt, patricia mg. 1990. statistics for biology and health. new york: springer verlag. https://studylib.net/doc/8687251/survival-analysis-methods-in-insurance-applications-in-car)%20(online%20published https://studylib.net/doc/8687251/survival-analysis-methods-in-insurance-applications-in-car)%20(online%20published http://scholarmine.mst.edu/%20doctoraldissertations/2340 sv-lncs ruhuna journal of science vol 6: 42-49, december 2015 issn: 1800-279x  faculty of science university of ruhuna  faculty of science, university of ruhuna 42 printed in sri lanka performance of zn rechargeable cells having polypyrrole cathodes doped with surfactant anion w.a.d.s.s. weerasinghe, l.k.m. madhushani, k.p. vidanapathirana 1 and k.s. perera 1 department of electronics, wayamba university of sri lanka, kuliyapitiya, sri lanka correpondence: 1 kamalpv41965@gmail.com received: 19 th october 2015, revised: 20 th december 2015, accepted: 21 st december 2015 abstract. currently there exists a very high demand for electrochemical energy storage devices such as rechargeable cells and super capacitors due to massive increase in the use of portable electronics. at the moment this demand is mainly supplied by li based cells. however, due to high cost, rapid reactivity and issues in disposal of li, attention has been diverted on non li based cells. zn has been identified as a good candidate to replace li. even though several zn based rechargeable cells were reported, not much work has been carried out on cells with conducting polymer cathode based zn cells. in this study, performance of zn rechargeable cells fabricated with conducting polymer polypyrrole (ppy) doped with a surfactant anion is reported. ppy films were galvanostatically electro polymerized on to stainless steel disc and cells were assembled with a polyvinylidenefluoride based gel polymer electrolyte having zinc-trifluoromethanesulfonate as the salt. cells having different cathode thicknesses were fabricated and were characterized using cyclic voltammetry, electrochemical impedance spectroscopy and continuous charge-discharge tests. cells exhibited open circuit voltages between 0.9 -1.0 v. cycling testes showed that cycleable capacity almost follows the thickness variation of the cathode. continuous charge-discharge tests revealed that the capacity decrease with the cycle number was higher in thinner films. keywords. gel polymer electrolyte, polypyrrole, rechargeable cell, surfactant. 1 introduction currently there exists a very high demand for electrochemical energy storage devices such as rechargeable cells and super capacitors due to massive increase in the use of portable electronics (chalamala 2007). at the moment this market is mainly covered by li based cells. however, li is expensive, highly reactive and it becomes problematic in the disposal of the cell. weerasinghe et al. performance of zn rechargeable cells ruhuna journal of science 43 vol 6: 42-49, december 2015 therefore, attention has been focused on developing low cost non-li based rechargeable cells. there are several substitutes such as zn, cu and mg to replace li. conducting polymers (cps) have been identified as one of the most promising candidates for cathode material due to their interesting features such as appreciable electronic conductivity, low cost, easy handling and environment friendliness (cheng et al. 2006; huang 2006; wang et al. 2006; green et al. 2010). several conducting polymers have been tested for use as cathode materials. among them polypyrrole (ppy), polyaniline (pani), polythiyophene (pth) are widely reported polymers (ghanbari et al. 2007; li et al. 2013; trinh et al. 2013). out of the three polymers mentioned above, ppy has been mostly studied as the cathode material due to some special features such as longer cycle life, lower self discharge rate, endurance to over discharging, low manufacturing cost and shape flexibility. during the synthesis of the cps, different anions are incorporated to the polymer backbone and it has been found that the properties of such polymer electrodes depend on the type of anions used (skaarup et al. 2000, 2003). in one of our early studies on li based cells, we have reported that better performance of cp could be obtained when it is doped with large anions and even more electro activity with surfactant based anions (vidanapathirana et al. 2000). in this study, the performance of a zn rechargeable cell with a ppy based cathode doped with sodium dodecylsulphonate (sds) is reported. a gel polymer electrolyte based on polyvinylideneflouride (pvdf), ethylene carbonate (ec), propylene carbonate (pc) and zinctrifluoromethanesulfonate (zntf) was used as the separator. 2 materials and methods 2.1 preparation of the cathode the monomer pyrrole (aldrich) was distilled and stored under refrigeration prior to use. polypyrrole (ppy) films were galvanostatically electropolymerized on to stainless steel discs using a three electrode set up. a ag/agcl and pt electrodes were used as the reference and the counter electrodes respectively. monomer concentration was 0.1 m. the salt used was 0.05 m sodium dodecylsulfonate (sds) (aldrich). it has been reported that a charge density of 240 mc/cm 2 gives rise to a film of thickness 1 μm (diaz et al, 1980). based on that, the thickness of the ppy films was varied from 0.5 to 1.0 µm. http://www.sciencedirect.com/science/article/pii/s0378775312012657 weerasinghe et al. performance of zn rechargeable cells ruhuna journal of science 44 vol 6: 42-49, december 2015 2.2 preparation of the gel polymer electrolyte (gpe) polyvinylidenefluoride (pvdf) (aldrich), zinc trifluoromethanesulfonate (zn(cf3so3)2 – zntf) (aldrich), ethylene carbonate (ec) (aldrich) and propylene carbonate (pc) (aldrich) were used as received. appropriate amounts were weighed and magnetically stirred well and heated at 120 ° c for 30 minutes. the hot mixture was pressed in between two glass plates. thereby, it was possible to obtain a bubble free thin film. composition of the gpe was chosen as 0.5 pvdf: 1 ec: 1 pc: 0.7 zntf (by weight). conductivity of the electrolyte was found to be in the range 10 -3 scm -1 . 2.3 fabrication of cells first, open circuit voltages (ocv) of the cells were measured using a digital multimeter. subsequently, cyclic voltammetry tests were carried out for the cells in the potential range 0.2 1.0 v using a computer controlled potentiostat/ galvanostat (metrohom-autolab m101) where ppy electrode was used as the working electrode and zn was used as both the counter and reference electrodes. cells were cycled at the scan rate of 10 mvs -1 . electrochemical impedance spectroscopy (eis) measurements of the cells were carried out for the frequencies ranging from 400 khz to 0.01 hz using metrohom frequency response analyzer. impedance data were analyzed using non-linear least square fitting programme. measurements were taken at time intervals of 15, 30, 60 and 90 minutes. thereafter these cells were tested for their ability to withstand continuous charge and discharge cycling. for this, the cells were first galvanostatically discharged to 0.5 v, immediately subjected to a galvanostatic charge up to 2.0 v and maintained at that potential until the desired current (20% of maximum charge current) was reached and then discharged using a computer controlled charge-discharge setup. the maximum charge and discharge currents were set to 75 µa. 3 results the open circuit voltages of the cells fabricated by varying the thickness of ppy/ds cathode are given in table 1. weerasinghe et al. performance of zn rechargeable cells ruhuna journal of science 45 vol 6: 42-49, december 2015 table 1. open circuit voltage values (ocv) of the cells with the variation of ppy thickness. cyclic voltammograms obtained for cells fabricated with different cathode thicknesses are shown in figure 1. figure 1(a) shows the capacity variation of the cells in 1 st cycles while figure 1(b) shows the same in the 10 th cycle. fig. 1. cyclic voltammograms with the capacity variations according to the ppy film thickness. scan rate 10 mvs -1 . (a) 1 st cycle, (b) 10 th cycle. the impedance plots obtained for the cells at different time intervals are given in figure 2. cells were subjected to continuous charge-discharge tests and the discharging capacity variation with the cycle number for the three cells is given in figure 3. thickness (µm) ocv (v) 0.50 0.951 0.75 0.995 1.00 1.000 (b) (a) weerasinghe et al. performance of zn rechargeable cells ruhuna journal of science 46 vol 6: 42-49, december 2015 fig. 2. impedance plots taken at different time intervals for cells having three ppy film thicknesses. (a) 0.5 µm , (b) 0.75 µm, (c) 1.0 µm. fig. 3. discharging capacity variation with the cycle number for the cells having different ppy film thicknesses. (a) (b) (c) weerasinghe et al. performance of zn rechargeable cells ruhuna journal of science 47 vol 6: 42-49, december 2015 4 discussion open circuit voltages (ocv) of the cells showed an increasing trend from 0.9 v to 1.0 v when the thickness of the cathode was varied from 0.5 to 1.0 µm. this indicates that there is no significant effect on the ocv with the thickness of the cathode. however this ocv is very much comparable with the results reported in early studies. zhang et al. have obtained only 1.15 v for znpolyaniline cells even with a liquid electrolyte (zhang et al. 2006). cyclic voltammograms (cvs) in figure 1 exhibit anodic peaks in the potential range 0.6 to 0.65 v and cathodic peaks in the range 0.4 to 0.55 v. the peaks at cvs appear at potentials on which redox reactions of the cells take place. the anodic peak is assigned to the oxidation of the ppy electrode while the cathodic peak to the reduction process (tang et al. 2010). these oxidation and reduction peaks can be assigned to the cation (zn 2+ ) insertion and deinsertion because one of our early studies has reported that when ppy is doped with a large anion, it tends to trap in the polymer chain and cation in the cycling electrolyte move in and out during the redox process (vidanapathirana et al. 2002). if the reactions are fully reversible, the peaks should appear at identical potentials resulting zero peak separation. however, the cvs obtained show a peak separation which may be due to difference in the rate of reactions during oxidation and reduction. furthermore, it is seen that the thickness of the cathode is also affecting the peak separation. as the thickness increased, peak positions shift towards higher potentials. this is because for thicker films, more energy is needed to obtain the maximum cycleable capacity and hence peak positions move to higher potentials (vidanapathirana et al. 2006). this shift in the peak position could also be due to the slower diffusion of ions in the thicker films as suggested by osaka et al. (osaka et al. 1987). in the 1 st cycle as well as in the 10 th cycle, when the cathode thickness is increased, the corresponding capacity has increased. this is because when thickness is high, more charges can be retained giving rise to high capacity. when the capacity variation with cycle number is considered, it is seen that between 1 st and 10 th cycle, (fig 1(b)), there is no significant loss in the cycling capacity. this proves that there exist good interfacial contacts between the electrodes and electrolyte without any side reaction. eis results on three different cells taken at 15, 30, 60, 90 minutes are given in figure 2. theoritically, high frequency intercept of the plot gives the bulk electrolyte resistance (rb). semi-circle arc at intermediate frequencies represents the zn anode resistance. low frequency part corresponds to the diffusion process and charge accumulation in ppy cathode. rb of all cells at each time period has not changed. this confirmed the stability of the gel polymer electrolyte. weerasinghe et al. performance of zn rechargeable cells ruhuna journal of science 48 vol 6: 42-49, december 2015 in cells based on conducting polymers, at low frequencies, two spikes are possible. one at about π/4 inclination representing warburg behavior and the other portion has an inclination close to π/2. warburg behaviour represents diffusion controlled kinetics. the portion with π/2 shows capacitive behavior indicating polymer electrode behaves like a capacitor at very low frequencies. from the three impedance plots, a special feature in figure 2(c) is that no charge accumulation has taken place in the cathode. always, diffusion controlled kinetics are present. probably this may be due to the mismatch of thickness of the cathode and the low frequency range. if the frequency was lowered too much, charge accumulation would have initiated. in the cell having the thin cathode, soon after fabrication, charges may accumulate on the cathode but as time goes on, diffusion controlled kinetics also take place. because of this effect, at 60 minutes and 90 minutes time periods, low frequency region shows two spikes having two different inclinations. similar results have been observed with the cell having the cathode with intermediate thickness. results obtained from continuous charge-discharge tests, (figure 3), show that discharge rate of thin films are higher than that of thicker films. this can be attributed to the fact that thin films have short ion diffusion and migration lengths compared to thicker films (wang et al. 2009; valaski et al. 2002). this phenomena has been observed in a previous study on ppy films deposited on stainless steel mesh (wang et al. 2005). in conclusion, zn rechargeable cells with conducting polymer cathode and gel polymer electrolyte were successfully fabricated with reasonable open circuit voltage and cycling capacity. it was evident that discharge rate of the cells with thinner cathodes are higher than that of thicker electrodes. acknowledgments financial assistance provided by national science foundation (grant no. rg/2014/bs/01), national research council (grant no. nrc-12/109) and wayamba university of sri lanka (research grant no. srhdc/rp/04/14/01) are highly acknowledged. references chalamala br. 2007. portable electronics and the widening energy gap. in: proceedings of the ieee, 95/11, 2106-2107. cheng f, tang w, li c, chen j, liu h, shen p, dou s. 2006. conducting poly(aniline) nanotubes and nano fibers: controlled synthesis and application in lithium/poly(aniline) rechargeable batteries. chemistry a european journal, 12/11, 3082−3088. weerasinghe et al. performance of zn rechargeable cells ruhuna journal of science 49 vol 6: 42-49, december 2015 diaz af, castillo ji. 1980. a polymer electrode with variable conductivity – polypyrrole. j. chem. soc. commun., 397-398. ghanbari k, mousavi mf, shamsipur m, karami h. 2007. synthesis of polyaniline/graphite composite as a cathode of zn-polyaniline rechargeable battery. journal of power sources, 170/2, 513-519. green ra, baek s, warren lap, martens pj. 2010. conducting polymer hydro gels for medical electrode applications. science and technology of advance materials, 11/1, 1−13. huang j. 2006. synthesis and applications of conducting polymer polyaniline nanofiberes. pure and applied chemistry, 78/1, 15 − 27. li w, zhang q, zheng g, seh zw, yao h, cui y. 2013. understanding the role of different conductive polymers in improving the nanostructured sulfur cathode performance. nano lett., 13/11, 5534-5540. osaka t, naoi s, ogano s, nakamura s. 1987. dependence of film thickness on electrochemical kinetics of polypyrrole and properties of li/ppy battery performance. journal of electrochem soc., 134/9, 2096. skaarup s, west k, gunarathne lmwk, vidanapathirana kp, careem ma. 2000. determination of ionic carriers in polypyrrole. solid state ionics, 136/137, 577−582. skaarup s, lasse b, vidanapathirana k, thybo s, tofte p, west k. 2003. simultaneous anion and cation mobility in polypyrrole. solid state ionics, 159, 143−147. tang z, wu j, li q, lan z, fan l, lin j, hung m. 2010. the preparation of poly(glycidyl acrylate) polypyrrole gel-electrolyte and its application in dye-sensitized solar cell. electrochim acta, 55, 4883-4888. trinh nd, saulnier m, lepage d, schougaard sb. 2013. conductive polymer film supporting lifepo4 as composite cathode for lithium ion batteries. journal of power sources, 221, 284–289. valaski r, ayoub s, micaroni l, hummdgen ia. 2002. influence of film thickness on charge transport of electrodeposited polypyrrole thin films. thin solid film, 41, 5206-5210. vidanapathirana kp, perera k, careem ma. 2000. application of conducting polymer (polypyrrole) as an electrode in li rechargeable batteries. in: proceedings of the technical sessions, institute of physics, sri lanka 16, 88−96. vidanapathirana k p, careem m a, skarup s, west k. 2002. ion movement in ppy/dbs films in aqueous and non-aqueous electrolytes. solid state ionics, 154/55, 331-335. vidanapathirana k p, careem ma. 2006. effect of cathode thickness on the performance of the cell li/pan:ec:pc:licf3so3/ppy:dbs. sri lankan journal of physics, 7, 29-34. wang j, chen j, wang cy, zhou d, too co, wallace gg. 2005. electrochemical synthesis of polypyrrole films using stainless steel mesh as substrate for battery application. synth. met., 153, 117-120. wang j, wang cy, too co, wallace gg. 2006. highly-flexible fiber battery incorporating polypyrrole cathode and carbon nanotubes anode. journal of power sources, 161, 1458−1462. wang j z, chou sl, liu h, wang gx, zhong c, chew sy, liu hk. 2009. highly flexible and bendable free-standing thin film polymer for battery application. mat. lett., 63, 2352-2354. zhang j, shan d, mu s. 2006. a rechargeable zn-poly(aniline-co-m-aminophenol) battery. journal of power sources, 161/1, 685-691. http://www.ncbi.nlm.nih.gov/pubmed/?term=zhang%20q%5bauthor%5d&cauthor=true&cauthor_uid=24127640 http://www.ncbi.nlm.nih.gov/pubmed/?term=zheng%20g%5bauthor%5d&cauthor=true&cauthor_uid=24127640 http://www.ncbi.nlm.nih.gov/pubmed/?term=seh%20zw%5bauthor%5d&cauthor=true&cauthor_uid=24127640 http://www.ncbi.nlm.nih.gov/pubmed/?term=yao%20h%5bauthor%5d&cauthor=true&cauthor_uid=24127640 http://www.ncbi.nlm.nih.gov/pubmed/?term=cui%20y%5bauthor%5d&cauthor=true&cauthor_uid=24127640 http://www.ncbi.nlm.nih.gov/pubmed/24127640 http://www.ncbi.nlm.nih.gov/pubmed/24127640 http://www.sciencedirect.com/science/article/pii/s0378775312012657 http://www.sciencedirect.com/science/article/pii/s0378775312012657 http://www.sciencedirect.com/science/article/pii/s0378775312012657 http://www.sciencedirect.com/science/article/pii/s0378775312012657 ruhuna journal of science vol 12 (2): 144 -154, december 2021 eissn: 2536-8400 © faculty of science http://doi.org/10.4038/rjs.v12i2.108 university of ruhuna © faculty of science, university of ruhuna sri lanka 144 determination of the distribution of calotropis gigantea (l.) in sri lanka using maxent modelling technique w.p.s.n. wijeweera1*, k.a.d.w. senaratne2, k. dhileepan2 and m.p.k.s.k. de silva1 1department of zoology, university of ruhuna, matara, sri lanka 2biosecurity queensland, department of agriculture and fisheries, queensland, australia *correspondence: surendi87nisha@gmail.com; orcid: https://orcid.org/0000-0003-1359-0188 received: 5th may 2021, revised: 03rd november 2021, accepted: 21st december 2021 abstract calotropis gigantea is a drought-resistant, salt-tolerant, native plant in sri lanka with ayurvedic medicinal values. the plant is used for fiber, fodder and fuel, as well as a fertilizer. despite its benefits, c. gigantea has become an emerging problem in countries where it has been introduced because of its invasiveness. although c. gigantea is widely distributed in sri lanka, precise information on its distribution is lacking. therefore, the present study was aimed at determining the distribution of c. gigantea in sri lanka. field surveys were conducted in 120 sites covering all provinces in sri lanka from december 2014 to june 2015 to record the occurrence of c. gigantea. c. gigantea was distributed in all provinces except the central province. it was more widespread along coastal regions, but its occurrence was low in the western and sabaragamuwa provinces. maxent modelling predicted that the entire coast of northern, north-central and eastern provinces contain the highest probability of c. gigantea distribution whereas the low probability was in north-western, western, southern, uva, central, and sabaragamuwa provinces. no occurrence probability was predicted in certain regions of southern, sabaragamuwa, uva, and central provinces of sri lanka. the study provides information on the current and potential distribution range of c. gigantea in sri lanka. keywords: calotropis gigantea, maxent modelling, plant distribution 1 introduction the spatial distribution of a species is the arrangement of that species across the earth’s surface. no species in the world is adapted to live in all environmental conditions of the earth. their spatial distribution is limited by biotic and abiotic factors (pidwirny 2018). climatic factors act as major limiting factors for the distribution of plants. the microclimate is the climatic condition that prevails in localized regions closer to the surface of the earth. it consists of environmental variables including temperature, moisture content, wind speed, and light (naiman et https://rjs.ruh.ac.lk/index.php/rjs/index https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v12i2.108 https://creativecommons.org/licenses/by-nc/4.0/ mailto:surendi87nisha@gmail.com w.p.s.n. wijeweera et al. distribution of calotropis gigantea in sri lanka ruhuna journal of science vol 12 (2): 144 -154, december 2021 145 al. 2005) and is closely bound with the habitat and influence on plants on a fine scale (bramer et al. 2018). climatic differences create significant changes in the vegetation and form climatic zonation (azarkhavarani et al. 2015). the close relationship of climate and vegetation is used to define and analyze bio-climatic zones using different vegetation and climatic maps (azarkhavarani et al. 2015). later, the use of computerized data related to climate and vegetation has led to species distribution modelling techniques (barbet-massin et al. 2018). primarily, species distribution models are developed to predict the distribution of species and secondarily to study the functional association of species with their living environment (austin 2002). since 2006, the maxent software package is widely used for modelling the species distribution covering more than 1000 publications. maxent program uses presenceonly data (species recorded locations) and data of environmental predictors as input data of the program (merow et al. 2013). modelling techniques have been applied to study the distribution range of calotropis procera, an invasive plant in australia. the driving factors for its distribution had not been identified and assuming “climatic conditions” as a driving factor on the plant distribution, maxent modelling has been applied to identify the current and potential distribution of c. procera. according to the model prediction, the distribution of plants is best explained through climatic variables and human disturbances (menge et al. 2016). c. gigantea is native to india, china, sri lanka, and malaysia (dhileepan 2014) and is also distributed in afghanistan, algeria, burkina faso, cameroon, ghana, guinea-bissau, and iran (kumar et al. 2013). it is considered a medicinal plant having ayurvedic value in its native range (kumar and kumar 2015). different plant parts of c. gigantea are used to cure a wide range of diseases including bronchial asthma, cholera, convulsions, pneumonia, ringworm infection, smallpox infection, toothache, epilepsy, skin diseases, and epilepsy (kumar and kumar 2015). in addition, c. gigantea is used as fodder, fiber source, fuel and febrifuge (kumar et al. 2013). however, in certain countries, the plant is considered as an invasive species. it is recorded as an exoticinvasive plant in australia, the virgin islands of the united states, mexico, and brazil (kumar et al. 2013). they prefer to grow on the abandoned over-cultivated areas, over-grazed grounds, roadsides and lagoon edges (kumar et al. 2013). c. gigantea is commonly found in sri lanka, but there are no published records on its distribution in sri lanka and the habitats they occupy. therefore, the objectives of the present study are to collect occurrence data of c. gigantea in sri lanka, to identify habitats and habitat characteristics of c. gigantea, to prepare a map of its distribution concerning selected environmental variables by using maxent modelling technique and to study the density of the plant in different regions of sri lanka. w.p.s.n. wijeweera et al. distribution of calotropis gigantea in sri lanka ruhuna journal of science vol 12 (2): 144 -154, december 2021 146 2 material and methods 2.1 collecting occurrence data of c. gigantea a field survey on the distribution of c. gigantea was conducted from december 2014 to june 2015 covering 120 sampling sites representing nine provinces of sri lanka (figure 1). each field visit covered eight (8) sampling sites on average. sampling was done only once for each site. the roadside sampling was done as calotropis spp. tend to grow closer to road-edges and for easy accessibility of sampling (sharma et al. 2010). roadside sampling sites were selected randomly maintaining equal distances on the main road at 30-minute intervals while travelling on a vehicle with a speed of 50 km per hour. if a new site with c. gigantea plants was not observed after 30 minutes, travelling continued until observing a site with c. gigantea plants. in every sampling site, c. gigantea distribution (gps coordinates) occurrence data were recorded. 2.2 mapping the distribution of c. gigantea in sri lanka the bioclimatic variables were downloaded from the worldclim global climate data website (https://www.worldclim.org/). bioclimatic variables used in the present study were derived from the monthly temperature and rainfall values. biologically meaningful variables derived from the monthly temperature and rainfall (e.g., mean annual temperature, annual precipitation, annual range in temperature and precipitation) and extreme or limiting environmental factors (e.g. temperature of the coldest and warmest month, and precipitation of the wet and dry quarters) are used to represent annual trends. the above data is in the form of layers in a grid format covering the global land area. they are in the latitude/longitude coordinate reference system. this data is available in esri grid (raster) format, geo tiff format and, generic grid (raster format). it is also available at 4 different spatial resolutions; from 30 seconds (0.93 x 0.93 = 0.86 km2 at the equator) to 2.5, 5 and 10 minutes (18.6 x 18.6 = 344 km2 at the equator). environmental variables such as temperature and rainfall should be included in raster arc/info ascii grid format. the maximum entropy model (maxent) software and diva-gis software were used. in worldclim, the globe is divided into 60 squares which refer as tiles. each tile consists of an enormous data package including climatic data of different regions of the world. the global climatic data is available in 1 km2 spatial resolution approximately (fick and hijimans 2017). as original data of worldclim is available in 30 seconds spatial resolution, it was selected for the study. for this study, bioclimatic data of worldclim related to 28 tiles (india sri lanka region) was downloaded from http://www.worldclim.org/tiles.php and in addition, the generic grid (raster) format was used. for the calculations of maxent, two file types are needed. they are comma separated values (.csv) and esri ascii gis (.asc) (young et al. 2011). however, in worldclim, the data is not in the .asc format and it w.p.s.n. wijeweera et al. distribution of calotropis gigantea in sri lanka ruhuna journal of science vol 12 (2): 144 -154, december 2021 147 does not support maxent. diva-gis software was used to convert downloaded bioclimatic data of worldclim into .asc format and mapping of c. gigantea distribution of sri lanka was done (figure 1). 2.3 model suitability and validation the suitability of the model was determined by the area under the curve /auc. it represents the ratio of sensitivity vs. specificity of c. gigantea. auc makes a comparison between the performances of one model with another. in addition, auc is useful to evaluate multiple models of maxent. acu contained a possible value range 0 to 1, and values above 0.5 represent the higher predictive power while values less than 0.5 represent lower model performance. in other words, if the value of auc is closer to 1 means the mode is extremely appropriate for the predicted distribution while a value close to 0 means the model is not suitable (young et al. 2011). model validation is necessary to assess the model's overall performance and application potential (uden et al. 2015). independent data or data which is not used for model training is required for the model validation (uden et al. 2015). if the data set is sufficient, observations are taken as randomly or spatially subset into training and testing data sets. as there were sufficient observations (n=120) in this study, 20 per cent of data was set aside for testing the model (uden et al. 2015). 2.4 habitats of c. gigantea in sri lanka the patterns of land use indicate the changes in the plant habitats. as a result, landuse patterns greatly influence plant distribution (honnay et al. 1999). to study the land-use pattern, habitats of c. gigantea were recorded during the field study. habitats were categorized as roadsides without disturbances, roadsides with wastelands, roadsides with abandoned lands, roadsides near to seashore, roadsides adjacent to a cemetery, roadsides closer to reservoirs and roadsides with continuous anthropogenic activities such as railway tracks, road construction sites and cattle grazing lands. 2.5 plant density of c. gigantea to calculate plant density, a randomly selected (3m ×3m) plant patch of c. gigantea in each site was observed. c. gigantea plants in randomly selected 9 m2 areas were counted and recorded. the recorded values were ranked as high density (h), moderate density (m) and, low density (l). these ranks were given according to the number of plants that were present in 9 m2 of selected c. gigantea patch (1 to 2 plants: low density, 3 to 4 plants: moderate density, more than 4 plants: high density). percentage of c. gigantea plant density was calculated as, w.p.s.n. wijeweera et al. distribution of calotropis gigantea in sri lanka ruhuna journal of science vol 12 (2): 144 -154, december 2021 148 c. gigantea plant density % = 3 results 3.1 predicted distribution of c. gigantea in sri lanka fig 1. sampling sites and predicted probability of calotropis gigantea distribution in sri lanka using maxent modelling the probability of c. gigantea distribution according to the environmental variables of the maxent modelling technique is shown in figure 1. in figure 1, warmer colours (red and orange) show the areas with a predicted high probability of c. gigantea distribution including the whole coastal belt of the country, northern, north-central, and eastern provinces of sri lanka. in addition, a high probability of c. gigantea is predicted in certain regions of southern and north-western provinces. the low number of sampling sites having the same density rank of the selected district number of c. gigantea sampling sites per selected district ×100% w.p.s.n. wijeweera et al. distribution of calotropis gigantea in sri lanka ruhuna journal of science vol 12 (2): 144 -154, december 2021 149 probability of c. gigantea distribution (yellow and green) is predicted in regions in north-western, western, southern, uva, central and, sabaragamuwa provinces of the country. according to the prediction of the model, there is no probability of occurrence of c. gigantea (white) in certain regions of southern, sabaragamuwa, uva and, central provinces of sri lanka. 3.2 evaluation of quality of maxent model auc is used to predict the accuracy of the model. it determines whether the probability of a present location is ranked higher than a random background location or not, and reddy et al. (2015) described the ranking system in values of auc as 0.50-0.60 (fail), 0.60-0.70 (poor), 0.70-0.80 (fair), 0.80-0.90 (good), and 0.90-1.0 (excellent). fig 2. auc curve of sensitivity versus specificity for calotropis gigantea in the present study, the auc was obtained based on the potential climatic factors which affect the distribution of c. gigantea in sri lanka. the auc values were 0.971 and 0.973, for training and test data, respectively (figure 2). it indicates the constructed model to be appropriate with an ‘excellent’ predictive accuracy. therefore, it is suitable to make predictions on the geographic distribution of c. gigantea in sri lanka. by entering ‘20’ in the settings of ‘random test percentage’, the program randomly set aside 20% of the sample records for testing. the analysis utilizes a threshold to make a binary prediction with suitable conditions predicted above the threshold level (suitable) and below the threshold level (unsuitable). figure 3 indicates the omission rate and predicted area as a function of the cumulative threshold. the calculation of the omission rate was done on the training records as well as the test records (80% and 20% of the presence records, respectively). w.p.s.n. wijeweera et al. distribution of calotropis gigantea in sri lanka ruhuna journal of science vol 12 (2): 144 -154, december 2021 150 according to reddy et al. (2015), the omission rate should be close to the predicted omission. fig 3. graph of omission and predicted area for calotropis gigantea figure 3 shows how testing and training omission and predicted area for c. gigantea vary with respect to the cumulative threshold. the omission on test samples has high compatibility with the predicted omission rate. in some situations, the test omission line lies well below the predicted omission line. on the other hand, in some situations, the test omission line lies well above the predicted omission line. such conditions appear due to the dependency of test and training data, as they are derived from the same spatially auto-correlated presence data. this denotes that the maxent model is significantly better than random in the binomial test of omission and predicted area curve. 3.3 habitats of c. gigantea in sri lanka the percentage occurrence of c. gigantea according to habitat category is given in table 1. the majority (66.9%) of c. gigantea plants were located on either side of the roads with undisturbed soil. during the survey, 12.1 % of c. gigantea plant sites were recorded in dumped lands closer to roadsides while 7.3% of them were recorded at sea-shore closer to roadsides. in the dry zone, c. gigantea plants were recorded closer to reservoirs such as tanks, lakes and, estuaries. it was also recorded in cultivations such as paddy and coconut. only 4% of c. gigantea habitats were associated with roads where continuous anthological activities prevail. w.p.s.n. wijeweera et al. distribution of calotropis gigantea in sri lanka ruhuna journal of science vol 12 (2): 144 -154, december 2021 151 table 1. percentage of occurrence of c. gigantea in sri lanka according to habitat category. habitat category % occurrence of c. gigantea according to habitat roadsides without disturbances 66.9 dumped lands near roadsides 12.1 seashore near roadsides 7.3 cultivations near roadsides 6 reservoirs close roadsides 4.8 roadsides with continuous anthropogenic activities 4.0 abandoned lands close to roadsides 2.4 cemetery at roadsides 2.4 3.4. plant density of c. gigantea the percentage of plant density according to the province is recorded in table 2. c. gigantea plants are highly preferred to grow in low plant densities in northern (73.68%) and eastern (72.0%) provinces. a higher percentage of moderate plant density was recorded in western (33.33%) and north-western (31.25%) provinces. in the uva province, c. gigantea plants tend to grow as high-density mass with a peak (50%) percentage. in contrast, western and northern provinces have no sites with high c. gigantea plant density. the plant was absent in central province. table 2: percentage calotropis gigantea plant density with respect to provinces in sri lanka. province low plant density % moderate plant density % high plant density % northern 73.68 26.32 0 eastern 72.0 4.0 24.0 western 66.67 33.33 0 northcentral 57.89 15.72 26.32 southern 50.0 19.44 30.56 northwestern 37.5 31.25 31.25 uva 30.0 20.0 50.0 central 0 0 0 sabaragamuwa ne ne ne nenot estimated due to less coverage area of sabaragamuwa province 4 discussion there are many records on roadside surveys on invasive species (baard and kraaij 2019). it may be due to two reasons; invasive species tend to grow on roadsides and easy accessibility for sampling. roads act as corridors that facilitate the distribution of invasive species to introduced areas. the development of road networks and frequent road constructions further facilitate the range expansion of invasive species. therefore, road edges are ideal habitats for invasive plants which facilitate their dispersal to different geographical regions (sharma et al. 2010). w.p.s.n. wijeweera et al. distribution of calotropis gigantea in sri lanka ruhuna journal of science vol 12 (2): 144 -154, december 2021 152 according to sharma et al. (2010), anthropogenic activities on roads pave the way for the invasion of calotropis procera into introduced areas. the study also describes that road usage, vehicle smoke, and vehicle gust further facilitate seed dispersal of c. procera, and that invasion from urban areas to rural regions is possible via road systems (sharma et al. 2010). in the present study, c. gigantea is in a broad range of habitats including undisturbed roadsides, cemeteries, abandoned lands, dumped lands, seashore and, cultivations. quazi et al. (2013) also mentions that calotropis spp. tend to grow on sand dunes closer to estuaries as well as over-grazed grasslands. kumar et al. (2013) also record that, it appears as a weed in over-grazed lands where there is no competition from grasses. however, if there is grass, c. gigantea may not act as a dominant plant in open areas which is well studied for c. procera in australia. according to menge et al. (2017), c. procera is a poor competitor with respect to the native grass (mitchell grass) and fails to invade grasslands. it also prefers to grow in desert regions due to its xerophytic nature such as milky latex in leaves, highly branched root system and waxy, thick leaves (kumar et al. 2013). low population densities limit the reproductive output of the plant. pollinator limitation is a leading factor for low population densities of calotropis spp. (menge et al. 2017). in the present study, low population densities are present in the western, eastern and, northern provinces on a large scale where there is a smaller number of pollinators. personal observations reveal that highly urbanized western province contains a low number of pollinators which may be the reason for low plant density. eastern and northern regions are under road construction after the civil war and as a result, c. gigantea populations have been cut down and cleared. it may lead to low population densities in northern and eastern regions. as calotropis spp. are salt and drought tolerant and resistant to low rainfall (300400 mm) (kumar et al. 2013), they are abundant in places with similar climatic conditions such as in coastal regions of southern, northern, eastern and, northcentral provinces of sri lanka. the observations (table 1) are compatible with maxent output (figure 1) indicating climatic conditions highly affecting the c. gigantea distribution. as an example, maxent predicts a low probability of c. gigantea in the western province which belongs to the wet zone of the country. a similar observation is recorded during field visits also. in contrast, a higher percentage occurrence of c. gigantea is recorded in the eastern, north-central, northern and southern provinces of sri lanka. the maxent model also predicts the same result indicating climatic suitability for c. gigantea growth prevailing in these areas. however, the model predicts a high probability of c. gigantea distribution in the coastal belt of puttalam to mannar which is not observed during field visits (figure 1). the fragmentation of c. gigantea distribution from puttalam to mannar may be due to human activities that occurred during road construction projects after the civil war prevailed in these regions although there are ideal environmental conditions available for c. gigantea plants. maxent model also predicted, a low probability of c. gigantea from induruwa to colombo, as it belongs to the wet zone of the country. on w.p.s.n. wijeweera et al. distribution of calotropis gigantea in sri lanka ruhuna journal of science vol 12 (2): 144 -154, december 2021 153 the other hand, field observations reveal that clearance of the c. gigantea coverage occurs due to high urbanization. the combined effect of lack of climatic suitability and urbanization may have led to the disappearance of c. gigantea from induruwa to colombo. auc values indicate that maxent produced significantly accurate results. the sensitivity versus 1-specificity graph indicated that the maxent model got an excellent predictive accuracy (mean auc = 0.972) concerning the relationship between the distribution of c. gigantea and the selected environmental variables. the results show that the maxent model can be used to study the climatic suitability for the distribution of c. gigantea in sri lanka. it acts as a tool to understand the potential distribution of c. gigantea in sri lanka. most of the observed results in table 2 are compatible with the predicted distributions in figure 1. field observations revealed high plant occurrence in the northern, northcentral and, eastern provinces of the country which is compatible with the model prediction. similarly, the model predicts a low probability of c. gigantea distribution in large land areas of uva, north-western, western and central provinces of sri lanka which is observed during field visits also (table 2). in addition, the model predicts a low probability of c. gigantea distribution in sabaragamuwa province where the percentage of occurrence is not evaluated due to lack of land coverage during field visits (table 2). therefore, the maxent model facilitated the prediction of species distributions where there is a lack of occurrence data. 5 conclusions the present study provides a detailed map of c. gigantea as well as detailed information on c. gigantea plant distribution, plant density and, habitats of the plant in sri lanka. the knowledge of the distribution of c. gigantea is important as it has medicinal value. on the other hand, the association of environmental factors for its distribution is greatly important to control its invasiveness in the introduced range. the present study would be the first of its kind in sri lanka using maxent to evaluate climatic suitability on c. gigantea. maxent modelling predicted the distribution of the plant within the whole country concerning environmental variables. according to the maxent, climatic factors highly influence on the distribution of c. gigantea. in addition, anthropological activities also play a considerable role in c. gigantea distribution. the study also provides occurrence data of c. gigantea in sri lanka which will be greatly important for further studies on its distribution in sri lanka as well as in the world. acknowledgements two anonymous reviewers are acknowledged for valuable comments on the initial draft of the manuscript. w.p.s.n. wijeweera et al. distribution of calotropis gigantea in sri lanka ruhuna journal of science vol 12 (2): 144 -154, december 2021 154 references austin m. 2002. spatial prediction of species distribution: an interface between ecological theory and statistical modelling. ecological modelling 157: 101-118. azarkhavarani sf, rahimi m, bernardi m. 2015. the most important climatic factors affecting distribution of zygophyllum atriplicoides in semi-arid region of iran (case study: isfahan province). desert 20 (2): 145-156. barbet-massin m, rome q, villemant c, courchamp f. 2018. can species distribution models really predict the expansion of invasive species? plos one 13(3): e0193085. baard ja, kraaij t. 2019. use of a rapid roadside survey to detect potentially invasive plant species along the garden route, south africa. koedoe 61(1): 1-10. bramer i, anderson bj, bennie j, bladon aj, de frenne, p, hemming d, hill ra, kearney mr, korner c, korstjens ah, lenoir j, maclean imd, marsh cd, marecroft md, ohlemuller r, slater hd, saggitt aj, zellweger f, gillingham pk. 2018. advances in monitoring and modelling climate at ecologically relevant scales. advances in ecological research 58: 101-161. dhileepan k. 2014. prospects for the classical biological control of calotropis procera (apocynaceae) using coevolved insects. biocontrol science and technology 24(9): 977-998. fick se, hijmans rj. 2017. worldclim 2: new 1-km spatial resolution climate surfaces for global land area. international journal of climatology, 37 (12). honnay o, hermy m, coppin, p. 1999. impact of habitat quality on forest plant species colonization. forest ecology and management 115 (2-3): 157-170. kumar d, kumar s. 2015. calotropis gigantea (l.) dryand – a review update. indian journal of research in pharmacy and biotechnology 3(3): 218-230. kumar p, suresh e, kalavath s. 2013. review on a potential herb calotropis gigantea (l.) r. br. scholars academic journal of pharmacy 2(2): 135-143. menge eo, bellairs sm, lawes mj. 2017. disturbance-dependent invasion of the woody weed, calotropis procera, in australian rangelands. the rangeland journal 39: 201-211. menge eo, stobo-wilson a, oliveira sl, lawes mj. 2016. the potential distribution of the woody weed calotropis procera (aiton) w.t. aiton (asclepiadaceae) in australia. the rangeland journal 38(1): 35-46. merow c, smith mj, silander ja. 2013. a practical guide to maxent for modelling species’ distributions: what it does, and why inputs and settings matter. ecography 36(10): 1058-1069. naiman rj, decamps h, mcclain me, likens ge. 2005. biotic functions of riparia, california, 1st edition, elsevier academic press, california, 1-488pp. pidwirny m. 2018. spatial distribution of species and ecosystems. british columbia, our planet earth publishing quazi s, mathur k, arora s. 2013. calotropis procera: an overview of its phytochemistry and pharmacology. indian journal of drugs 1(2): 63-69. reddy m, begum h, sunil n, pandravada s, sivara n. 2015. mapping the climate suitability using maxent modelling approach for ceylon spinach (basellaalba l.) cultivation in india. the journal of agricultural sciences 10(2): 87-97. sharma gp, kumar m, raghubanshi as. 2010. urbanization and road-use determines calotropis procera distribution in the eastern indo-gangetic plain, india. ambio 39(2): 194–197. uden dr, allen cr, angeler dg, corral l, fricke, ka. 2015. adaptive invasive species distribution models: a framework for modelling incipient invasions. biological invasions 17 (10): 831-2850. young n, carter l, evangelista p. 2011. a maxent model v3.3.3e tutorial (arcgis v10). colarado: colarado state university. microsoft word rjs-vol-2-wijetunga-etal.doc.doc © 2007 faculty of science university of ruhuna ruhuna journal of science vol. 2, september 2007, pp. 96-110 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x 96 abstracts removal mechanisms of acid red 131, acid yellow 79 and acid blue 204 dyes with different chemical groups under anaerobic process using mixed anaerobic granular sludge were studied. the uv-visible spectrum obtained and dissolved residual chemical oxygen demand measured at the end of incubation suggest that acid red 131 and acid yellow 79 were biodegraded and no further degradation of dye metabolites have occurred. acid red 131 and acid yellow 79 were decolourized by biodegradation achieving 81% and 97% of colour removal at 300 mgl-1 dye concentrations, respectively. acid red 131 biodegradation followed the first order kinetic model with respect to dye concentration while acid yellow 79 biodegradation followed the second order kinetic model. although over 90% of colour was removed in acid blue 204 dye, no biodegradation was noticed. it is clear that different dyes can be decolourized under anaerobic condition even though mechanism of decolourization is not same. keywords: acid dyes; anaerobic; biodegradation; decolourization; kinetics; textile wastewater; 1. introduction wastewater generated from various industries creates severe detrimental effects to the environment leading to bio-systems imbalance. the textile industry, which is one of the largest water consumers in the world, produces wastewater composed of various recalcitrant agents such as dye, sizing agents, and dyeing aids, that should be of concern in releasing into the environment. the removal of colour is needed to be considered in the disposal of textile wastewater due to aesthetic deterioration as well as the obstruction of penetration of dissolved oxygen and sunlight into water bodies, which seriously affects aquatic life. besides, the dye precursors and degradation products are proven carcinogenic and mutagenic in nature (kalyuzhnyi and sklyar, 2000). it is estimated that 10-20% of dye produced is lost into wastewater during textile dyeing process (graça et al., 2001).the largest class of dyes used at present is referred to as acid dyes, which are anionic compounds mainly used for dyeing nitrogen containing fabrics like wool, polyamine and silk. out of 12 classes of different chromogenic groups of dyes, most common group is azo group, which contributes up to 70% of all textile dyes produced, followed by anthraquinone group (vandevivere et al., 1998). azo dyes, characterized by their typical azo bond (r1-n=n-r2), which is responsible for colour, are a more popular dye group used in the textile industry and due to their poor exhaustion properties, as removal mechanisms of acid dyes of different chemical groups under anaerobic mixed culture s. wijetunga1, l. xiufen, r. wenquan, j. chen 1department of agricultural engineering, faculty of agriculture, university of ruhuna, kamburupitita, sri lanka. laboratory of environmental biotechnology, school of biotechnology, southern yangtze university, no 170, huihe road, wuxi 214036, pr china correspondence: : 1 swije@ageng.ruh.ac.lk, swije2001@yahoo.com wijetunga et al.: removal mechanisms. . . 9 7 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100( 2 0 0 7 ) much as 30% of initial dye applied remain unfixed and end up in effluents (manu and sanjeev, 2003). anthraquinone dye chromophore is characterized by its carbonyl group, and it may be present once or repeated times. the colour of anthraquinone dyes is partially associated with the anthraquinone nucleus and is modified by the type, number, and position of the substitutes. the colour of wastewater decreases when the cleavage of -n=n-, -c=cbonds and heterocyclic and aromatic rings occurs. then, absorption of light by the associated molecules shifts from visible to uv or infrared region of the electromagnetic spectrum (slokar et al., 1998). even though there are various physical and chemical wastewater treatment methods that can be applied for textile wastewater treatment, the applicability of those methods are limited due to various limitations. advanced oxidation process such as fenton’s regent (h2o2 and fe2+), h2o2, and ozonation are costly in terms of operational costs (stanislaw and monika, 1999). coagulation and flocculation with lime, alum, polyelectrolyte, and ferrous salts produce large amount of sludge, which impose handling and disposal problems (huseyin, 2005; georgiou et al., 2003; daneshvar et al., 2003). the adsorption of dyes using activated carbon and various other absorbents are also costly (zhemin et al., 2001). the photochemical oxidation of dyes using uv and sunlight with oxidation agents like catalyses, h2o2 are also not economically viable (muruganandham and swaminathan, 2004). most of the studies with respect to decolourization of textile wastewater containing dyes have been focused on azo dyes and have shown that they could be effectively decolourized by biological means, such as anaerobic and combined anaerobic/aerobic processes (frank et al., 2001; manu and sanjeev, 2003; méndezpaz et al., 2005; andré et al., 2004; georgiou et al., 2004; rui et al., 2001; pearce et al., 2003). the information on decolourization and biodegradation of dyes having different chemical groups under anaerobic process are inadequately reported. therefore, the objectives of this study were to find removal mechanisms of different acid dyes performed by mixed anaerobic consortia and compare them with respect to each dye, and monitor the inhibition caused to microorganisms by acid dyes widely used in the textile industry. furthermore, the dye removal kinetics was also evaluated. 2. materials and methods 2.1 laboratory scale batch reactors and inocula the anaerobic batch reactors, glass vessels of 500 ml capacity having sample ports for withdrawing samples for the analysis of colour and chemical oxygen demand (cod), were used in the study. the reactors were seeded by anaerobic granular sludge obtained from a laboratory scale upflow anaerobic sludge blanket (uasb) reactor, which had been steadily operated for six months in synthetic wastewater. the total suspended solids (tss) and the total volatile suspended solid (tvss) of granules were 104.74 gl-1 and 84.90 gl-1, respectively. 98 wijetunga et al.: removal mechanisms. . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100 ( 2 0 0 7 ) 2.2 incubation medium the incubation medium was prepared according to ergüder and co-workers (2003) with minor modifications. the composition of medium consisted of following inorganic compounds (mgl-1): nh4cl, 1200; mgso4.7h2o, 400; kcl, 400; cacl2.4h2o, 50; (nh)2hpo4, 170; fecl2.4h2o, 40; cocl2.6h2o, 10; ki, 10; mncl2.4h2o, 0.5; cucl2.2h2o, 0.5; zncl2, 0.5; alcl3.6h2o, 0.5; na2moo4.2h2o, 0.5; h3po3, 0.5; nicl2.6h2o, 0.5; nawo4.2h2o, 0.5; and na2seo3, 0.5. the sufficient buffering capacity in reactors was maintained by nahco3 4500 mgl-1. glucose concentration of 2500 mgl-1 as cod was used as the carbon source as well as to provide reducing equivalents required to breakdown the azo bond in the case of azo dyes (georgiou et al., 2004). 2.3 chemicals used all chemicals used in this study were of analytical grade (ar), except for gas chromatography (gc), where gc grade chemicals were used. the dyes; c.i. acid red 131 (telon red m-3b belongs to azo dye group), c.i acid blue 204 (telon blue m-rlw belongs to anthraquinone dye group) and c.i acid yellow 79 (telon yellow m-4gl, of which the chemical group of dye has not been defined) were of commercial grade and used for this study without further purification. 2.4 experimental procedure four levels of dye concentration of 50, 100, 150 and 300 mgl-1 of each dye were studied under anaerobic condition at 37 °c in a dark incubator and a control reactor without dye was also incubated to find out the effect of dye on anaerobic microorganisms. the reactors were seeded with 5.1 g anaerobic granules in tvss and other macro, micro, buffering agent, and a carbon source was added in aforementioned concentrations. before keeping in the incubator, headspace of the reactors was flushed with nitrogen gas for 5 minutes. abiotic assays using autoclaved biomass were also done in order to determine the dye removal by nonbiological processes. 2.5 analytical procedure total alkalinity, ph, tss and tvss were determined according to the procedure outlined in standard methods (apha, 1985). chemical oxygen demand was determined spectrometrically by digestion in 5b-1 quick cod analyzer (the lianhua environmental instrument institute, langzhou, pr china). volatile fatty acids (vfa) at the end of the incubation were determined by shimadzu gc-14a gas chromatography with peg-20m capillary column (sge international, australia), 30 m length and internal diameter 0.53 mm. prior to injection, samples were acidified with 1 m h2so4, centrifuged for 15 minutes at 13000 rpm, and filtered through 0.22 µm filter (tahar and sami, 2005). the column temperature was initially set at 100 °c for 1 minute and then increased at a rate of 5 °c per minute up to 240 °c and held for 20 minutes. the injector and detector temperature wijetunga et al.: removal mechanisms. . . 9 9 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100( 2 0 0 7 ) were set at 240 °c and 250 °c, respectively. nitrogen was used as the carrier gas at a flow rate of 1.5 ml min-1. the colour of the samples was measured spectrometrically using the maximum absorbance of each dye by uv-visible spectrometer (shimadzu uv-2450). the samples were filtered through microfiber filter and centrifuged at 7000 rpm for 10 minutes prior to absorbance measurements. the residual dyes after incubation were quantified by standard curves based on the maximum absorbance of respective dyes. colour removal was determined according to eq. (1). 100 a aa 0 t0 × − (1) where a0 is absorbance at the beginning of incubation and at is absorbance at selected time. the values given are the mean values of independent replicates and error bars show standard deviation. 3. results and discussion 3.1 primary removal mechanism of c.i. acid red 131 by anaerobic consortia the maximum absorbance wavelength and uv-visible spectrum were scanned at the beginning and end of incubation (where more than 90% of colour was removed and no more changes of uv spectrum were observed after this period) to find out whether the colour removal is merely physical adsorption of dyes onto granules or biodegradation by microbial granules. the uv-visible spectrum of c.i. acid red 131 (ar131) dye containing medium before and after incubation of 45 hours are shown in fig.1. the remarkable changes of uv-visible spectra could be seen in the samples obtained after incubation and explained by the structural modifications of the dye molecules performed by anaerobic microorganism. changes in the spectrum especially in uv region provide the evidences of degradation of dye molecules (manu and sanjeev, 2003; rui et al., 2001; pinheiro et al., 2004; robert and sanjeev, 2005). 100 wijetunga et al.: removal mechanisms. . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100 ( 2 0 0 7 ) figure.1 uv-visible spectrum obtained before and after incubation (300mg l-1) ar131 acid red 131 showed its maximum absorbance in visible region (400-800 nm) at 548 nm before incubation. however, after incubation for 45 hours, the maximum absorbance peak disappeared and a new peak appeared in uv region at 366 nm, thereby confirming biodegradation of the dye. at the same time, red colour of the dye disappeared and a strong straw colour appeared indicating biodegradation of ar131 after the anaerobic incubation. the previous studies showed that anaerobic digestion of several azo dyes produce relevant amines (or aromatic amines) by cleavage of azo bond (andré et al., 2004; georgiou et al., 2004). according to pinheiro and co-workers (2004), the peak resulted at 270 nm after incubation is probably due to aromatic amines produced after biodegradation of ar131 by anaerobic consortia. the emergence of the new colour might be due to partial break down of azo bond or breakdown products. similar observations have been noted by frank et al. (2001) in their study of reactive red 2 and reactive red 4, azo dyes, under anaerobic process. the colour of these dyes had been shifted to yellow under anaerobic process. the aromatic amines produced are not generally mineralized to co2 and ch4 by anaerobic process, with the exceptions of a few aromatic amines characterized by the presence of hydroxyl and or carboxyl groups (elías et al., 1996; razo-flores et al, 1997). consequently, ar131 was probably not mineralized and it was confirmed by uv-visible spectra and cod obtained after incubation. wijetunga et al.: removal mechanisms. . . 1 0 1 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100( 2 0 0 7 ) 0 500 1000 1500 2000 2500 3000 0 20 40 60 80 100 incubation time (h) r es id ua l c o d (m g/ l) without dye 100 mg/l 300 mg/l figure.2 residual dissolved cod at different ar131 dye concentrations the residual dissolved cod in the reactors containing ar131 under 0 mgl-1, 100 mgl-1 and 300 mgl-1 dye concentration with respect to time are shown in fig. 2. it could be seen that more than 90% of cod was removed in all reactors after 88 hours of incubation, except 300 mgl-1 dye concentrations, where ca. 84% cod was removed. it appears that cod removal declined as dye concentration increased. chemical oxygen demand in controlled reactor was ca. 95 mgl-1 while it was 450 mgl-1 at 300 mgl-1 of dye concentration. this indicates that substrate was degraded faster without ar131 or with its low concentrations. it was also confirmed by substrate removal kinetics. the substrate removal seems to be followed first order kinetics and the first order kinetic constants decreased from 0.0353 h-1 in dye free reactor to 0.0193 h-1 in reactors supplemented with 300 mgl-1 of dye, suggesting a slight inhibition on substrate removal. the dye contribution to feed cod was ca.1200 mg cod per gram of ar131 dye, so the measured residual cod indicates that the dye metabolites seem to contribute same cod as the initial dye. it means that considerable mineralization of dye metabolites have not taken place. similar observations have reported by mustafa and delia (2005) and méndez-paz and coworkers (2005) with regard to azo dye degradation under anaerobic conditions. 3.2 primary removal mechanism of c.i. acid yellow 79 by anaerobic consortia the maximum absorbance of c.i acid yellow 79 (ay79) before incubation was observed at 400 nm, while it disappeared after incubation leading to decolourization (fig.3). the substantial changes of uv-visible spectrum can be seen after incubation. the new peak at about 336 nm emerged and it might be due to degraded products of ay79. at the same time, it can be concluded that ay79 could be biodegraded under anaerobic condition by mixed anaerobic consortia. no new 102 wijetunga et al.: removal mechanisms. . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100 ( 2 0 0 7 ) colour developments were observed in ay79 containing effluents. the effluents were colourless. figure.3 uv-visible spectrum obtained before and after incubation of ay79 (300mg l-1) with anaerobic consortium more than 90% of cod removal was achieved in all batch reactors except in 300 mgl-1 concentration, where 87% cod was removed indicating cod removals tend to decrease as increasing dye concentration in influent (fig. 4). cod in controlled reactor was ca.111 mgl-1 and it was 345 mgl-1 in reactors containing 300 mgl-1 dye. cod were not considerably decreased after 72 hours indicating almost all cosubstrate cod provided by glucose was exhausted. cod removal was high in reactors with low or without dye, exhibiting higher dye concentrations inhibited cod degradation. the substrate removal followed the first order kinetic model and kinetic constant was decreased from 0.0393 h-1 to 0.0266 h-1 in 0 to 300 mgl-1 dye concentrations, respectively, confirming the little inhibition on substrate removal. the dye itself contributed ca. 848 mg cod per gram of ay79 dye, so the most dissolved cod at the end of the study was provided by the dye metabolites suggesting that the dye was not completely degraded by anaerobic microorganisms. wijetunga et al.: removal mechanisms. . . 1 0 3 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100( 2 0 0 7 ) 0 500 1000 1500 2000 2500 3000 0 20 40 60 80 incubation time (h) r es id ua l d is so lv ed c o d (m g/ l) without dye 100 mg/l 300 mg/l fig.4 residual dissolved cod at different ay79 dye concentrations 3.3 primary removal mechanism of c.i. acid blue 204 by anaerobic consortia the dye c.i. acid blue 204 (ab204) seems to be resistant to biodegradation by anaerobic consortia and uv-visible spectrum obtained reveals that the colour removal observed was a result of physical adsorption (fig. 5). figure.5 uv-visible spectrum obtained before and after incubation of ab204 (300mg l-1) with anaerobic consortium decolourization of dyes by microbes occurs by biodegradation or adsorption to microbial cells or both. in adsorption, examination of uv-visible spectrum will 104 wijetunga et al.: removal mechanisms. . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100 ( 2 0 0 7 ) reveal that all peaks decrease approximately in proportion to each other. if dye removal is attributed to biodegradation, either the major visible light absorbance peak will completely disappear or a new peak will appear (kuo-chen et al., 2003). the prominent peaks that contributed to the colour of ab204 can be seen at 625 and 580 nm, and these peaks can be seen in both media before and after incubation, suggesting that decolourization is due to physical adsorption onto microbial granules. this was confirmed by observation of microbial granules after incubation. the granules were dark in colour and when they were washed slightly, colour became visible in water. it seems that a greater removal of colour of ab204 was by adsorption onto microbial granules. several authors have reported that most anthraquinone dyes studied under anaerobic process are quite resistant to biodegradation. thongchai and worrawit (2000) have reported that anthraquinone dye, reactive blue 5 and reactive blue 19 were not biodegraded under anaerobic processes but the decolourization of these dyes was observed. it was merely as a result of adsorption on to bacterial floc materials. anthraquinone dyes, brillant red resolin bls, c.i. acid blue 40 and tectilon blue 4r-01 have been examined under anaerobic conditions and reported that except tectilon blue 4r-01, where single microbial strains were used, other dyes are resistant to biodegradation and colour removal observed only by adsorption (malpei et al., 1998; stanislaw et al., 2001; walker et al., 2000). in this study, ab204 was not biodegraded but considerable colour removal was observed and the slight disintegration of granules after incubation with this dye was also noticed. the decolourization or removal mechanisms of different dyes under anaerobic process are different. therefore, this study shows the possibility of the use of anaerobic process for the treatment of textile wastewater containing different types of dyes with different chemical groups. it has to be realized that some dyes, such as anthraquinone dye ab204 could not be biodegraded although high colour removal can be achieved. 0 500 1000 1500 2000 2500 3000 0 20 40 60 80 100 incubation time (h) r es id ua l d is so lv ed c o d (m g/ l) without dye 100 mg/l 300 mg/l wijetunga et al.: removal mechanisms. . . 1 0 5 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100( 2 0 0 7 ) figure.6 residual dissolved cod at different ab204 dye concentrations over 90% of cod removal was observed in 100 mgl-1 of ab204 dye concentration (fig.6). chemical oxygen demand concentrations at the end of the incubation were 89 and 420 mgl-1 in reactors without dye and with 300 mgl-1 dye, respectively. since, dye contribution for dissolved cod is ca. 657 mg cod per gram of ab204 dye, the residual dissolved at the end of the study was primarily supplied by the dye. the rate of low cod depletion at later stage of incubation with ab204 dye might be owing to the toxic effects of dye on anaerobic consortia. the probable reason for toxicity could be that the active sites of enzymes involved in co-substrate degradation is occupied by anthraquinone ab204 dye molecules, thus, blocking the binding of substrate and cofactors (prestera et al., 1992). another proposed inhibition mechanism is the uncoupling of electron transfer from atp synthesis via an anthraquinone-mediated electron transfer reaction (cooling iii et al., 1996). however, from this study, it cannot be justified that the above one or both mechanisms are involved in the inhibition process till the end of incubation, since cod at the end of study was not high in comparison to controlled reactor, suggesting that no high accumulation of glucose or vfa in reactors. therefore, inhibition extended by ab204 only caused to slow down the cod depletion process but it does not affect long-term inhibition. 3.4 decolourization kinetics of ar131, ay79 and ab204 dyeing wastewater decolourization rate constants, k0, k1 and k2 for zero, first and second order kinetic models were determined by fitting the data obtained during study period to eq. (2), (3) and (4), respectively. tkdd 00t −= (2) t1k 0t edd −= (3) tk d 1 d 1 2 0t += (4) where dt is the dye concentration in the medium at a given time; d0 the dye concentration at time 0; k0 zero order rate constant; k1 first order rate constant; k2 second order rate constant and t is the time. rate constants determined and r2 values are given in table 1. table 1. decolourization rate constants for ar131 and ay79 dyes constant 50 mgl-1 100 mgl-2 150 mgl-3 300 mgl-1 k0 (mgl -1h-1) 0.8788 1.8111 2.7372 5.2168 r2 0.79 0.78 0.80 0.75 k1 (h -1) 0.0497 0.0560 0.0559 0.0517 r2 0.97 0.97 0.98 0.97 k2 (lmg -1h-1) 0.0040 0.0026 0.0017 0.0008 ar131 r2 0.96 0.95 0.90 0.95 k0 (mgl -1h-1) 0.9055 1.9829 3.0247 6.3753 r2 0.52 0.59 0.60 0.65 ay79 k1 (h -1) 0.0543 0.0735 0.0792 0.0932 106 wijetunga et al.: removal mechanisms. . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100 ( 2 0 0 7 ) r2 0.74 0.88 0.90 0.96 k2 (lmg -1h-1) 0.0050 0.0049 0.0039 0.0030 r2 0.93 0.99 0.99 0.94 figure.7 variation of residual dye levels of (a) ar131, (b) ay79 and (c) ab204 during incubation period the residual dyes of different reactors having different dyes after incubation are shown in fig. 7. as it can be seen, all levels of ar131 dye concentrations reduced to ca.10 mgl-1 except in 300 mgl-1 dye concentration (fig. 7(a)). the color removals were almost 90% in all reactors excluding 50 mgl-1 dye where comparatively lower removal was noticed with 87%. it seems that the removal mechanism of different dyes under anaerobic process is the physical adsorption followed by biodegradation. at the lower dye concentrations, dye adsorbed onto microbial granules appears to be lower than high dye concentrations providing less opportunity for biodegradation and removal. the residual dye estimated in the reactors of 300 mgl-1 of ar131 dye concentration is ca. 25 mgl-1 after 45 hours of incubation. even after 300 hours of incubation (not shown), no further reduction of residual dye was observed ay79 (b) 0 50 100 150 200 250 300 350 0 10 20 30 40 incubation tim e (h) r es id u al d ye ( m g /l) 50 mg/l 100 mg/l 150 mg/l 300 mg/l ar131 (a) 0 50 100 150 200 250 300 350 0 10 20 30 40 50 incubation time (h) r es id u al d ye ( m g /l) 50 mg/l 100 mg/l 150 mg/l 300 mg/l ab204 (c) 0 50 100 150 200 250 300 350 0 50 100 incubation tim e (h) r es id u al d ye ( m g /l) 50 mg/l 100 mg/l 150 mg/l 300 mg/l wijetunga et al.: removal mechanisms. . . 1 0 7 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100( 2 0 0 7 ) suggesting that the long incubation period does not bring additional decolourization. abiotic test performed with autoclaved anaerobic granules removed 10-14% of colour and it probably is as a result of adsorption. this implies that physical adsorption of dyes onto anaerobic granules is not the primary mechanism of colour removal of ar131 dye. the r2 values (>0.95) indicated that colour of ar131 was removed according to first order kinetic model. the highest first order rate constant, 0.0560 h-1 was observed at 100 mgl-1 of ar131 dye concentration, whereas it was 0.0559 and 0.0517 h-1 at 150 and 300 mgl-1 dye concentrations, respectively (table 1). decolourization rate seems to be decreased slightly with increased dye concentrations. first order kinetics of degradation by anaerobic granules with respect to dye concentration from 60 mgl-1 to 3200 mgl-1 with different azo dyes has been reported by several researchers (mustafa and delia, 2004; frank et al., 2001; méndez-paz et al., 2005; robert and sanjeev, 2005; pearce et al, 2003). andré and co-workers (2004) have stated that mordant yellow 10 followed zero order models. it appears that different azo dyes may follow different degradation rates with regard to dye concentration. the rate and the amount of decolourization of azo dyes depend on their molecular structure, molecular weight, substitution groups of the dye molecule and the intermolecular hydrogen bond between azo and hydroxyl groups (nuttapum et al., 2004). acid yellow 79 was incubated up to 36 hours where almost total decolourization was achieved with more than 95% of colour removal, except in 50 mgl-1 dye concentration, where only 90% removal was attained (fig.7(b)). the residual dye concentration for each initial concentration was less than 10 mgl-1 at the end of incubation. biodegradation rate constants, determined based on the data obtained during the study, are also shown in table 1. acid yellow 79 biodegradation with respect to time, seems to follow the second order reaction kinetic model, where r2 is higher than that of zero and first order kinetics. the initial biodegradation was very high; more than 70% of colour was removed after 6 hours of incubation, except 300 mgl-1 dye concentration, where 60% was removed. after 6 hours of incubation, colour removal rate appears to have declined. it may be due to the less availability of dye in the medium due to being adsorbed into microbial granules for further biodegradation. therefore, the remaining dye dissolved in the supernatant could probably be degraded by microbes present there and not by the anaerobic granules. as aforementioned, no biodegradation of ab204 was observed during 88 hours of incubation period but more than 90% decolourization was observed (fig.7(c)). the residual dyes in reactors are almost constant after 12 h of incubation, indicating that the adsorption of ab204 dye onto anaerobic granules reaches equilibrium within 12h. although the main mechanism of colour removal of ab204 appears to be physical adsorption, abiotic study conducted using autoclaved granules yielded only ca. 16% removal of color in all reactors. reasons for these phenomena might be the dense structure of live anaerobic granules that can absorb more dye into cell mass, and dye precipitation by the actions of granules. on autoclaving, the structure of 108 wijetunga et al.: removal mechanisms. . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100 ( 2 0 0 7 ) granules is destroyed and it appears loose, causing poor adsorption. the similar results have been observed by stanislaw et al. (2001) and reported that viable cells of activated sludge can remove more colour than dead cell mass in anthraquinone dye (c.i. acid blue 40) but they did not find any biodegradation. they noticed that adsorption and dye precipitation were the main mechanism of colour removal. 4. conclusions the uv-visible spectrum obtained and dissolved residual cod measured at the end of incubation suggest that ar131 and ay79 could be biodegraded and no further degradation of dye metabolites occurred. however, both dyes were decolourized by anaerobic consortia achieving 81% and 97% of colour removal at 300 mgl-1 dye concentrations, respectively. the ar131 biodegradation followed first order kinetic model with respect to dye concentration while ay79 biodegradation process approximates to second order kinetic model. no evidences were found for biodegradation of ab204 by anaerobic process under examined conditions. however, almost 90% of colour caused by ab204 dye was removed by adsorption. since real textile wastewater contains many kinds of auxiliary chemicals except dyes; further studies are needed to find out the overall effect of these chemicals on decolourization and degradation of dyes by anaerobic process before application. references apha, standard methods for the examination of water and wastewater. 16th edn. apha-awwa and wpcf, washington dc; 1985 andré, b.d.s.; iemke, a.e.b.; francisco, j.c.; jules, b.v.l. effects of different redox mediators during thermophilic azo dye reduction by anaerobic granular sludge and comparative study between mesophilic (30°c) and thermophilic (55°) treatments for decolourisation of textile wastewaters. chemosphere 55: 1149-1157; 2004 cooling iii, f.b.; maloney, c.l.; tabinowski, j.; odom, j.m. inhibition of sulfate respiration by 1,8-dihydroxyanthraquinone and other anthraquinone derivatives. appl. environ. microbiol. 62: 2999-3004; 1996 daneshvar, n.; ashassi, s.h.; tizpar, a. decolorization of orange ii by electrocoagulation method. seperation and purification technology 31(2): 153162; 2003 elías, r.f.; brian, d.; jim, f.; gatze, l. biodegradation of n-substituted aromatics and alkyl phenols under methanogenic conditions using granular sludge. wat. sci. tech. 33(3): 47-57; 1996 ergüder, t.h.; guven, e.; demirer, g.n. inhibitory effects of lindane in batch and upflow anaerobic sludge blanket reactors. chemosphere 50: 165-169; 2003 frank, p.v.z.; lettinga, g.; field, j.a. azo dye decolourization by anaerobic granular sludge. chemosphere 44: 1169-1176; 2001 wijetunga et al.: removal mechanisms. . . 1 0 9 r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100( 2 0 0 7 ) georgiou, d.; aivazidis, a.; hatiras, j.; gimouhopoulos, k. treatment of cotton textile wastewater using linme and ferrous sulfate. wat. res. 37: 2248-2250; 2003 georgiou, d.; metallinou, c.; aivasidis, a.; voudrias, e.; gimouhopoulos, k. decolorization of azo-reactive dyes and cotton-textile wastewater using anaerobic digestion and acetate-consuming bacteria. biochem. eng. j. 19: 7579; 2004 graça, m.b.s.; maria, c.f.; pessoa, d.a.m.t. decolorization of an anthraquinonetype dye using a laccase formulation. bioresource technol. 79: 171-177; 2001 huseyin, s. decolorization and detoxification of textile wastewater by ozonation and coagulation. dyes pigments 64: 217-222; 2005 kalyuzhnyi, s.; sklyar, v. biomineralization of azo dyes and their breakdown products in anaerobic-aerobic hybrid and uasb reactors. wat. sci. technol. 41(12): 23-30; 2000 kuo-chen, c.; jane-yii, w.; dar-jen, l.; sz-chwun, j.h. decolorization of the textile dyes by newly isolated bacterial strains. j. biotechnol. 101: 57-68; 2003 malpei, f.; andreoni, v.; daffonchio, d.; rozzi, a. anaerobic digestion of print paste: a preliminary screening of inhibition by dyes and biodegradability of thickeners. bioresource technol. 63: 49-56; 1998 manu, b.; sanjeev, c. decolorization of indigo and azo dyes in semicontinuous reactors with long hydraulic retention time. process biochem. 38: 1213-1221; 2003 méndez-paz, d.; omil, f.; lema, j.m. anaerobic treatment azo dye acid orange 7 under batch conditions. enzyme microb. tech. 36: 264-272; 2005 muruganandham, m.; swaminathan, m. photochemical oxidation of reactive azo dye with uv-h2o2 process. dyes pigments 62(3): 269-275; 2004 mustafa, i.; delia, t.s. a batch kinetic study on decolorization and inhibition of reactive black 5 and direct brown 2 in an anaerobic mixed culture. chemosphere 55: 119-128; 2004 mustafa, i.; delia, t.s. a batch study for assessing the inhibition effect of direct yellow 12 in a mixed methanogenic culture. process biochem. 40: 1053-1062; 2005 nuttapum, s.; kanchana, j.; somsak, d.; marie-line, d.; pierre, s. microbial decolorization of reactive azo dyes in a sequential anaerobic-aerobic system. chem. eng. j. 99: 169-176; 2004 pearce, c.i.; lloyd, j.r.; guthrie, j.t. the removal of colour from textile wastewater using whole bacterial cells: a review. dyes pigments 58: 179-196; 2003 pinheiro, h.m.; touraud, e.; thomas, o. aromatic amines from azo dye reduction: status review with emphasis on direct uv spectrometric detection in textile industry wastewaters. dyes pigments 61: 121-139; 2004 prestera, t.; prochaska, j.; talalay, p. inhibition of nad(p)h: (quinone-acceptor) oxireductase by cibracon blue and related anthraquinone dyes: a structureactivity study. biochemistry 31: 824-833; 1992 110 wijetunga et al.: removal mechanisms. . . r u h u n a j o u r n a l o f s c i e n c e 2 , p p . 96-100 ( 2 0 0 7 ) razo-flores, e.; luijten, m.; dolon, b.a.; lettinga, g.; field, j.a. complete biodegradation azo dye azodisalicylate under anaerobic conditions. environ. sci. technol. 31:2098-2103; 1997 robert, m.; sanjeev, c. adsorption and biological decolourisation of azo dye reactive red 2 in semicontinuous anaerobic reactors. process biochem. 40: 699-705; 2005 rui, b.; isabel, m.a.f.; helena, m.p.; isolina, c.g. batch tests for assessing decolurisation of azo dyes by methanogenic and mixed cultures. j. biotechnol. 89: 155-162; 2001 slokar, y.m.; majcen, a.l.m. methods of decolourisation of textile wastewaters. dyes pigments 37(4): 335-356; 1998 stanislaw, l.; monika, s.; renata, z. biodegradation, decolourisation and detoxification of textile wastewater enhanced by advanced oxidation process. j. biotechnol. 89: 175-184; 2001 stanislaw, l.; monika, g. optimization of oxidants dose for combined chemical and biological treatment of textile wastewater, wat. res. 33(11): 2511-2516; 1999 tahar, m.; sami, s. evaluating process imbalance of anaerobic digestion of olive mill wastewaters. process biochem. 40(1): 139-145; 2005 thongchai, p.; worrawit, l. decolorization of reactive dyes with different molecular structures under different environmental conditions. wat. res. 34(17): 4177-4184; 2000 vandevivere, p.c.; bianchi, r.; verstraete, w. treatment and reuse of wastewater from the textile wet-processing industry: review of emerging technologies. j. chem. technol. biotechnol. 72: 289-302; 1998 walker, g.m.; weatherley, l.r. biodegradation and biosorption of acid anthraquinone dye. environ. pollut. 108: 219-223; 2000 zhemin, s.; wenhua, w.; jinping, j.; jianchang, y.; xue, f.; an, p. degradation of dye solution by an activated carbon fiber electrode electrolysis system, j. hazard. mater. 84(1): 107:116; 2001 sv-lncs ruhuna journal of science vol 7: 58-63, december 2016 eissn: 2536-8400  faculty of science university of ruhuna  faculty of science, university of ruhuna 58 sri lanka short paper soil amendment with raw rice hull as a source of silicon in enhancing anthracnose disease resistance, plant growth and fruit qualities of chili pepper (capsicum annuum l.) h.a.r.k. jayawardana and h. l. d. weerahewa1 department of botany, the open university of sri lanka, nawala, nugegoda, sri lanka correspondence: 1 weerahewa@gmail.com received: 11th june 2016, revised: 27th august 2016, accepted: 06th october 2016 abstract. this study was conducted to investigate the effect of soil incorporation of raw rice hull (rrh) as a source of silicon on plant growth, yield and anthracnose disease development of capsicum annuum l. ‘muria f1’. experiment was conducted using five different ratios of rrh:soil:sand (t1-3:2:0, t2-3:0:2, t3-3:1:1, t4-2:3:0, t51:1:3) in the first 5 treatments and applying potassium silicate (ps) as a si source at the rate of 75 mg/l to the soil replacing rrh in the 6th (positive control). the plants grown in soil medium with no addition of a silicon source were used as the negative control. all treatments were arranged in a completely randomized design with 4 replicates. anthracnose disease development was assessed as number of days for the appearance of symptoms and the mean lesion areas by challenge inoculation with colletotrichum gloeosporioides on fruits. the highest reduction of disease severity (61%) was observed in crops grown in soils amended with rrh (t1) compared to control plants. the disease reduction observed in fruits from plants provided with potassium silicate was only about 15%, compared to that of si-free plants. it was concluded that incorporation of rice hull (60% of the media) in the growing substrate would be an effective solution against anthracnose disease of chili pepper plants. keywords. capsicum annuum, silicon, rice hull, soil amendment 1 introduction anthracnose disease caused by colletotrichum species is a major disease of chili pepper (capsicum annuum l.) in tropical and sub-tropical climates and causes severe postharvest losses (oanh et al. 2004). the disease is controlled by seed treatment or foliar sprays with fungicides. however, continuous use of fungicides can leads to environmental and health hazards and development of fungicide resistance of pathogen populations. therefore, to address the jayawardana and weerahewa chili grown in raw rice hull amendment soil ruhuna journal of science vol 7: 58-63, december 2016 59 current demand in sustainable crop production, environmental friendly alternatives should be investigated to control one of the important diseases, anthracnose. silicon is beneficial for plants in terms of growth (kamenidou et al. 2008), yield (ghasemi et al. 2013) and resistance to biotic (huang et al. 2011) and abiotic stresses (savant et al. 1999). silicon can be supplied using both inorganic and organic sources. rice hull is an agricultural by-product, which is underutilized. it is a natural source of silicon. according to patel et al. (1987) the si content in raw rice husk is 10.3 (w/w% in wet basis). this study was conducted to investigate the effect of incorporation of rice hull (rrh) and potassium silicate (ps) as silicon sources into potting media on plant growth, fruit quality parameters and anthracnose disease development of chili pepper. 2 materials and methods 2.1 plant material seeds of capsicum annuum l. ‘muria f1’ (east-west seed international ltd., thailand) were sown on coir dust and compost medium (1:1) and were maintained in the nursery. four weeks old chili pepper plants were transplanted in grow bags (300 gauge polythene bags, size 30 cm × 12 cm) filled with equal volumes of the respective growing media according to treatments. 2.2 treatments and experimental design rrh: soil: sand in a ratio of (3:2:0, 3:0:2, 3:1:1, 2:3:0, 1:1:3) were used in first five treatments; t1-t5 while ps (75 mg/l of si) was applied to the soil medium instead of adding rrh in t6 (control 1). the plants grown in soil medium applied with no silicon source were used as the t7 (control 2). treatments were arranged in a completely randomized design (crd) with four replicates having three plants per experimental unit. the data were analyzed using one way anova in spss 16 statistical package. 2.3 fertilizer application nutrients were supplied using nfv (nutrient formulation for vegetative stage) and nff (nutrient formulation for fruiting stage) (jayawardana et al. 2015) for all the other treatments except t6, in which the nutrient formulations were amended with potassium silicate to supply 75 mg/l of si by making proper adjustment in the nutrient composition (jayawardana et al. 2015). fertigation (50 ml of nutrient solution per grow bag per day during first 4 weeks and 100 ml per growing bag per day thereafter) was practiced with relevant nutrient jayawardana and weerahewa chili grown in raw rice hull amendment soil ruhuna journal of science 60 vol 7: 58-63, december 2016 solution according to the treatments. weeding and watering was practiced when required. 2.4 pathogen isolation and identification c. gloeosporioides from anthracnose lesions of diseased chili pepper fruits were cultured on potato dextrose agar (pda), following surface sterilization with 1% (v/v) naocl for 1 min, followed by washing with sterile distilled water. ten culture plates were incubated at 27°–30°c and observed for mycelial growth, morphology of the cultures, and the shape of conidia were observed using a compound microscope (daffodil mcx100, vienna, australia). c. capsici was identified by its sickle-shaped conidia, the presence of prominent setae (sutton, 1992), and its brown colony colour (rajapakse and ranasighe 2002). c. gloeosporioides was identified by its orange cottonlike mycelium (sutton, 1980) and ovoid-shape conidia (du et al. 2005). 2.5 fruit inoculation and assessment of disease severity conidial suspension (105 conidia ml-1) of c. gloeosporioides or c. capsici were prepared by scraping the mycelium from pure 7 days old cultures and suspending them in sterilized distilled water, followed by filtering through glass wool. harvested fruits were challenge-inoculated by placing a 20 µl drop of conidial suspension at three different points on the fruit surface. twelve fruits were inoculated per treatment. the inoculated fruits were maintained in a moist chamber (95 – 100% rh) at 28 ± 2°c. numbers of days for the appearance of disease symptoms in each treatment was recorded. lesion areas were recorded each day for 10 days and the mean lesion area per fruit was calculated. 2.6 plant growth and fruit parameters the shoot length, number of leaves, average leaf area, internodes girth (3-4 nodes) and number of fruits per plant were recorded at 12 weeks after transplanting. fruit length and fruit fresh weight of each harvested fruit were measured. 3 results and discussion 3.1 anthracnose disease resistance in chili pepper fruits the highest lesion area after 10 days of pathogen c. gloeosporioides inoculation was recorded in the control treatment (77.5 mm2) and the lowest jayawardana and weerahewa chili grown in raw rice hull amendment soil ruhuna journal of science vol 7: 58-63, december 2016 61 was recorded in t1 treatment (29.5 mm 2) (fig. 1.). the significantly high reduction of lesion area compared to the control 2 (t7) (61%) was shown in crops treated with rice hull in the growing substrate as the si source (t1). however, the lesion area observed in fruits from t1, t2 and t3 treatments at 10 days after inoculation was not significant. the reduction of lesion area was significantly lower both in t4 and t6 treatments than that of t1, t2 and t3. there was no significant difference observed between the lesion areas of t5 and control treatments. a comparatively large percentage of rice hull was included in t1, t2 and t3 treatments compared to the other treatments, suggesting that si supplied by rice hull in the media might have actively contributed for anthracnose disease suppression. fig. 1. mean lesion area (mm2) on chili pepper fruits at 10 days after inoculation of c. gloeosporioides. bars with same letters are not significantly different (p ≤ 0.05) as determined by tukey hsd test. (t1= rrh:soil:sand at 3:2:0, t2= rrh:soil:sand at 3:0:2, t3= rrh:soil:sand at 3:1:1, t4= rrh:soil:sand at 2:3:0, t5= rrh:soil:sand at 1:1:3, t6-soil medium treated with ps and t7si-free medium) only a 15% reduction of anthracnose lesion area was observed in fruits from t6 compared to that of control fruits. however, our previous study indicated that supplementation of 75 mg/l of potassium silicate in hydroponic nutrient solution can significantly reduce the anthracnose disease in chili pepper (jayawardana et al. 2015). there could have been a significant reduction in the anthracnose disease by application of greater concentrations than 75 mg/l of potassium silicate into the soil medium. it has to be further investigated. jayawardana and weerahewa chili grown in raw rice hull amendment soil ruhuna journal of science 62 vol 7: 58-63, december 2016 table 1. plant growth and fruit parameters of chili pepper plants grown in combinations of raw rice hull, soil, sand and ps incorporated substrates (means followed by the same letter within the same row are not significantly different at p ≤ 0.05 as determined by tukey hsd test; ns= not significant) parameter t1 t2 t3 t4 t5 t6 t7 shoot length (cm) 63a 61 a 65 a 59 a 69b 60a 59 a inter node (3-4) girth (cm) 2.9 2.8 3.1 2.9 2.9 2.8 2.8 ns number of leaves 53 a 50 a 52 a 49 a 57b 50a 48 a number of fruits/plant 9 a 9 a 10a 9 a 12b 8a 8a fruit length (cm) 12.8 13.0 12.7 12.2 13.3 12.9 12.0 ns fruit fresh weight (g) 25.0 27.2 24.1 26.3 26.8 27.0 25.5 ns average leaf area (cm2) 45.2 40.0 43.6 42.9 45.0 44.1 42.2 ns t1= rrh:soil:sand at 3:2:0, t2= rrh:soil:sand at 3:0:2, t3= rrh:soil:sand at 3:1:1, t4= rrh:soil:sand at 2:3:0, t5= rrh:soil:sand at 1:1:3, t6-soil medium treated with ps and t7si-free medium) 3.2 plant growth and fruit parameters there was no significant effect of potassium silicate applied into the soil medium (t6) on growth parameters of chili pepper plants compared to the control. however, the shoot length, number of leaves and number of fruits were significantly greater in the plants from t5 treatment (plants were grown in rh:soil:sand at 1:1:3 medium) compared with the rest of the treatments (table 1). sand in the medium enhances the porosity. it can be suggested that not only the si provided by the medium but also the properties of the medium could have affected the results particularly, on the plant growth and fruit parameters. the ratio of 1:1:3 of rrh:soil:sand might be a good combination in the substrate to enhance plant growth and yield parameters of chili pepper. however, the disease resistance showed in the t5 treatment was not significant, which could have been attributed to lower rrh applied. incorporation of rrh in the form of ash into growing media has improved growth of sunflower plants (kamenidou 2005; kamenidou et al. 2008). in the current study rrh was used in the growth medium. in a previous study, rrh incorporation in a simplified hydroponic system, showed a significant increase in anthracnose disease resistance, plant growth and fruit qualities of chili pepper plants (jayawardana et al. 2016). jayawardana and weerahewa chili grown in raw rice hull amendment soil ruhuna journal of science vol 7: 58-63, december 2016 63 4 conclusions incorporation of rice hull (60% of the media) in the ratios of 3:2:0, 3:0:2 and 3:1:1 of rrh:soil:sand in the growing substrate would be effective against anthracnose disease of chili pepper plants. si provided by rrh in the substrate would be the main factor influencing the disease reduction. however, more sand in combination with rrh in the growing substrate (rrh:soil:sand in 1:1:3 ratio) would enhance shoot length, number of leaves and number of fruits of chili pepper plants. therefore, combination of rrh with soil and sand in the ratio of 3:1:1would be better in overall performance of chili pepper crop. acknowledgments. national science foundation (rg/2012/ag/04) and the faculty research grant of the faculty of natural science, ousl are kindly acknowledged for the financial assistance. references du m, schardl cl, nuckles em, vaillancourt lj. 2005. using mating-type gene sequences for improved phylogenetic resolution of colletotrichum species complexes. mycologia 97: 641-658. ghasemi a, ejraei a, rajaei m. 2013. effect of silicon on vegetative and generative performance of broad bean (vicia faba l.). journal of novel applied science. 2 (s): 881– 884. huang c, roberts pd, datnoff le .2011. silicon suppresses fusarium crown and root rot of tomato. journal of phytopathology 159: 546–554. jayawardana hark, weerahewa hld, saparamadu mdjs. 2015. enhanced resistance to anthracnose disease in chili pepper (capsicum annuum l.) by amendment of the nutrient solution with silicon. journal of horticultural science and biotechnology 90(5): 557–562. jayawardana hark, weerahewa hld, saparamadu mdjs. 2016. the effect of rice hull as a silicon source on anthracnose disease resistance and some growth and fruit parameters of capsicum grown in simplified hydroponics. international journal of recycling of organic waste in agriculture 5: 9-15. kamenidou s. 2005. silicon supplementation affects greenhouse produced cut flowers, m,sc. thesis, faculty of the graduate college, oklahoma state university. kamenidou s, cavins tj, marek s. 2008. silicon supplements affect horticultural traits of greenhouse-produced ornamental sunflowers. hortscience 43 (1): 236-239. oanh ltk, korpraditskul v, rattanakreetakul c. 2004. a pathogenicity of anthracnose fungus, colletotrichum capsici on various thai chili varieties. kasetsart journal (natural science) 3: 103-108. patel m, karera p, prasanna p. 1987. effect of thermal and chemical treatments on carbon and silica contents in rice husk. journal of materials science 22: 2457-2464. rajapakse rgas, ranasinghe, jadar. 2002. development of variety screening method for anthracnose disease in chili. tropical agricultural reseach and extention, 5: 7-11. savant nk, snyder gh, datnoff le. 1999. silicon management and sustainable rice production. advanced agronomy 58: 151-199. sutton bc. 1980. the coleomycetes. fungi imperfect with pycnidia, acervula and stromata. commonwealth mycological institute. kew, uk. 522-537. sutton bc. 1992. the genus glomerella and its anamorph colletotrichum. in : colletotrichum : biology, pathology and control. (bailey ja, jeger mj (eds). cab international: wallingford, oxon, uk. 1-26. ruhuna journal of science vol 10 (2): 161-173, december 2019 eissn: 2536-8400 ©faculty of science doi: http://doi.org/10.4038/rjs.v10i2.81 university of ruhuna © faculty of science, university of ruhuna sri lanka s1 wijeweera et al. 2019 – supplementary material appendix the required materials are available in the link given below. https://drive.google.com/drive/folders/1ypwhjouf4xn9bfxvfzc_hijxodsas wsr the folder ‘application’ contains an application in c# programming language that can be used to demonstrate the proposed algorithm. the graphical user interface of the application is given in figure 1. fig 1. graphical user interface of the c# application • the number of vertices of the convex polygon should be entered in the text box with the label ‘vertices’. • “0” appears in the text box with the label ‘v’ after pressing the button ‘ok’. that means the user has to enter each coordinate of the 0th vertex of the polygon in the text boxes with labels ‘x’ and ‘y’ respectively. • “1” appears in the text box with the label ‘v’ after pressing the button ‘insert’. that means the user has to enter each coordinate of the 1st vertex of the polygon in the text boxes with labels ‘x’ and ‘y’ respectively. • similarly, coordinates of each vertex of the convex polygon can be inserted. the convex polygon can be displayed on the panel by clicking the button ‘draw’. • random line segments can be generated on the panel. the number of random line segments should be entered in the text box with the label ‘number’. • the algorithm extracts the parts of the generated line segments that are inside the convex polygon. the button ‘run’ should be clicked to see the output on the panel. • the details of the execution can be written into a text file by pressing the button ‘save’. wijeweera et al. line segment clipping against a convex polygon ruhuna journal of science vol 10 (2): 161-173, december 2019 s2 • the content on the panel can be cleared by pressing the button ‘clear’. each convex polygon in table 2 (in wijeweera et al. 2019) was tested by generating 100 random line segments. the corresponding snapshots of results are given in figure 2. note that the interior parts of the line segments are shown in green color and the exterior parts of the line segments are shown in red color. the convex polygon is shown in blue color. fig 2. snapshots for each convex polygon the folder ‘complete project’ contains the complete c# project of the application. follow the file path: complete project →clipping1→clipping1→form1.cs. the c# programming code of the proposed algorithm is available in the file form1.cs. ruhuna journal of science vol 12 (1): 26-39, june 2021 eissn: 2536-8400 © faculty of science http://doi.org/10.4038/rjs.v12i1.98 university of ruhuna sri lanka © faculty of science, university of ruhuna sri lanka 26 forest resources as ecotourism attraction: cross river national park, nigeria bukola o. adetola*1, abideen a. alarape2 and ibukun a. ayodele2 1department of ecotourism and wildlife management, federal university of technology, akure, ondo state, nigeria 2department of wildlife and ecotourism management, university of ibadan, oyo state, nigeria *correspondence: boadetola@futa.edu.ng; orcid: https://orcid.org/0000-0002-1675-3344 received: 5th june 2020, revised: 29th april 2021, accepted: 10th june 2021 abstract cross river national park (crnp) in nigeria with its natural resources of tourist significance can help to foster a culture of conservation and recreation among the public. this study identified ecotourism attractions in crnp aiming to promote its sustainable use and to enhance its planning as a haven for ecotourism in nigeria. a field survey of ecotourism resources was conducted in the park during 2010-2014. fauna resources were identified along the 5 km transects that covered existing trails and jeep tracks in both dry and wet seasons. quadrat method (15 sample plots of 50 m x 20 m) was used for the vegetation assessment and shannon-weiner diversity index (h) and descriptive statistics were calculated. findings revealed that the rainforest, game viewing, birdwatching, mountain climbing, camping, rock formation, village tour, botanical garden tour and water recreation as core ecotourism attractions in crnp. fifteen wild animal species belonging to nine families were observed (endangered 01, least concerned 12, near threatened 01, vulnerable 01). wild animals included drill monkey mandrillus leucophaecus (endangered), elephant loxodonta africana cyclotis (vulnerable), bat eidolon helvum (near threatened), and bare-headed rock fowl picarthates oreas (endemic) which are “ecostars” and “a must see” that warrant a visit to the park. other animals were least concerned. a total of 81 tree species belonging to 26 families were identified. shannon-wiener diversity index (h’) of 3.88 and 3.84 were recorded for oban and okwango divisions respectively. the most dominant family in oban was leguminosae (12 species) and in okwango, it was apocynaceae (10 species). managing and promoting the sustainable use of the park’s numerous resources to actualize the potential value non-consumptively through ecotourism is hereby recommended. keywords: attraction, conservation ecotourism, national park, rainforest 1 introduction tourism is both leisure and a self-indulgent pursuit. it is a global phenomenon that has experienced rapid growth in the developing countries of the world. smith (2004) https://rjs.ruh.ac.lk/index.php/rjs/index https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v12i1.98 https://creativecommons.org/licenses/by-nc/4.0/ mailto:boadetola@futa.edu.ng https://orcid.org/0000-0002-1675-3344 b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 27 noted that tourism and travel have been part of the human experience for millennia, describing it as a form of movement that characterizes homo sapiens. ecotourism is a type of specialty travel incorporating a diverse array of activities and tourism type, from bird watching, game viewing, scientific study, photography, diving, trekking, to regeneration of damaged ecosystems. the international ecotourism society defined ecotourism as “responsible travel to natural areas that conserves the environment, sustains the well-being of the local people, and involves interpretation and education” (ties 2015). it has been able to capitalize on the increased motivation to experience and preserve natural environments (diamantis 2004). tourism industry has wholly embraced ecotourism, and its global importance has been recognized by the un through the declaration of the year 2002 as the “international year of ecotourism” the concern of tourists for environmental issues has increased, and ecotourism has become a known sector of tourism that should be developed (holden 2003), because of its strategy of understanding nature and imparting conservation measure on biodiversity. biodiversity is the backbone of the ecotourism industry. the role of biodiversity in tourism ranges from biodiversity as an attraction (i.e., many wildlife are focal species), resources for consumptive goods (i.e., culinary), natural components to support ecological services (i.e., pollination), to aesthetics (i.e., ornamental plants) (higginbottom 2004, newsome et al. 2012). biodiversity, however, is not a tourist attraction unless its tourism potential value is converted and actualized as objects which are able to attract tourists (drumm and moore 2002). the success of tourism attraction, therefore, depends on the ability of tourism planners and managers to actualize the potential value of biodiversity as a tourist attraction (luchman 2017). ecotourism is seen as an interfacing of conservation concerns and tourism interest. essentially, ecotourism derives its attraction from a combination of groups to be made from marketing a product that exists in its natural state in a specific geographic location and the potential to make such ecologically, economically, and socially sustainable (ijeoma 2007). the national park divisions at oban and okwango in cross river state, nigeria is a wildlife park suitable for ecotourism. the cross river national park (crnp) is an important ecological gene pool containing one of the oldest rainforests in africa. because of its critical conservation status, it has been designated as one of the 25 un biodiversity hot spots in the world (crnp 2010). the forest resources represent a tourism haven capable of instilling conservation consciousness in the mind of the tourists that patronize the park. therefore, adequate knowledge of the occurrence of the tourism products in an ecosystem is vital to enhance its effective marketing, increased patronage and sustainable use through ecotourism. despite the enviable potential nigeria is endowed with, the country has remained an unpopular tourist destination. therefore, this study was conducted to identify the park’s core tourist attractions, to determine the potential recreation activities of ecotourism, to assess the fauna and flora resources of crnp forest as a draw for ecotourism and to promote argument for its successful conservation and sustainable use. this will enhance b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 28 effective decision-making in the competing economic, social and environmental demands of sustainable development in nigeria. 2 material and methods 2.1 study area cross river national park (crnp) is located in cross river state between longitudes 5o.05'-6o.29' n and latitudes 8o.15'90.30' e in nigeria (figure 1). it covers an area of approximately 4000 km2, divided into two non-contiguous divisions – the oban hills in the southern part covering 3000km2, and the okwangwo division in the northern part covering 1000 km2 of primary moist rainforest ecosystem in the northern and central parts, and montane mosaic on the obudu plateau (crnp 2010). fig. 1: oban and okwangwo divisions of cross river national park with support zone communities b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 29 the terrain is rough, and elevation rises from the river valleys to over 1000 m in mountainous areas. the park has a tropical climate characterized by a rainy season between april and november. the annual rainfall ranges between 2000 mm to 3000 mm; relative humidity in and around the park range from 80-90%. the temperature rarely falls below 19°c with an annual mean of 27°c (udoidung et al. 2007). many ecosystems including the tropical rainforest and mangrove forest occur in cross river state. the geographical location of the state favours ecotourism as the state is accessible by air through the calabar airport, by sea through the calabar sea port and by road through the trunk roads from akwa ibom, ebonyi and abia states (udoidung et al. 2007). the ecology of the state has made it a renowned wildlife sanctuary and the government has made substantial investment to develop its tourism potentials across the state in order to realize its aspiration of making the state the flagship of tourism in the country. the rich cultural heritage and strategic coastal location are great impetus for the rapidly developing culture and business tourism in the state. 2.2 data collection and analysis ecotourism resources available in the park were carefully observed and assessed. the global positioning system (gps) surveying method was used in the mapping of ecotourism resources in the study area (shrestha 2006). relevant pictures that depict specific and exciting characteristic features of interest were taken using digital camera-samsung zoom lens 5x (14.2 megapixels) and gps (gps550 magellan) for the geographic coordinates reading. fauna resources were identified along the existing 5 km transects, trails and jeep tracks within the park. observations were recorded between 6.30 am to 11.00 am in the morning and 4.00 to 6.30 pm in the evening three days a week to identify the fauna species. twenty-four days of observation were carried out in february (dry season) and august (rainy season), 2012 , making up 12 days in each season. spoors such as faecal droppings, footprints spines, calls, and trails were recorded and used to determine the presence of the animals in the park in addition to the direct observations. there are no recent studies so these data will provide some baseline information. quadrat method was used for the vegetation assessment of oban and okwango divisions of the park. the method adopted by fao (2009) was modified and a total of 15 plots with 50m x 20m quadrant size each were established on three lines transversed evenly along the existing jeep track within the study area. five plots were marked on each line at 10 km interval. identification of tree species within each plot was carried out. all identified trees were allocated to families and the number of species was obtained for the diversity classification of tree species in each family using keay (1989). the total density was calculated as well as the diversity and species evenness. species relative density (rd) and relative frequency (rf) of each site were computed using equations 1 and 2. species diversity index was calculated b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 30 using shannon–wiener diversity index (eq. 3), while shannon’s equitability index (eh) (eq. 4) was adopted for estimating species evenness. species relative density (rd): this refers to the number of individuals of a given species divided by the total number of individuals of all species. rd = [ ni n ] x 100 (1) where, rd = relative density, ni = number of individual species i, and n = total number of individual in the entire population. relative frequency (rf): rf was obtained using the formula given by oduwaiye et al. (2002). rf = ∑ fi x 100 fn (2) where, rf = relative frequency, fi = number of plot where species was found, and fn = total frequency of all species. shannon-wiener diversity index (hi): the equation given by price (1997) was adopted. h1 = ∑ pi𝑙𝑛pisi=1 (3) where, s= total number of species in the community, pi= proportion of a species to the total number of plant in the community, ln = natural logarithm. species evenness index (eh): in each forest community, species evenness was determined using shannon’s equitability (eh). eh = h hmax = ∑ piln(pi)si=1 ln(s) (4) b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 31 3 results 3.1 ecotourism resources in cross river national park prominent recreational activities that can be conducted in the cross river national park (crnp) are hiking in the rainforest vegetation, mountaineering, water recreation in the natural water body (e.g. kwa river), game viewing, bird watching, botanical garden tour and visit the natural history museum as presented in table 1. the topography presents very interesting sceneries for ecotourism activities and harbors rare species of animals such as the critically endangered cross river gorilla (gorilla gorilla diehli), and the endangered chimpanzee (pan troglodytes). village tours can provide visitors the opportunity to explore the way of life of the local communities. table 1: ecotourism attractions in cross river national park, nigeria. attractions location wilderness hiking the entire park game viewing/ bird watching the entire park mountain viewing/ climbing park rugged terrain rising from 100 m in the river valleys to over 1000m in the mountains mbe mountain forest camping the entire park rock formation bat caves in oban division village tours 105 support zone communities of crnp tour of botanical garden okwangwo division water recreation natural water pool – kwariver, bemi river, oyi river and okon river natural history museum oban division crnp is endowed with biodiversity resources, which serve as a major source to attract ecotourists to this area. fifteen wild animal species belonging to 9 families were observed (table 2) and a total of 81 tree species belonging to 26 families were identified in oban and okwango divisions of crnp (tables 3 and 4). the diversity index (h') and equitability (evenness) (eh) for oban division is 3.88 and 0.62 and for okwango division h is 3.84 and (eh) 0.53, respectively (table 5). according to the redlist conservation status in 2015, the animals encountered included endangered drill monkey (mandrillus leucophaecus) vulnerable elephant (loxodonta africana cyclotis) near threatened bat (eidolon helvum), and others were at the least concerned status. the endemic bare-headed rock fowl picarthates oreas b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 32 and the cameroonian vine, ancistrocladus korupensis are “ecostars” and “a must see” that warrant a visit to the park. table 2: fauna species encountered in cross river national park, nigeria. family common name scientific name evidence remark iucn status * population trend bovidae red flanked duiker cephalophus rufilatus faecal dropping fresh lc decreasing bovidae blue duiker cephalophus monticola footprint/ trail fresh lc stable bovidae bush buck tragelaphus scriptus footprint fresh lc stable hystricidae porcupine altherurus africanus footprint/ trail fresh lc unknown suidae red river hog potamochoerus porcus footprint fresh lc decreasing sciuridae giant tree squirrel protoxerus stangeri sighting on a tree lc unknown sciuridae ground squirrel xerus erythropus sighting crossing transect line lc stable elephantidae forest elephant loxodonta africana cyclotis footprint/ feeding trail/ dung dung about 2 weeks old v increasing viverridae palm civet viverra civetta sighting individual on tree lc decreasing viverridae mongoose ichneumia albicauda footprint more than 10 sighting lc stable cercopithecidae drill monkey mandrillus leucophaeus sighting/ vocalisation about 15 individuals on a tree feeding on fruit en unknown cercopithecidae mona monkey cercopithecus mona sighting/ vocalisation 5 individuals lc unknown cercopithecidae tantalus monkey cercopithecus aethiops sighting 1 individual lc stable capitonidae yellowspotted barbet buccanodon duchaillui sighting 1 individual lc stable pteropodidae fruit bats eidolon helvum sighting more than 20 sighted at erokut rock nt decreasing *lc-least concern, v-vulnerable, ntnear threatened, enendangered b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 33 table 3: tree species observed in oban division of cross river national park. species family iucn status # density (ha-1) relative density pi*ln(pi) allanblackia floribunda guttiferae vu 5 0.581395 -0.02993 alstonia congensis apocynaceae lc 44 1.98915 -0.08204 antrocaryon micraster anacardiaceae vu 8 0.180832 -0.0121 anonidium mannii annonaceae lc 12 0.542495 -0.02993 antiaris toxicaria moraceae lc 5 0.361664 -0.02152 anthocleista vogelii apocynaceae lc 5 0.180832 -0.0121 baphia nitida fabaceae lc 24 0.542495 -0.02993 bridelia micrantha euphorbiaceae lc 40 2.712477 -0.10285 brenania brelie leguminosae lc 2 0.180832 -0.0121 carapa procera meliaceae lc 22 2.531646 -0.09787 chrysophyllum albidum sapotaceae nt 5 0.361664 -0.02152 christiana africana tiliaceae lc 16 0.180832 -0.0121 cleistopholis patens annonaceae lc 8 0.361664 -0.02152 cola acuminate sterculiaceae lc 64 0.723327 -0.03767 corynanthe pachyceras rubiaceae lc 64 0.723327 -0.03767 coula edulis olacaceae lc 154 5.244123 -0.16179 cylicodiscus gabunensis leguminosae lc 32 0.361664 -0.02152 distemonanthus benthamianus olacaceae lc 34 3.435805 -0.12157 dialium guineens leguminosae lc 192 2.169982 -0.08747 desplatsia dewevrei tiliaceae lc 96 2.169982 -0.08747 pycnanthus angolensis myristicaceae lc 16 0.180832 -0.0121 dacryodes edulis burseraceae nl 32 0.361664 -0.02152 diospyros mespiliformis ebenaceae lc 13 0.723327 -0.03767 desplatsia dewevrei tiliaceae lc 6 0.361664 -0.02152 diospyros suaveolens ebenaceae nl 5 0.361664 -0.02152 diospyros zenkeri ebenaceae lc 84 3.797468 -0.1303 enantia chlorantha annonaceae lc 21 0.723327 -0.03767 entandrophragma cylindricum meliaceae vu 11 0.361664 -0.02152 ficus capensis moraceae nl 32 0.723327 -0.03767 funtumia elastica apocynaceae lc 48 0.542495 -0.02993 garcinia mannii guttiferae vu 88 1.98915 -0.08204 garcinia staudtii guttiferae vu 32 0.361664 -0.02152 guibourtia ehie leguminosae lc 16 0.180832 -0.0121 guarea glomerulata meliaceae nl 21 0.723327 -0.03767 guarea thompsonii meliaceae vu 16 0.361664 -0.02152 hannoa klaineana simaroubaceae nl 112 1.265823 -0.05834 hunteria umbellata apocynaceae lc 112 1.265823 -0.05834 distemonanthus benthamianus leguminosae lc 48 0.542495 -0.02993 irvingia gabonensis irvingaceae nt 32 0.361664 -0.02152 khaya ivorensis meliaceae vu 2 0.180832 -0.0121 lophira alata ochnaceae vu 8 0.361664 -0.02152 maesobotrya barteri moraceae lc 88 1.98915 -0.08204 microdesmis puberula euphorbiaceae nl 12 0.542495 -0.02993 milicia excelsa moraceae nt 2 0.180832 -0.0121 millettia griffoniana leguminosae lc 64 2.169982 -0.08747 musanga cecropioides moraceae lc 52 2.350814 -0.09274 neoboutonia glabrescens euphorbiaceae nt 5 0.361664 -0.02152 nesogordonia papaverifera sterculiaceae vu 8 0.180832 -0.0121 newbouldia laevis bignoniaceae nl 8 0.180832 -0.0121 b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 34 table 3. continued species family iucn status # density (ha-1) relative density pi*ln(pi) octoknema affinis octoknemaceae nl 11 0.723327 -0.03767 omphalocarpum procerum sapotaceae lc 2 0.180832 -0.0121 ongokea gore olacaceae lc 3.2 0.361664 -0.02152 ouratea calophylla ochnaceae nl 5 0.361664 -0.02152 acacia albida leguminosae lc 60 0.180832 -0.10285 parkia bicolor leguminosae lc 112 2.712477 -0.05834 pentaclethra macrophylla leguminosae lc 48 1.265823 -0.02993 piptadeniastrum africanum leguminosae lc 5 0.542495 -0.0121 poga oleosa anisophylleaceae lc 16 0.180832 -0.0121 prosopis africana leguminosae lc 18 0.180832 -0.07642 pycnanthus angolensis myristicaceae nl 32 1.808318 -0.02152 rauvolfia mannii apocynaceae lc 32 0.723327 -0.03767 rauvolfia vomitoria apocynaceae lc 128 1.446655 -0.0646 rothmannia hispida rubiaceae lc 96 1.084991 -0.05179 amphimas pterocarpoids leguminosae lc 3 0.180832 -0.0121 spathodea campanulate bignoniaceae lc 3 0.180832 -0.0121 staudtia stipitate myristicaceae lc 24 1.627486 -0.07062 dacryodes edulis burseraceae nl 4 0.180832 -0.0121 sterculia oblonga sterculiaceae lc 16 0.542495 -0.02993 sterculia rhinopetala sterculiaceae lc 11 0.361664 -0.02152 sterculia tragacantha sterculiaceae lc 32 0.723327 -0.03767 strombosia grandifolia olacaceae nl 448 5.063291 -0.15812 strombosia pustulata olacaceae lc 368 8.318264 -0.21551 tabernaemontana pachysiphon apocynaceae lc 80 0.904159 -0.04493 trichilia lanata meliaceae lc 112 1.265823 -0.05834 trilepisium madagascariense moraceae cr 48 1.627486 -0.07062 uapaca staudtii euphorbiaceae lc 48 1.084991 -0.05179 uapaca togoensis euphorbiaceae lc 64 0.723327 -0.03767 vitex doniana verbenaceae lc 32 0.361664 -0.02152 xylopia aethiopica annonaceae lc 144 1.627486 -0.07062 zanthoxylum zanthoxyloides rutaceae en 64 0.723327 -0.03767 3.88336 *ls (least concerned), v (vulnerable), nt (near threatened), nl (not listed), e (endangered), cr (critically endangered) a total of 80 tree species belonging to 25 families was identified in oban division while 78 species belonging to 22 families were recorded in okwango division. leguminosae (12) and apocynaceae (10) were the most dominant families in oban and okwango divisions respectively. other dominant families in both divisions were moraceae, olacaceae, and guttiferae. the slight variation in the diversity index signified that oban division is more diverse in woody tree species than okwango division. notable amongst the flora species are the critically endangered (trilepisium madagascariense), endangered (zanthoxylum zanthoxyloides), near threatened (irvingia gabonensis, neoboutonia glabrescens, milicia excels, chrysophyllum albidum) and vulnerable (nesogordonia papaverifera, lophira alata, khaya b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 35 ivorensis, guarea thompsonii, garcinia mannii, garcinia staudtii, entandrophragma cylindricum, antrocaryon micraster, allanblackia floribunda). table 4: tree species observed in okwango division of cross river national park. species family iucn status # density (ha-1) relative density pi*ln(pi) acacia albida apocynaceae lc 2 0.246914 -0.01218 allanblackia floribunda guttiferae vu 7 0.740741 -0.03012 alstonia congensis apocynaceae lc 59 2.716049 -0.08251 amphimas pterocarpoides apocynaceae lc 6 0.493827 -0.02166 anonidium mannii annonaceae lc 8 0.740741 -0.03012 anthocleista vogelii apocynaceae lc 5 0.246914 -0.01218 antiaris toxicaria moraceae lc 6 0.493827 -0.02166 antrocaryon micraster apocynaceae vu 176 8.148148 -0.17672 baphia nitida fabaceae lc 24 0.740741 -0.03012 brenania brelie fabaceae nl 16 0.246914 -0.01218 bridelia micrantha fabaceae lc 80 3.703704 -0.10343 carapa procera meliaceae lc 45 3.45679 -0.09842 christiana africana olacaceae lc 3 0.246914 -0.01218 chrysophyllum albidum sapotaceae nt 3 66.66667 -0.02166 cleistopholis patens annonaceae lc 11 0.493827 -0.02166 sterculia oblonga sterculiaceae lc 4 0.987654 -0.03791 staudtia stipitate olacaceae lc 4 0.987654 -0.03791 coula edulis olacaceae lc 240 7.407407 -0.16624 cylicodiscus gabunensis olacaceae lc 16 0.493827 -0.02166 dacryodes edulis burseraceae nl 16 0.493827 -0.02166 desplatsia dewevrei tiliaceae lc 5 0.493827 -0.02166 dialium guineense leguminosae lc 96 2.962963 -0.08797 diospyros mespiliformis ebenaceae lc 21 0.987654 -0.03791 diospyros suaveolens ebenaceae nl 16 0.493827 -0.02166 diospyros zenkeri ebenaceae lc 168 5.185185 -0.131 distemonanthus benthamianus leguminosae lc 304 4.691358 -0.12223 drypetes chevalieri euphorbiaceae nl 64 2.962963 -0.08797 enantia chlorantha annonaceae lc 32 0.987654 -0.03791 entandrophragma cylindricum meliaceae vu 32 0.493827 -0.02166 ficus capensis moraceae nl 64 0.987654 -0.03791 funtumia elastica apocynaceae lc 10 0.740741 -0.03012 garcinia staudtii guttiferae vu 16 0.493827 -0.02166 garcinia mannii guttiferae vu 88 2.716049 -0.08251 guarea glomerulata guttiferae nl 96 1.481481 -0.05211 guarea thompsonii meliaceae vu 32 0.493827 -0.02166 guibourtia ehie guttiferae lc 8 0.246914 -0.01218 hannoa klaineana simaroubaceae nl 112 1.728395 -0.05869 hunteria umbellate guttiferae lc 112 1.728395 -0.05869 irvingia gabonensis irvingiaceae nt 32 0.493827 -0.02166 khaya ivorensis meliaceae vu 16 0.246914 -0.01218 lophira alata ochnaceae vu 16 0.493827 -0.02166 maesobotrya barteri euphorbiaceae lc 176 2.716049 -0.08251 microdesmis puberula euphorbiaceae nl 16 0.740741 -0.03012 milicia excelsa moraceae nt 8 0.246914 -0.01218 b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 36 table 4. continued species family iucn status # density (ha-1) relative density pi*ln(pi) millettia griffoniana leguminosae lc 96 2.962963 -0.08797 musanga cecropioides moraceae lc 208 3.209877 -0.09327 neoboutonia glabrescens moraceae nt 11 0.493827 -0.02166 nesogordonia papaverifera moraceae vu 8 0.246914 -0.01218 newbouldia laevis moraceae nl 16 0.246914 -0.01218 octoknema affinis olacaceae nl 64 0.987654 -0.03791 omphalocarpum procerum moraceae lc 3 0.246914 -0.01218 ongokea gore moraceae lc 16 0.493827 -0.02166 ouratea calophylla ochnaceae nl 16 0.493827 -0.02166 parinari macrophylla rosaceae nl 112 1.728395 -0.05869 parkia bicolor leguminosae lc 240 3.703704 -0.10343 pentaclethra macrophylla leguminosae lc 8 0.246914 -0.01218 piptadeniastrum africanum leguminosae lc 64 0.987654 -0.03791 amphima spterocarpoides apocynaceae nl 16 0.246914 -0.01218 prosopis africana leguminosae lc 16 0.246914 -0.01218 pycnanthus angolensis myristicaceae nl 176 2.716049 -0.08251 rauvolfia mannii apocynaceae lc 32 0.987654 -0.03791 rauvolfia vomitoria apocynaceae lc 128 1.975309 -0.06498 rothmannia hispida rubiaceae lc 32 1.481481 -0.05211 irvingia gabonensis irvingaceae nt 8 0.246914 -0.01218 staudtia stipitate myristicaceae lc 72 2.222222 -0.07103 sterculia oblonga sterculiaceae lc 48 0.740741 -0.03012 sterculia rhinopetala sterculiaceae lc 11 0.493827 -0.02166 sterculia tragacantha sterculiaceae lc 32 0.987654 -0.03791 strombosia grandifolia olacaceae nl 448 6.91358 -0.15893 strombosia pustulata olacaceae lc 736 11.35802 -0.21649 tabernaemontana pachysiphon apocynaceae lc 16 1.234568 -0.0452 trichilia lanata meliaceae lc 56 1.728395 -0.05869 trilepisium madagascariense moraceae cr 72 2.222222 -0.07103 uapaca staudtii euphorbiaceae lc 96 1.481481 -0.05211 uapaca togoensis euphorbiaceae lc 64 0.987654 -0.03791 vitex doniana verbenaceae lc 16 0.493827 -0.02166 xylopia aethiopica annonaceae lc 144 2.222222 -0.07103 zanthoxylum zanthoxyloides rutaceae en 64 0.987654 -0.03791 3.83659 # ls (least concerned), v (vulnerable), nt (near threatened), nl (not listed), e (endangered), cr (critically endangered), rd (relative density) table 5: summary of tree species diversity indices and distribution in oban and okwangwo divisions of cross river national park (crnp). crnp division no. of tree species no. of families shannon-weiner diversity index (h') species evenness (eh) oban 80 25 3.88 0.62 okwango 78 22 3.84 0.53 b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 37 4 discussion cross river national park is a unique ecotourism site with its rich natural resources, geomorphological structure, rivers, climate and rich biological diversity. the core attractions of the park include wilderness hiking, game viewing, mountaineering, forest camping, rock formation, visits to the botanical garden, water recreation and village tours. hiking in the rainforest vegetation was the prominent activity as the park provides wonderful wilderness experience with its dense canopy forests, consisting of tall trees with huge buttresses, the coolness of the environment, the freshness of the air and the sweet songs of the forestdwelling birds. its rich flora provides suitable habitats to many wild animals such as the cross river gorilla, (gorilla gorilla diehli) and the chimpanzee (pan troglodytes) and the bare-headed rock fowl (picarthates oreas). birds and monkey call audible in the forest would be of interest to ecotourists. this agrees with drumm and more (2002) assertion about ecotourism attraction whether they are wildlife viewing possibilities or dramatic natural landscapes, tend to be found in protected natural areas. moreover, rajib and jaba (2012) stated that natural environment was the major purpose of tourists’ visit to bangladesh. the erokut rock formation is a major attraction to tourist where bat watching is prominent. bemi river (referred to as natural swimming pool) in butatong, okwango division and kwa river (shallow water) in erokut, oban division are good ecotourism attractions where tourist can dive and swim. plans are also on the ground by the park management to utilize the kwa river for sport fishing and canoe riding for ecotourists. other attractions include nature trail in kayang at the foot of mbe mountain, natural history museum in oban division, botanical garden in okwango division where anceistrocladus kurupensis that was discovered to have high medicinal properties effective against hiv/aids can be found (obot et al. 1996, crnp 2008). rock of picathartes oreas (rock fowl) endemic bird species is found in bashu (okwango division) a distance of 2 km gentle slope from the community known as bashu bird sanctuary. traditional lifestyle and culture are important and attractive elements for ecotourists. traditional architecture is an important element of the cultural landscape, with buildings of wood, bamboo and mud walls and grassthatched, high, pyramid-like roofs. the attribute of these resources in cross river national park makes the resources to be location specific for ecotourists and will have wide utility in the management of ecotourism in the park (shrestha 2006, srivastava and anitha 2010). although fifteen wild animal species belonging to nine families were identified, the conservation status of the animals encountered show that six of the species are stable (blue duiker cephalophus rufilatus, ground squirrel xerus erythropus, mongoose ichneumia albicauda, bush buck tragelaphus scriptus, tantalus monkey cercopithecus aethiops and yellow-spotted barbet buccanodon duchaillui), four of the species’ population is decreasing (red flanked duiker cephalophus rufilatus, red b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 38 river hog potamochoerus porcus, palm civet viverra civetta, fruit bats eidolon helvum) and the population change of four are unknown (porcupine altherurus africana, giant tree squirrel protoxerus stangeri, drill monkey mandrillus leucophaecus, and mona monkey cercopithecus mona). elephant loxodonta africana cyclotis is the only species whose population is increasing. the family leguminosae and apocynacea were dominant in oban and okwangwo divisions respectively. other dominant families in both divisions were meliaceae, euphorbiaceae, moraceae, olacaceae, guttiferae and sterculiaceae which are mostly found in nigerian rainforests as reported by adekunle and olagoke (2008) and onyekwelu et al. (2008). moreover, the diversity index and equitability (evenness) for the two divisions slightly differ. this slight variation signified that oban division is more diverse in woody tree species than okwango division. these diversities were above 3.5 and fall within the general limit recommended by kent and coker (1992). the species diversities for the two divisions further elucidate zakaria et al. (2016) assertion that tropical rainforests are the most diverse among the world ecosystem in vegetation structure and composition. 5 conclusions cross river national park with its natural resources of touristic significance is a powerful ecotourism destination. in spite of the tremendous resources available in crnp, ecotourism development in the park is still at its infancy stage, its potentials to support activities like wilderness hiking, game viewing, bird watching, mountaineering, village tour, forest camping, water recreation where tourist can dive and swim, participate in sport fishing and canoe ride cannot be overemphasized. there was a presence of diverse biological resources that are endemic to this area including those declared threatened or endangered. the diversity assessment shows that the two divisions were able to conserve both fauna and flora diversity however oban division was more diverse in woody tree species than okwango division. hence, the species richness of the park can be sustainably managed if the park is further prevented from anthropogenic effects occasioned by logging, hunting, and collection of forest products to improve the vegetation status and invariably enhance wildlife population in the park. acknowledgements the conservator general of nigeria national park service is appreciated for granting approval to carry out this study in the park, cross river national park management is acknowledged for providing research unit staff during the field work and dr. o.o. sobola’s contribution and insightful remarks is highly commended. three anonymous reviewers are thanked for providing constructive feedback on the initial and revised versions. b.o. adetola et al. cross river national park ecotourism resources ruhuna journal of science vol 12 (1): 26-39, june 2021 39 references adekunle vaj, olagoke ao. 2008. diversity and biovolume of tree species in natural forest ecosystem in the bitumen-producing areas of ondo state, nigeria. a baseline study. biodiversity and conservation 17: 2735-2755. crnp. 2008. federal ministry of environment: national park service. cross river national park annual report, 120pp. crnp. 2010. federal ministry of environment: national park service. cross river national park annual report, 139pp. diamantis d. 2004. ecotourism management: an overview. in d. diamantis (ed.), ecotourism–management and assessment. thompson learning, belmont, ca. london. 1-26pp. drumm a, moore a. 2002. ecotourism development: an introduction to ecotourism planning volume1, the nature conservancy, usa. 85pp. fao. 2009. food and agriculture organisation. national forest monitoring and assessment. manual for local level assessment of land degradation, sustainable land management and livelihood, land degradation assessment in dry lands (lada) project, 39-66pp. higginbottom k. 2004. wildlife tourism: impacts, management and planning. wildlife tourism. altona, vic.: common ground publishing for crc for sustainable tourism. 277pp. holden a. 2003. environment and tourism, 1st ed., rutledge, england. 624pp. ijeomah hm. 2007. the impact of tourism on poverty levels of households in adjoining ecotourism destinations in plateau state. ph.d thesis. department of wildlife and fisheries management. university of ibadan. nigeria, 86-105pp. keay rwj. 1989. trees of nigeria. a revised version of “nigerian trees” (keay et al., 1964). clarendon press oxford, 476pp. kent m, coker p. 1992. vegetation description and analysis. a practical approach, belhaven press, london, 363pp. luchman h. 2017. managing biodiversity for a competitive ecotourism industry in tropical developing countries. new opportunities in biological fields aip conference proceedings 1908, 030008; doi: 10.1063/1.5012708. newsome d, moore sa, dowling r. 2012. natural area tourism: ecology, impacts, and management. volume 58, channel view publications. obot ea, baker j, ogar g. 1996. biological surveys in cross river national park as contribution to park management. cross river national park (okwango division) occasional paper no. 2. proceedings of workshop on african rainforest and the conservation of biodiveristy. limbe botanical gardens, cameroon, 17th -24th january, 1996, 3-4pp. oduwyaiye ea, oyeleye b, oguntala, ab. 2002. species diversity and potentiality for forest regeneration in okomu sample plot in abu je, oni, p. i. and popoo;a. l. (eds) forestry and challnges of sustainable livelihood. in proceeding of the 28th annual conference of forestry assocciation of nigeria, akure. onyekwelu jc, mosandl r, stimm b. 2008. tree species diversities and soil status of primary and degraded tropical rainforest ecosystems in south-western nigeria. journal of tropical forest science 20: 193-204. price pw. 1997. insect ecology. john wiley and sons. isbn: 978-0-471-16184-4 august 1997 888pp. rajib kd, jaba c. 2012. an evaluative study on tourism in bangladesh. research journal of finance and accounting 2(1):17-28 shrestha h. 2006. globa positioning systems (gps) and geographical information system (gis) in participatory mapping of community forest in nepal. ejisdc 25 (5):1-11. smith slj. 2004. the measurement of global tourism: old debates, new consensus and continuing challenges. in: a. lew, c.m. hall and a. williams (eds) a companion to tourism. oxford: blackwell. srivastava v, anitha d. 2010. mapping of non-timber forest products using remote sensing and gis. tropical ecology 51(1): 107-116 udoidung nig, braide ie, opara kn, adie ha. 2007. perstans filariasis in rural communities of lower cross river basin nigeria. parasitological observation international journal of zoological research 3:207-212. zakaria m, rajpar mn, ozdemir i, rosli z. 2016. fauna diversity in tropical rainforest: threats from landuse change. available from: https://www.intechopen.com/books. retrieved on 22/01/2020. https://www.intechopen.com/books rjs-vol-1-sept-2006-16.dvi ruhuna journal of science vol. 1, september 2006, pp. 158–173 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. vertebrate diversity in a thirty year old analogue forest in pitigala, elpitiya, in the galle district of southern sri lanka s. n. gamage, w. k. d. d. liyanage, a. gunawardena department of animal science, faculty of agriculture, university of ruhuna, kamburupitiya, matara, sri lanka. asokag2006@yahoo.com s. wimalasuriya land owners restore rainforest in sri lanka, bangamukanda estate, pitigala, galle, sri lanka. most of the natural ecosystems in the wet zone are severely fragmented and interspersed between human managed agro ecosystems and home gardens. there is growing evidence that traditional agro-ecosystems contribute to sustain the regional biodiversity of many invertebrate and vertebrate species. analogue forest as a concept is accepted by agronomists and conservationists, which would bring profits in the long-term sustainable basis. the bangamukanda estate is an example of a 18 hectares plantation (tea, rubber and cinnamon) that has been converted into an analogue forest. objective of the study was to assess the current vertebrate diversity in this 30-year-old analogue forest. total of 206 species of vertebrates belonging to 74 families were observed during the study period, out of that 58 species were endemic to sri lanka. the findings of the survey clearly highlighted the contribution of analogue forest systems towards sustaining a rich biodiversity. in addition analogue forest systems can be used to link the forest patches in the wet zone. key words : vertebrate diversity, analogue forest, conservation 1. introduction since sounding the alarm of the biodiversity crisis in the 1970s, most conservationists have focused on establishing protected areas to conserve endangered habitats and species (scherr & shames 2006). agricultural production areas have been seen as useless for conservation activities and their growth viewed as a threat (scherr & shames 2006). in recent decades, sustainable farmers and researchers around the world have responded to the extractive industrial model with ecology based approaches such as eco-agriculture, agro-forestry or analogue forest (earles 2005, scherr & shames 2006). all of them, representing thousands of farms, have contributed to our understanding of what sustainable systems are, and each of them shares a vision of “farming with nature”, an agro-ecology that promotes biodiversity, recycles plant nutrients, protects soil from erosion, conserves and protect water. also uncultivated portions of mainly agricultural landscapes can provide patches of habitat for forest wildlife and form corridors that connect protected areas and allow species to 158 gamage et al.,: vertebrate diversity in a thirty year... ruhuna journal of science 1, pp. 158–173, (2006) 159 continue genetic contact with populations that would otherwise be isolated (scherr & shames 2006). for example, millions of hectares of multi-strata ‘agro-forests’ in indonesia produce commercial rubber, fruits, spices and timber. the number of wild plant and animal species in these agro-forests is often nearly as high as in natural forests (scherr & shames 2006). there is growing evidence that traditional agroecosystems contribute to sustain the regional biodiversity of many invertebrate and vertebrate species (lawler, 2001). vast extents of sri lanka’s biodiversity rich lands which were transferred into mono crop plantations during the colonial era are regenerating in many places due to various reasons, both natural and man made. the bangamukanda estate is an example which consists of 18 hectares plantation land (tea, rubber and cinnamon) that has been deliberately converted to an analogue forest as a direct result of the far sighted, land use policy of the 1970 -1977 government which introduced crop diversification of uneconomic tea lands. the bangamukande estate is situated in pitigala, galle, sri lanka. the land was transformed into an undulating terrain that consists of a series of ridges and valleys with an altitudinal range from 100 m to 300 m. it has an intricate network of small streams, which drains into the river benthara. in 1904 ancestors of the present owner planted agricultural mono crops such as cinnamon, rubber, and tea . this practice continued up to 1973. it was changed in 1973 and 12 hectares of cinnamon and tea land were transferred into analogue forest using a government subsidy, under crop diversification of uneconomic tea lands. the remaining rubber field of 6 hectares is presently allowed to regenerate into forestland while being tapped. sri lanka’s point of view is that, the primary natural ecosystems found in the low country wet zone consist of lowland rainforests, which are severely fragmented and interspersed between human managed agro ecosystems and home gardens. these wet-zone ecosystems harbour a high percentage of endemic and globally threatened species of animals and plants (gunatilleke et al., 2005; pethiyagoda, 2005). wetzone of sri lanka along with the western ghats is recognized as one of the world’s 11 biodiversity “hyperhot” hotspots, which demand extensive conservation investment (myers et al. 2000, brookes et al. 2002). however agro ecosystems and human settlements cover most of the land area in the wet-zone of sri lanka (gamage 2005). these habitats are frequently subjected to human modification and therefore the environment of these habitats is constantly changing whose impact on the biodiversity is little known. however, these man made habitats function as an integral part of the habitats of large number of fauna and flora but most of the studies are presently confined to herpetofauna and freshwater fish (gamage 2000, gamage et al. 2002, gamage et al. 2005). analogue forest is a tree-dominated ecosystem that is analogous in structure and function to the original climax and sub-climax forest community. with time, the natural succession of any undisturbed forest community is to increase in diversity and stability until a highly complex ecosystem or climax state is reached. when an ecosystem is designed to mimic the indigenous climax state, the efficiency and gamage et al.,: vertebrate diversity in a thirty year... 160 ruhuna journal of science 1, pp. 158–173, (2006) dynamics of the natural processes can be replicated. such forests are referred to as analogue forests which are considered to provide economic benefits. a wide range of supplies can be produced that may include: fruit, nuts, herbs, cut flowers and cut-foliage, pharmaceuticals, and timber. furthermore this type of concept can be used to link the fragmented forest patches in the wet zone of sri lanka. therefore the main objective of this study was to assess the diversity of vertebrate fauna in this 30 years old analogue forest (bangamukanda estate). 2. materials and methods the bangamukande estate is situated in niyagama divisional secretate area in galle district of southern province of sri lanka, at 060 20’ 46” n 0800 16’ 26” e, average annual rainfall 2300 mm, average temperature 280, and humidity 90%. approximate distances from bke to the larger forest complexes are as follows: south 4 km to polgahakande-malabure forest reserve east 1 km to hiniduma forest reserve southwest 8 km to beraliya forest reserve southeast 100 m to bangamukanda proposed forest southeast 8 km to kannaliya-dediyagala-nakiyadeniya forest reserve northeast 12.5 km to sinharaja forest reserve world heritage site north 11 km to kalugalkande forest hermitage and reserve different methods were used to assess the vertebrate fauna in the bangamukanda estate. the study was carried out during august 2003 to november 2005. 2.1. herpetofauna the quadrate sampling method was the main method used to study the herpetofauna. it involves placing small squares (quadrates) at randomly selected sites within a habitat and thoroughly searching these squares for presence of herpetofauna (heinen 1992). quadrate sampling was done during september 2003 to november 2003. a total of eighteen 8 × 8 m quadrates were placed at randomly selected points in each study site. in placing of quadrates, areas with a high relief angle or areas adjacent to tree-fall gaps were omitted. all of the sampled quadrates in the agroecosystems were located within 1-2 km from natural vegetations. a 45cm height polythene fence was placed along the sides of the quadrate to prevent animals from escaping. a minimum of two people was engaged in all of the sampling sessions. sampling involved sorting through all leaf litter in the plot, tree trunks, branches, under stones and logs (fauth et al. 1989, heinen 1992). furthermore fixed line traces were also used to assess the herpetofauna. 2.2. avifauna and mammals fixed line transect method was used to assess avian and mammalian richness of the study site (sutherland 1996). day and night surveys were carried out during a period from august 2003 to november 2005. field observations were made from 6.30 am to 9.00 am and 4.00 pm to 6.00 pm. in addition night observations were made from 7.00 pm to 10.00 pm and 2.00 am to 6.00 am. gamage et al.,: vertebrate diversity in a thirty year... ruhuna journal of science 1, pp. 158–173, (2006) 161 table 1 the number of vertebrate species, families and endemic species recorded in each taxonomic group during the study period. vertebrate group total number number of number of number of of endemic species families endemic species species recorded in sri lanka amphibians 18 3 12 79 snakes 25 5 9 46 tetra pods reptiles 17 5 8 48 fish 23 9 13 44 birds 89 34 10 25 mammals 34 18 6 16 total 206 74 58 258 2.3. identification the different groups of vertebrates were identified using the most recent taxonomic keys and guides available for the respective taxonomic group (freshwater fish: pethiyagoda (1991), deraniyagala (1949); amphibians: dutta & manamendraaarachchi (1996), manamendra-arachchi & pethiyagoda (1998), manamendraarachchi & pethiyagoda (2005), meegaskumbura & manamendra-arachchi (2005); serpentoid reptiles: de silva (1990), pethiyagoda & manamendra-arachchi, (1998), das & de silva (2005); birds: henry (1971), kotagama & fernando (1992); mammals: phillips (1981), corbet & hill (1992), groves (2001). bambaradeniya eds. (2006) was used for the confirmation of nomenclature.) 3. results in the course of this study 206 species of vertebrates belonging to 74 families were observed out of which 58 species were endemic to sri lanka (table 1). the vertebrate fauna was comprised 18 species of amphibians, 25 species of snakes, 17 species of tetrapod, reptiles, 23 species of fish, 89 species of birds and 34 species of mammals (table 01 & appendix). a total of 54 endemic vertebrate species were recorded during the survey which include 12 amphibians and 13 fresh water fish. table 2 shows the conservation status of some threatened species found in the study site of which four were vulnerable, six were endangered, one critically endangered and one was data deficient. this critically endangered frog (philatus nemus) is a newly discovered species which was only found in hiniduma kanda forest reserve previously (manamendra-arachchi & pethiyagoda, 2005). 4. discussion the results indicate that the bangamukanda estate (analogue forest) is an agroecosystem that sustains a high diversity of vertebrate fauna. a variety of methods targeting at different groups enabled the documentation of vertebrate diversity in bangamukanda estate expressed in terms of species richness. the total vertebrate richness shows that bangamukanda estate harbours a comparatively high number of species. in addition to the species richness, the study gamage et al.,: vertebrate diversity in a thirty year... 162 ruhuna journal of science 1, pp. 158–173, (2006) table 2 conservation status of some threatened species, recorded in bangamukanda estate. species conservation status rana aurantiaca vulnerable (vu) nanophrys ceylonensis vulnerable (vu) polipedates longinasus endangered (en) polypedates eques endangered (en) philatus nemus critically endangered (cr) philatus folicola endangered (en) lepidocephalichthys jonklaasi endangered (en) sicyopus jonkalaasi data deficient (dd) centropus chlorohyncho vulnerable (vu) loris tardigradus tardigradus endangered (en) trachypithecus vetullus vetullus endangered (en) macaca sinica aurifrons vulnerable (vu) site is also evident for providing niches for a large number of endemic vertebrates. the results clearly indicate that such agro-forestry systems are closer to natural conditions and maintain high biodiversity. the study site is providing niches for nine globally threatened species of which one species is critically endangered. this clearly shows the importance of this ecosystem. most of the birds and mammal species are using this estate as a temporary refuge ground or feeding area, while they move from one forest patch to another suggesting that further studies are necessary to evaluate the importance of agricultural systems as means of connecting forest patches in the country. 5. conclusion according to the results, it can be concluded that analogue forest systems are sustaining high level of vertebrate diversity and endemism. as a concept analogue forestry system is biodynamic and environmentally friendly (hochegger 1998; earles 2005). our study confirmed this concept. in addition the findings of the survey clearly highlighted the contribution of analogue forest systems towards sustaining rich biodiversity. such agro-ecosystems can be used to link the forest patches in the wet zone. however, a detailed study on analogue forest systems has to be carried out for further confirmation of the validity of the concept and to plan conservation strategies to increase biodiversity in the agro ecosystems. references brookes, t.m., mittermeier, r.a., mittermeier, c.g., fonseca, g.a.b.da, rylands, a.b., konstant, w.r., flick, p., pilgrim, j., oldfield, s., magin, g. & hiltontaylor, c. (2002). habitat loss and extinction in the hotspots of biodiversity. conservation biology. 16: 909-923. bambaradeniya, c.n.b. (editor). (2006). fauna of sri lanka: status of taxonomy, research & conservation. the world conservation union, colombo, sri lanka & government of sri lanka. viii + 308pp. gamage et al.,: vertebrate diversity in a thirty year... ruhuna journal of science 1, pp. 158–173, (2006) 163 corbert, g.b. & hill, j.e. (1992). mammals of the indomalayan region: a systematic review. oxford university, oxford, uk. deraniyagala, p.e.p. (1949). some vertebrate animals of ceylon. national museum pictorial series, 1: 1-41. de silva a. (1990). colour guide to the snakes of sri lanka. r & a publishing ltd., avon, england, 130pp. dutta, s.k., & manamendra-arachchi, k. (1996). the amphibian fauna of sri lanka. wildlife heritage trust of sri lanka, colombo, 280pp. earles, r. (2005). sustainable agriculture: an introduction publication of attra, the national sustainable agriculture information service. usa. fauth, j.e., crother, b.i. & slowinski, j.b. (1989). elevational patterns of richness, evenness, and abundance of the costa rican leaf litter herpetofauna. biotropica 21(2): 55-73. gamage, s.n. (2000). fresh water fish diversity and habitat variations in streams under different vegetation in hiniduma, sri lankain . bsc thesis. faculty of agriculture, university of ruhuna. kamburupitiya, sri lanka. gamage, s.n., liyanage, w.k.d.d. & guanawardena, a. (2002). comparison of fish diversity and the habitat structure of streams in rubber and oil-palm plantations in nakiyadeniya galle. sri lanka association for the advancement of science, annual conference 2002, university of colombo. gamage, s.n. (2005). a comparative study on biodiversity of selected manmade and natural habitats in low country wet zone of sri lanka. mphil thesis. faculty of agriculture, university of ruhuna. kamburupitiya, sri lanka. groves, p.c. (2001). primate taxonomy. smithsonian institution press, washington d.c. gunatilleke, i.a.u.n., gunatilleke, c.v.s. & dilhan, m.a.a.b. (2005). plant biogeography and conservation of the southwestern hill forests of sri lanka. the raffles bulletin of zoology, supplement no. 12: 9-22 henry, g.m. (1971) a guide to the birds of ceylon (sri lanka) with 30 half-tone plates of which 27 are coloured and 136 black and white drawings. (2nd edition). k.v.g. de silva & sons, kandy, ceylon (sri lanka). 457pp. heinen, j.h. (1992). comparisons of the leaf litter herpetofauna in abundoned cacao plantaions and primary rain forest in costa rica: some implications for faunal restoration. biotropica 24(3): 431-439. hochegger, k. (1998). farming like the forest: traditional home garden systems in sri lanka. tropical agroecology 191, margraf verlag, wikersheim, germany, 203pp. kotagama, s.w. & fernando, p. (1992). a field guide to the birds of sri lanka. wildlife heritage trust of sri lanka, colombo 8, srilanka. 226pp. lawler, s. p. (2001). rice fields as temporary wetlands: a review. israel j. of zoology, 47 (3), 513-528. gamage et al.,: vertebrate diversity in a thirty year... 164 ruhuna journal of science 1, pp. 158–173, (2006) manamendra-arachchi, k., & pethiyagoda, r. (1998). a synopsis of sri lankan bufonidae (amphibia: anura) with discription of new species. j. south asian nat. hist., 3(1), 213-247. manamendra-arachchi, k., & pethiyagoda, r. (2005). the sri lankan shrub-frogs of genus philautus gistel., 1848 (ranidae: rhacophorinae), with discription of 27 new species. the raffles bulletin of zoology, supplement no. 12: 163-303. meegaskumbura, m., & manamendra-arachchi, k. (2005). description of eight new species of shrub frogs (ranidae: rhacophorinae: philautus) from sri lanka. the raffles bulletin of zoology, supplement no. 12: 305-338 myers, n., mittermeier, r.a., mittermeier, c.g., fonseca, g.a.b.da & kent, j. (2000). biodiversity hot spots for conservation priorities. nature. 403: 853-858. phillips, w.a.a. (1981). manual of the mamals of sri lanka. wildlife and nature protection society of ceylon (sri lanka). colombo. vol. i, ii, & iii. pethiyagoda, r. (1991). freshwater fishes of sri lanka. wildlife heritage trust, colombo. xiv+362pp. pethiyagoda, r. (2005). exploring sri lanka’s biodiversity. the raffles bulletin of zoology, supplement no. 12: 1-4 pethiyagoda, r., & manamendra-arachchi, k. (1998) evaluating sri lanka’s amphibian diversity. occ. pap. wildlife heritage trust. 2:1-12. pethiyagoda, r. & kottelat, m. (2005). a review of the barbs of the puntius filamentosus group (teleostei: cyprinidae) of southern india and sri lanka. in: yeo, d.c.j.p.k.l. ng & r. pethiyagoda (eds.), contributions to biodiversity exploration and research in sri lanka. the raffles bulletin of zoology, supplement no. 12: 127-144. scherr, s.j. & shames, s. (2006). agriculture: a threat or promise for biodiversity conservation. arborvitæ the iucn/wwf forest conservation newsletter. sutherland, j.w. (1996) ecological census techniques. cambridge university press. uk. 336pp. acknowledgments we wish to acknowledge members of lorris and beog for invaluable assistance provided in the field. very special thank to mr. dharma sri kandambii (curator, national maritime museum, galle) for providing advise and guidance in the taxonomic identification. member’s of the galle wild life conservation society are greatly acknowledge for their great assistance in the taxonomic identification of fauna. dr. k.a.i. nekaris and lilia bernede provided necessary literature and equipment for the survey. we also thank nhk japan and nocturnal primate research group (oxford brooks university) for providing some funds. appendix gamage et al.,: vertebrate diversity in a thirty year... ruhuna journal of science 1, pp. 158–173, (2006) 165 list of freshwater fauna observed at bke. sub class: osteichtheys family: aplocheilidae 1. *aplocheilus werneri werner’s killifish family: anguillidae 2. anguilla bicolour level finned eel family: bagridae 3. mystus gulio long whiskered catfish 4. mystus keletius yellow catfish family: balitoridae 5. *schistura notostigma banded mountain loach family: cobitidae 6. *lepidocephalichthys jonklaasi jonklas loach (en) 7. lepidocephalichthys thermalis common spiny loach family: cyprinidae 8. danio malabaricus giant danio 9. *esomus thermoicos flying barb 10. *garra ceylonensis stone sucker 11. puntius amphibious scarlet-banded barb 12. puntius bimaculatus redside barb 13. *puntius cumingii cuming’s barb (endemic * ) list of amphibian species observed at bke. class: amphibia order: apoda family: ichthyophiidae 1. *ichthyophis glutinosus (linnaeus 1758) common yellow-band cecillian order: anura family: bufonidae 2. bufo melanostictus schneider 1799 common house toad 3. *bufo atukoralei bogert & senanayaka 1966 athukorala’s dwarf toad family: ranidae subfamily: raninae 4. rana aurantiaca boulenger 1904 golden frog (vu) / small wood frog 5. rana temporalis (günther, 1864) bronzed frog / common wood frog 6. *fejervarya (limnonectes) kirtisinghei kirtisinghe’s frog manamendra-arachchi & gabadage, 1994 7. fejervarya (limnonectes) limnocharis (boie,1835) common paddy field frog 8. hoplobatrachus crassus (jerdon 1853) jerdon’s bull frog gamage et al.,: vertebrate diversity in a thirty year... 166 ruhuna journal of science 1, pp. 158–173, (2006) 9. *nanophrys ceylonensis (günther, 1864) sri lankan rock frog (vu) 10. *lankanectes corrugatus (peter 1863) corrugated water frog 11. euphlyctis hexadactylus (lesson, 1834) indian green frog / sixtoe green frog 12. euphlyctis cyanophlyctis (schneider, 1799) skipper frog subfamily: rhacophorinae 13. *polypedates eques(günther, 1858) saddled tree frog (en) 14. *polypedates cruciger blyth, 1852 common hourglass treefrog 15. *polypedates longinasus (ahl, 1931) long-snouted tree frog (en) / sharp snout saddle tree frog 16. *philautus abundus labugama shrub frog manamendra-arachchi & pethiyagoda, 2005 17. *philautus folicola anthropogenic shrub frog (en) manamendra-arachchi & pethiyagoda, 2005 18. *philautus. nemus southern shrub frog (cr) manamendra-arachchi & pethiyagoda, 2005 list of reptile species observed at bke. class-reptiliya order: serpentes family: boidae 1. python molurus (linnaeus, 1758) rock python / indian python family: elapidae 2. naja naja (linnaeus, 1758) indian cobra 3. *bungarus ceylonicus günther, 1864 sri lanka krait family: colubridae 4. *xenochrophis asperrimus (boulenger,1891) sri lanka keelback / common pond snake 5. xenochrophis piscator (schneider, 1799) checkered keelback 6. *balanophis ceylonensis (günther, 1858) sri lanka blossom krait / sri lanka keelback 7. *aspidura guentheri ferguson, 1876 gunther’s rough-side 8. amphiesma stolata (linnaeus, 1758) buff strip keelback 9. ahaetulla nasutus (lacepede, 1789) green vine snake gamage et al.,: vertebrate diversity in a thirty year... ruhuna journal of science 1, pp. 158–173, (2006) 167 10. boiga ceylonensis (günther, 1858) sri lanka cat snake 11. boiga forsteni (dumeril, bibron & dumeril, 1854) forten’s cat snake 12. chrysopelea ornate (shaw, 1802) ornate flying snake 13. * dendrelaphis bifernalis (boulenger, 1890) boulenger’s bronze-back 14. dendrelaphis tristis (daudin, 1803) common bronze-back 15. dryocalamus nympha (daudin, 1803) common bridal snake 16. elaphe helena (daudin, 1803) trinket snake 17. lycodon aulicus (linnneaus, 1758) common wolf snake 18. *lycodon striatus (shaw, 1802) shaw’s wolf snake 19. coluber mucosus (linnaeus, 1758) rat snake 20. oligodon arnensis (shaw, 1802) common banded kukri snake 21. *oligodon sublineatus streaked kukri snake / dumeril’s kukri snake (dumeril, bibron & dumeril, 1854) family: cylindrophiidae / europeltidae 22. *cylindrophis maculatus (daudin, 1803) sri lanka pipe snake family: viperidae 23. hypnale hypnale (merrem, 1820) hump-nosed viper 24. *trimeresurus trigonocephalaus sri lanka green pit viper (sonnini & latreille, 1801) 25. daboia russelii (shaw & nodder, 1797) russell’s viper order: sauria family: agamidae 1. calotes calotes (linnaeus, 1758) green garden lizard 2. calotes versicolor (daudin, 1802) common garden lizard 3. *calotes liolepis boulenger, 1885 whistling lizard 4. *ceratophora aspera günther, 1864 rough-horn lizard 5. *otocryptis wiegmanni wagler, 1830 sri lanka kangaroo lizard 6. *lyriocephalus scutatus (linnaeus, 1758) hump nose lizard family: scincidae 7. mabuya carinata lanka deraniyagala, 1953 rat snake skink 8. *nessia burtonii gray, 1839 three-toe snake skink 9. *lankascincus fallax (peters, 1860) common lanka skink 10. *lankascincus gansi (greer, 1991) gans’s lanka skink family: varanidae 11. varanus bengalensis (daudin, 1802) land monitor 12. varanus salvator (laurenti, 1768) water monitor gamage et al.,: vertebrate diversity in a thirty year... 168 ruhuna journal of science 1, pp. 158–173, (2006) family: gekkonidae 13. *cnemaspis podihuna deraniyagala, 1953 dwarf day gecko 14. hemidactylus frenatus dumeril & bibron, 1836 asian house gecko 15. hemidactylus brooki gray, 1845 brooke’s house gecko / spotted house gecko 16. gehyra mutilata (wiegmann, 1834) four-clawed gecko family: trionychidae 17. lissemys punctata (lacepede, 1788) flap-shell turtle list of bird species observed at bke. order: ciconiiformes family: phalacrocoracidae 1. phalacrocorax niger little cormorant family: ardeidae 2. bubulcus ibis cattle egret 3. egretta garzetta little egret 4. ardeola grayii indian pond heron family: accipitridae 5. ictinaetus malayensis black eagle 6. haliastur indus brahmini kite 7. spizaetus cirrhatus changable hawk eagle 8. spilornis cheela crested serpent eagle 9. accipiter badius shikra order: galiformes family: phasianidae 10. *gallus lafayetii sri lanka junglefowl 11. *galloperdix bicalcarata sri lanka spurfowl order: gruiformes family: rallidae 12. amaurornis phoenicurus white breasted water hen order: columbiformes family: columbidae 13. chalcophaps indica emerald dove 14. ducula aenea green imperial pigeon 15. treron bicincta orange breasted green pigeon 16. treron pompadora pompadour green pigeon gamage et al.,: vertebrate diversity in a thirty year... ruhuna journal of science 1, pp. 158–173, (2006) 169 17. streptopelia chinensis spotted dove order: psittaciformes family: psittacidae 18. psittacula krameri rose-ringed parakeet 19. psittacula cyanocephala plum-headed parakeet 20. *loriculus beryllinus sri lanka hanging parakeet order: cuculiformes family: cuculidae 21. eudynamys scolopacea asian koel family: centropodidae 22. centropus sinensis greater coucal 23. * centropus chlororhynchus sri lanka green-billed coucal order: strigiformes family: strigidae 24. bubo nipalensis spot-bellied eagle owl 25. *glaucidium castanonotum sri lanka chestnutbacked owlet 26. strix leptogrammica brown wood owl 27. ketupa zeylonensis brown fish owl family: batrachostomidae 28. batrachostomus monilieger frog mouth order: apodiformes family: apodidae 29. cypsiurus balasiensis asian palm swift family: hemiprocnidae 30. hemiprocne coronata crested tree swift order: trogoniformes family: trogonidae 31. harpactes fasciatus malabar trogon order: coraciiformes family: alcedinidae 32. ceyx erithacus oriental dwarf kingfisher 33. alcedo atthis common kingfisher family halcyonidae 34. halcyon smyrnensis white-breasted kingfisher family: meropidae 35. merops philippinus blue-tailed bee-eater gamage et al.,: vertebrate diversity in a thirty year... 170 ruhuna journal of science 1, pp. 158–173, (2006) 36. merops leschenaultia chestnutheaded beeeater family: coraciidae 37. eurystomus orientalis dollarbird order: bucerotiformes family: bucerotidae 38. *ocyceros gingalensis sri lanka grey hornbill order: piciformes family: megalaimidae 39. megalaima zeylanica brown-headed barbet 40. megalaima rubricapilla crimson-fronted barbet 41. * megalaima flavifrons sri lanka yellow-fronted barbet family: picidae 42. chrysocolaptes lucidus greater flame-back 43. dendrocopus nanus pigmy woodpecker 44. dinopium benghalense red-backed woodpecker order: passeriformes family: pittidae 45. pitta brachyura indian pitta family: passeridae 46. dendrolanthus indicus forest wagtail family: hirundinidae 47. hirundo daurica red-rumped swallow family: corvidae 48. crovus macrorhynchos large-billed crow / black crow/ jungle crow 49. pericrocotus flammeus scarlet minivet 50. pericrocotus cinnamomeus small minivet 51. aegithina tiphia common iora 52. terpsiphone paradisi asian paradise flycatcher 53. rhipidura aureola white-browed fantail 54. oriolus xanthornus black-hooded oriole / black headed oriole 55. dicrurus caerulescens white-bellied drongo 56. dicrurus paradisius lophorhinus crested drongo / great racket-tailed drongo 57. artamus fuscus ashy woodswallow family: pycnonotidae 58. hypsipetes leucocephalus black bulbul 59. pycnonotus melanicterus black-crested bulbul 60. pycnonotus cafer red-vented bulbul gamage et al.,: vertebrate diversity in a thirty year... ruhuna journal of science 1, pp. 158–173, (2006) 171 61. pycnonotus luteolus white-browed bulbul 62. iole indica yellow-browed bulbul family: passeridae 63. lonchura striata white-rumped muniya 64. lonchura punctulata scaly-breasted muniya family: irenidae 65. chloropsis cochinchinensis blue-winged leafbird 66. chloropsis aurifrons gold-fronted leafbird family: laniidae 67. lanius cristatus cristatus brown shrike family: muscicapidae 68. muscicapa daurica asian brown flycatcher 69. hypothymis azurea black-naped monarch 70. copsychus saularis oriental magpie robin 71. cyornis tickelliae tickell’s blue flycatcher family: sittidae 72. sitta frontalis velvet-fronted nuthatch family: silviidae 73. orthotomus sutorius common tailorbird 74. phylloscopus trochiloides greenish warbler 75. phylloscopus magnirostris large-billed leaf warbler 76. turdoides affinis yellow-billed babbler 77. rhopocichla atriceps dark-fronted babbler 78. *pellorneum fuscocapillum sri lanka brown capped babbler 79. pomatorhinus horsfieldii scimitar babbler family: paridae 80. parus major great tit family: nectarinidae 81. *dicaeum vincens sri lanka legge’s flowerpecker 82. dicaeum erythrorhynchos small flowerpecker 83. nectarina zeylonica purple-rumped sunbird 84. nectarina lotenia long-billed sunbird / loten’s sunbird 85. nectarina asiatica purple sunbird family: zosteropidae 86. zosterops palpebrosa small white-eye / oriental white eye family: sturnidae 87. acridotheres tristis common myna 88. gracula religiosa hill myna 89. * gracula ptilogenys sri lanka myna gamage et al.,: vertebrate diversity in a thirty year... 172 ruhuna journal of science 1, pp. 158–173, (2006) list of mammal species observed at bke. class-mammalia subclass-theria order: chiroptera family: pteropidae 1. cynopterus sphinx short-nosed fruit bat 2. pteropus giganteus flying fox family: emballonuridae 3. taphozous melanopogon black-bearded sheath-tailed bat family: rhinolophidae 4. rhinolophus rouxii rufus horseshoe bat 5. hipposideros lankadiva great leaf-nosed bat family: vespertilionidae 6. pipistrellus ceylonicus kelaart’s pipistrel 7. kirivoula pictus painted bat order: primates family: loridae 8. *loris tardigradus tardigradus sri lanka western red slender loris (en) family: cercopithecidae 9. *macaca sinica aurifrons dusky toque macaque (vu) / sri lanka toque monkey 10. *trachypithecus vetullus vetullus sri lanka purple faced leaf monkey (en) order: rodentia family: sciuridae 11. funambulus palmarum palm squirrel 12. *funambulus layardi sri lanka flame-striped jungle squirrel 13. funambulus sublineatus dusky-striped jungle squirrel 14. ratufa macroura melanochra black and yellow giant squirrel / giant squirrel family: muridae 15. bandicota indica malabar bandicoot 16. mus booduga field mouse 17. mus musculus indian house mouse 18. rattus rattus common house rat 19. vandeleuria oleracea long-tailed tree mouse family: hystricidae 20. hystrix indica porcupine order: pholidota family: manidae 21. manis crassicaudata indian pangolin order: lagomorpha family: leporidae 22. lepus nigricollis black-napped hare gamage et al.,: vertebrate diversity in a thirty year... ruhuna journal of science 1, pp. 158–173, (2006) 173 order: carnivora family: viverridae 23. viverricula indica ring-tailed civet 24. paradoxurus hermaphroditus palm cat 25. *paradoxurus zeylonensis sri lanka golden palm civet family: herpestidae 26. herpestes brachyurus brown mongoose 27. herpestes smithii black-tipped mongoose family: felidae 28. prionailurus rubiginosa rusty-spotted cat 29. panthera paradus kotiy leopard 30. prionailurus viverrinus fishing cat family: mustelidae 31. lutra lutra otter family: canidae 32. canis aureus jackal order: artiodactyla family: suidae 33. sus scrofa wild boar family: tragulidae 34. *moschiola meminna sri lanka mouse deer ruhuna journal of science vol 12 (2): 155-166, december 2021 eissn: 2536-8400 © faculty of science http://doi.org/10.4038/rjs.v12i2.109 university of ruhuna © faculty of science, university of ruhuna sri lanka 155 effect of nitrogen fertilizer, weed control and seed rate on incidence and severity of narrow brown leaf spot in rice cultivation under the dry zone of sri lanka w.m.d.m. wickramasinghe1*, t.d.c. priyadarshani1, u.s. herath1, w.c.p. egodawatta1, d.i.d.s. beneragama2 and u.g.a.i. sirisena1 1department of plant sciences, faculty of agriculture, rajarata university of sri lanka, anuradhapura, sri lanka 2 department of plant sciences, university of saskatchewan, canada *correspondence: dharshikawickramasinghe@gmail.com ; orcid: https://orcid.org/0000-0002-0267-9067 received: 31st march 2021, revised: 24th october, accepted: 14th december 2021 abstract narrow brown leaf spot (nbls) caused by cercospora janseana is a common disease of rice causing severe yield loss. in order to identify the factors favourable for disease development under field conditions, a study was carried out during the 2017/2018 maha and 2018 yala seasons. effects of mineral n fertilizer, weed control, and seed rate on the incidence and severity of nbls disease were determined. two levels of mineral n, i.e., department of agriculture, sri lanka (dof) recommended level and half of the dof level were used with weedy and weed-free conditions, under four different seed rates, 100 kg/ha, 125 kg/ha, 150 kg/ha and 175 kg/ha. nbls incidence was significantly higher (p<0.0001) in the maha season compared to the yala season. in the maha season, mineral n fertilizer by weed interaction was significant on disease severity. in the yala season, the disease incidence was significantly (p<0.05) higher in weedy conditions and fertilizer, weed and seed rate interaction and fertilizer and seed rate interaction were significantly (p<0.05) higher. the lowest disease severity was recorded in both seasons with the 100kg/ha seed rate, 100% fertilizer and weed-free conditions. when the weedy conditions prevailed in the field in maha season, a disease severity scale value of 2 was observed at the highest frequency. although only the weed condition affects nbls incidence, it was found that the seed rate, fertilizer and, weed condition interaction was critical to control the severity of nbls in paddy cultivation. keywords: cercospora janseana, n fertilizer, paddy cultivation, yala/maha seasons 1 introduction rice (oryza sativa l.) is the staple food and the most important grain cultivated in sri lanka. the total land extent under rice in the country is 792,000 ha and the https://rjs.ruh.ac.lk/index.php/rjs/index https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v12i2.109 https://creativecommons.org/licenses/by-nc/4.0/ mailto:dharshikawickramasinghe@gmail.com wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 156 annual production stands at 2,383,000 mt (economic and social statistics of sri lanka, central bank of sri lanka, 2018). the rice is cultivated in two seasons of the year, i.e., maha and yala, determined by the annual monsoonal rainfall pattern. with the increasing population, there is a demand for increased production of rice. however, it can only be fulfilled through overcoming abiotic and biotic stress conditions (weerakoon et al. 2018). common biotic stresses encountered in rice cultivation include diseases, insect pests, and weeds. among these, pathogens causing diseases are major constraints in paddy production that cause considerable yield losses. the fungal disease, narrow brown leaf spot (nbls) causes a considerable reduction in the yield of rice cultivation in sri lanka (department of agriculture, sri lanka, 2019). narrow brown leaf spot caused by the fungus cercospora janseana infects leaves, sheaths, and panicles of the rice plant. the occurrence of nbls has been recorded in many rice growing areas in the world including the tropics and subtropics of asia, australia, africa and america (simanjuntak et al. 2020). the symptoms appear in the mature stages of the crop reducing the market value of the grains. light to dark brown linear lesions that are parallel to the veins occur in infected leaves and upper leaf sheaths, enlarging to make large necrotic areas in more susceptible varieties (irri 2019). the low external input and organic crop production are gaining momentum both globally and nationally due to verified negative effects of high input conventional crop production systems on human health and the environment (reganold and wachter 2016). the transition towards more sustainable alternative crop production systems is required to cut down external inputs and substitute them with more ecological and biological approaches to manage biotic and abiotic stresses. ecologically based weed management and integrated soil nutrient management with fewer herbicides and fertilizers have become important in this scenario. weeds account for 9.5% of annual rice yield losses globally (alam 2003, rabbani et al. 2011, hakim et al. 2013) and 30–40% in sri lanka (abeysekera 2001, hakim et al. 2013). these unwanted plants can significantly influence disease incidence by acting as pests, serving as an alternative host for pests, pathogens, and vectors (wisler 2009). the crop has to compete with the weeds for resources subsequently weakening and increasing the probability of pathogen infection (schreiber et al. 2018). it is recommended to remove weeds in the field in order to remove alternative hosts of the fungus, cercospora janseana since the presence of weeds in the field has a direct impact on the nbls. furthermore, increasing crop density by increasing seeding rates has been identified as one of the major ecological strategies to outcompete weeds in cereals as well as to increase resource use efficiency. the competitiveness of the crop can be increased by employing a high seed rate (chauhan 2012, ahmed et al. 2014) to facilitate a quick canopy closure and decreased weed growth (guillermo 2009, ahmed et al. 2014). however, such ecological changes in the cropping systems can alter other biological and ecological processes due to changes in the microclimate, particularly by negatively influencing the disease development in the crop. wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 157 moving from conventional to alternative production systems can cause a reduction in plant-available n at least during the transition periods. the low availability of n can also influence disease dynamics in crops. plant growth at high n availability may result in increased plant susceptibility to pathogens due to increased foliar n concentrations (mitchell et al. 2003, veresoglou et al. 2013). however, panique et al. (1997) stated that disease resistance of plants is occasionally reduced due to fertilization, but disease tolerance is increased due to the plant growth stimulating effect caused by nutrient availability (veresoglou et al. 2013). we hypothesized that the agronomy of rice crops; especially the n status of the crop, and weeds may have an impact on the occurrence of nbls disease. therefore, this study was conducted to compare the seasonal variation of incidence and severity of the disease in the maha and yala seasons and to assess the effect of mineral n fertilizer dosage, control of weeds, and seed rates on disease incidence and severity of nbls. 2 material and methods the experiment was carried out in 2017/2018 maha (2017 november – 2018 march) and 2018 yala (april – august) seasons at the research field of the faculty of agriculture, rajarata university of sri lanka (rusl), puliyankulama, anuradhapura. a field trial was established with three treatments; two mineral n fertilizer dosages i.e., full (100%) department of agriculture (doa) recommendation and a half (50%) of doa recommendation (table 1) in the presence of weeds, i.e., weedy and weed-free, and at four seed rates, i.e., 100 kg/ha, 125 kg/ha (control),150 kg/ha and 175 kg/ha. the treatments were laid out on a factorial, splitsplit plot design with four replicates. the size of the plot was 2m x 2m. the n fertilizer rate was the main plot factor and the weed population was the sub-plot factor, while the seeding rate was the sub-sub plot factor. main plots and subplots were separated with bunds to escape the lateral movement of fertilizers. to avoid the movement of air-borne pathogens, 1.5m height white colour polyethylene sheets were used to cover each plot. 2.1 crop establishment and management pre-germinated rice of variety bg 300 was direct seeded into puddled soil following the doa recommendations, in 64 plots established to maintain the treatments during the experiment. fertilizer application was carried out as split applications according to doa recommendations. weeds were controlled in the weed-free plots by the application of pretilachlor + safener 300 ec according to the recommended rate of 1.6 l/ha, three days after establishment. bispyribac sodium was then applied at a recommended rate of 225 wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 158 g/ha 14 days after sowing. hand weeding was used to control excess weeds in weedfree maintained plots. table 1. treatments and their respective nutrient contents. treatment mineral nutrient (kg/ha) n (urea 46%) p (p2o5 16%) k (k2o 60%) full (100%) doa recommendation 103.5 3.9 30 half (50%) doa recommendation 51.75 1.95 15 doa: department of agriculture, sri lanka 2.2 data collection and statistical analysis the incidence and severity of nbls were recorded from the 50 days after sowing (das) with four days sampling intervals. four random quadrate (50cm x 50cm) samples were selected from each plot as described by lin et al. (1979). the number of infected plants in each quadrate sample was counted, and the disease incidence was calculated using the following formula. disease incidence = number of infected plants per quadrat total number of plants per quadrat the severity of the nbls was assessed by counting the number of infected leaves on four randomly selected plants inside the quadrate (severity 1). selected plants were tagged with plot number and treatment. the number of spots on the flag leaf of earlier tagged plants was counted. the severity of the disease in the rice plant was estimated using the counted values according to the predetermined scale in table 2 (severity 2). table 2. severity scale is based on the number of spots on the flag and its description. spot numbers/flag leaf severity scale description 0-30 1 less number of spots with individual occurrence. the diameter remains small. 31-60 2 medium number of spots with individual occurrence. the diameter remains small and medium. 61-90 3 high number of spots with individual and combined patches. the diameter remains small to medium size. > 90 4 very large number of spots with individual and combined patches. the diameter remains small to large. disease incidence and severity 1 data were tested for normality and heteroscedasticity and the data were transformed to square roots. non-linear wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 159 regression analysis was performed to assess disease incidences and severity 1 in both 2017/2018 maha and 2018 yala seasons to identify the pattern of the disease incidence and severity 1. and the effects of each experimental factor in each season were statistically interpreted using sas statistical software. all treatments and the seasons were considered as fixed factors and the replicate (block) and interactions with sub-plot factors were considered as random factors. all means were separated using tukey’s lsd at p=0.05. to understand the severity of the disease, the severity scale values identified for each block were averaged under each treatment combination. 3 results and discussion symptoms of nbls were observed in both the 2017/2018 maha and 2018 yala seasons. the disease incidence and severity were compared with the season and the three crop management factors. the cumulative effect of all treatments for the disease incidence was significantly higher (p<0.05) in the 2017/2018 maha season compared to that of the 2018 yala season. maha season recorded an 18% greater disease incidence than that of yala season (figure 1). fig. 1: cumulative effect of treatment on nbls disease incidence in 2017/2018 maha and 2018 yala seasons the observed difference between the incidences of the disease during the two seasons could be associated with the changes in optimal environmental conditions (elevated temperature, humidity and leaf wetness (pool and mckay 1916, shane and teng 1983) or disease development. in the maha season, the development of pathogen may be facilitated by comparatively higher wet conditions due to higher rainfall (data not shown) experienced than that of yala season. it is the most observed trend in the a b 0 0.2 0.4 0.6 0.8 2017/2018 maha 2018 yala d is e a se i n c id e n c e season wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 160 pattern of disease transmission (garcía-guzmán and dirzo 2004), and present results provide evidence parallel to the universal pattern. the disease incidence in the maha season progressively increased during the period of data collection and reached the maximum by the final sampling day (figure 2a). fig. 2. the cumulative effect of all treatments for nbls disease. (a) disease incidence in 2017/2018 maha season, (b) disease severity 1 (number of infected leaves per plant) in 2017/2018 maha season, (c) disease incidence in 2018 yala season and, (d) disease severity 1 (number of infected leaves per plant) in 2018 yala season. the favourable microclimate and nutrient uptake during the growing season may be conducive for the growth and spread of the disease during the vegetative and reproductive growing seasons without any hindrance (krupinsky et al. 2002). although disease incidence gradually increased, disease severity 1 gradually 0 1 2 3 4 5 6 7 8 9 0.0 0.2 0.4 0.6 0.8 1.0 2 0 1 7 /2 0 1 8 m a h a d is e a s e i n c id e n c e sampling date (a) 0 1 2 3 4 5 6 7 8 9 0 5 10 15 20 2 0 1 7 /2 0 1 8 m a h a d is e a s e s e v e ri ty sampling date (b) 0 1 2 3 4 5 6 7 8 9 0.0 0.2 0.4 0.6 0.8 1.0 2 0 1 8 y a la d is e a s e i n c id e n c e sampling date (c) 0 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 2 0 1 8 y a la d is e a s e s e v e ri ty sampling date (d) wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 161 decreased for the first four sampling days (62 das), which is also the end of the vegetative growth stage in the maha season (figure 2b). during that period, plants developed, and the leaf number increased. in proportion to the increasing leaf number, there was no increase in infected leaf number. from the fourth sampling date (62 das) to the sixth sampling date (72 das), there was a minor increasing trend of disease severity 1 and thereafter a gradual decrease (figure 2b). from the fourth sampling day to the sixth sampling date, i.e., from the later vegetative period to the initial reproductive stage, leaf growth is not shown while leaf senescence is accelerated. after the sixth sampling date, plants reached the end of the reproductive growth and it might have caused senescence of leaves. it may be the major reason for the decrease in the infected leaves per plant (vergara 1991). in the yala season, the disease incidence increased up to the sixth sampling date, thereafter, decreased towards the end of the season (figure 2c). the incidence of the disease in the yala season increased from a lower level in the vegetative stage of the crop to reach a peak at the initial stage of reproductive growth. the incidence remained the same or decreased during the latter part of the reproductive growth (figure 2c). similar results were observed by long et al. (2000) in a study on the incidence of blast disease of rice under different n levels. this pattern of disease progression may be attributed to an aspect of the development of resistance to the pathogen in the plant with maturity. at the end of the vegetative period, the plants become more resistant to certain pathogen due to the physiological maturity (bastiaans 1993), which alter the chemical composition of cells and permeability of cell walls. the disease incidences were not increased beyond this point as the death of the diseased leaves occur with the formation of new healthy leaves (long et al. 2000). the present study also observed that the severity of the disease decreases with the growth of the plant. disease severity 1 in the yala season also showed the same trend of disease incidence and the sixth sampling date reported higher disease severity (figure 2d). disease incidence in the yala season and disease severity 1 in both maha and yala was reported maximum values on the sixth sampling date. thus, the effects of experimental factors on the incidence of the disease and severity were assessed on the sixth sampling date (figure 2c). fertilizer, weed, seed rate and their interactions during the 2017/2018 maha season did not have a significant (p<0.05) impact on disease incidence. it was reported that the severity 1 of the disease during the 2017/2018 maha season was significantly different (p<0.05) with fertilizer and weed interactions. however, during the 2018 yala season, there was a significant difference (p<0.05) in the incidence of diseases with weedy and weed-free conditions. fertilizer, weed and seed rate interaction significantly (p<0.05) influence the disease severity 1 in the yala season. similar to the incidence of the 2018 yala disease, the severity 1 of the disease varied significantly with weedy and non-weedy conditions (table 3). wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 162 table 3: associated probability values for the effect of nitrogen fertilizer, weed control and seed rate on the disease incidence and disease severity 1 in 2017/2018 maha season and 2018 yala season. treatment disease incidence disease severity 1 2017/2018 maha season 2018 yala season 2017/2018 maha season 2018 yala season fertilizer (f) 0.33 0.09 0.77 0.86 weed (w) 0.78 <0.01 0.36 0.04 seed rate (sr) 0.70 0.89 0.92 0.57 f*w 0.95 0.43 0.03 0.07 f*sr 0.92 0.48 0.70 0.1 w*sr 0.90 0.21 0.57 0.06 f*w*sr 0.97 0.58 0.72 0.04 weed management and mineral n fertilizer dosage interaction using anova mixed model was significant (p<0.05) for the disease severity 1 in maha season while all factors did not significantly (p>0.05) affect disease incidence in maha season. a higher disease severity (3.3 leaves/ plant) was observed in plants grown in weedy conditions and supplied with the full recommendation of mineral n fertilizer than that of the treatment with half the dosage of n fertilizer in weedy condition (figure 3). fig. 3. nbls disease severity in 2017/2018 maha season with the rate of nitrogen fertilizer and control of weeds significantly lower disease severity (2.06 leaves/ plant) was reported with the weedfree condition with the full recommendation of mineral n fertilizer. effects of other factors and interactions were insignificant (p>0.05) (table 3). generally, the presence of high n in plants with a high n supply can increase the susceptibility to diseases (makheti mutebi et al. 2021). the invasion and growth of fungi can be facilitated by the weakening of the mechanical resistance induced by cell wall materials due to high n applications (talukder et al. 2005). since some weeds act as a host for diseases, weedy conditions may negatively affect the crop (table 4). more sedges were a b 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 weed weed free d is e a se i n c id e n c e i n 2 0 1 8 y a la s e a so n presence of weeds wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 163 reported than grasses, broad leaves and aquatic weeds in both seasons. as sedges, cyperus rotundus (kaladuru) (more abundance), echinochloa crus-galli (bajiri), and fimbristylis dichotoma (kudametta) (low abundance) were reported (data not shown) with this experiment. biswas (2006), also reported, sedges are the host plant for the narrow brown leaf spot. complete control of weeds under the full fertilizer dose can be used as a remedy to control the severity of the disease in the maha season (groth et al. 1991). table 4: mean disease incidence, severity 1 and severity scale values of treatment combinations in 2017/2018 maha season and 2018 yala season on sixth sampling date of rice crop. doa: department of agriculture; absence of shared superscripts within columns indicate significant difference (anova, p<0.05). disease incidence was significantly (p<0.05) higher when rice was grown under weedy conditions (69%) than weed-free conditions (58%) in the yala season (figure 4), while the impacts of other factors were constant. islam et al. (2002), observed similar results, where there was a significant (p<0.05) effect from weeds to increase diseases in plants. the same study explained the reduction of the disease when the source of the pathogen was reduced in the field through the removal of weeds. moreover, the presence of weeds increased the relative humidity and decrease the temperature surrounding the crop creating a favourable microenvironment for the treatments disease incidence severity 1 severity scale value seed rate n fertilizer dosage as % doa presence of weed 2017/2018 maha season 2018 yala season 2017/2018 maha season 2018 yala season 2017/2018 maha season 2018 yala season 100 kg/ha 100% weedy 0.81a 0.63abcd 3.73ab 2.83a 2 1 weed free 0.86a 0.51d 2.38b 1.39b 1 1 50% weedy 0.82a 0.78a 2.94ab 2.87a 2 1 weed free 0.84a 0.62abcd 2.87ab 1.90ab 2 1 125 kg/ha 100% weedy 0.84a 0.71abc 2.93ab 2.24ab 1 1 weed free 0.85a 0.54cd 2.79ab 1.72ab 2 1 50% weedy 0.88a 0.78ab 3.00ab 2.90a 2 1 weed free 0.85a 0.51d 3.16ab 2.05ab 2 1 150 kg/ha 100% weedy 0.87a 0.66abcd 3.21ab 2.69ab 2 1 weed free 0.81a 0.59bcd 2.56ab 2.04ab 1 1 50% weedy 0.91a 0.61abcd 2.81ab 2.40ab 2 1 weed free 0.89a 0.62abcd 3.82a 3.01a 2 1 175 kg/ha 100% weedy 0.87a 0.60abcd 3.31ab 2.00ab 2 1 weed free 0.85a 0.63abcd 2.69ab 2.16ab 1 1 50% weedy 0.92a 0.77ab 2.75ab 2.6ab 2 1 weed free 0.90a 0.59bcd 2.77ab 1.8ab 1 1 cv % 7.78 10.56 15.09 22.73 wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 164 pathogen (sall 1981) making the plants grown among weeds more susceptible to the disease. this might be the reason for the recorded higher disease incidence in the yala season. also, weed cause to increase disease susceptibility of the crop as a result of competition exerts by weed for the resources such as nutrients water, etc. in this study, it was stated that this factor also contributed to the increase in the disease. fig. 4. nbls disease incidence in 2018 yala season with control of weeds. during the yala season, the incidence of disease in the plot with the seed rate of 125 kg/ha, 50% fertilizer and weedy were higher than the disease incidence in the weedfree plot (table 4). half of the fertilizer application rate reduces the vigour of the plant, thereby increasing the susceptibility of the plant to disease (huber et al. 2012). disease severity 1 in the yala season with 100 kg/ha rate, 100% fertilizer dosage and weedy condition was reported significantly (p<0.05) higher value than disease severity in non-weedy conditions with same seed rate and same nutrient management conditions. the 100kg/ha seed rate, 100% fertilizer and weed-free condition reported the lowest disease severity in both yala and maha seasons (table 4). it also interprets that weed control is important under all seed rate and fertilizer levels to control the disease development (groth et al. 1991). in addition to identifying the incidence of the disease and the number of infected leaves per plant, the severity of the disease was determined by calculating the number of spots on a flag leaf with the scale values (table 4). when the weedy conditions prevailed in the field in the maha season, the disease severity scale value of 2 occurred at the highest frequency. groth et al. (1991) also reported that weedy conditions caused higher disease severity than weed-free conditions at any seed rate and any fertilizer management practices. 4 conclusions according to the findings, the incidence of the nbls disease was higher during the maha season than that of the yala season. the disease incidence and severity a b ab ab 0 0.5 1 1.5 2 2.5 3 3.5 4 weed weed free d is e a se s e v e ri ty i n 2 0 1 7 /2 0 1 8 m a h a s e a so n presence of weeds fertilizer half fertilizer wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 165 increased with the application of high mineral n fertilizers with the presence of weeds in the maha season, while the disease incidence increased with weedy conditions only in the yala season. disease severity in the yala season significantly changed with the seed rate, weed and fertilizer dosage interaction. the seed rate had no significant effect on the disease incidence in the two seasons. this study concludes that controlling weed significantly contributes to the reduction of nbls disease. acknowledgements the research project titled ‘designing sustainable cropping systems for the dry zone’ funded by the university research grant – rjt/rp&hdc/2016/agri/r/01 is highly acknowledged. s.n. nugaliyadda and k.m. jayasundara are greatly appreciated for their enormous support during data collection. comments from two anonymous reviewers are acknowledged. references abeysekera ask. 2001. management of echinochloa spp. in rice in sri lanka. in fao workshop on echinochloa spp. control, beijing, china. 3: 13. ahmed s, salim m, chauhan bs. 2014. effect of weed management and seed rate on crop growth under direct dry seeded rice systems in bangladesh. plos one. 9. e101919. 10.1371/journal.pone.0101919. doi:10.1371/journal.pone.0101919. alam sm. 2003. weeds and their ill effects on main crops. [accessed on 25.05.2019]. available at http://dawn.com. bastiaans l. 1993. understanding yield reduction in rice due to leaf blast. phd thesis. wageningen agricultural university, wageningen: netherlands. biswas a. 2006. narrow brown leaf spot disease of rice: a review. journal of mycopathological research 44 (1): 113-115. chauhan bs, singh rg, mahajan g. 2012. ecology and management of weeds under conservation agriculture: a review. crop protection 38: 57–65. doi: 10.1016/j.cropro.2012.03.010. economic and social statistics of sri lanka 2018. central bank of sri lanka, 2018. garcía-guzmán g, dirzo r. 2004. incidence of leaf pathogens in the canopy of a mexican tropical wet forest. plant ecology 172 (1): 41-50. doi:10.1023/b: vege.0000026034.24664.73 groth de, rush mc, hollier ca. 1991. rice diseases and disorders in louisiana. louisiana state university, agricultural center, louisiana agricultural experiment station, bulletin no. 828. baton rouge, la, usa. guillermo da, pedersen p, hartzler rg. 2009. soybean seeding rate effects on weed management. weed technology 23: 17–22. doi:10.2307/40587035 hakim ma, juraimi as, musa mh, ismail mr, rahman mm, selamat a. 2013. impacts of weed competition on plant characters and the critical period of weed control in rice under saline environment. australian journal of crop science 7 (8): 1141-1151. huber d, römheld v, weinmann m. 2012. relationship between nutrition, plant diseases and pests. in: marschner's mineral nutrition of higher plants (pp. 283-298). academic press. doi:10.1016/b978-012-384905-2.00010-8 islam mr, ali ma, islam ms, golam afm. 2002. effect of nutrients and weeding on the incidence of mung bean mosaic. plant pathology journal 1: 48-50. https://doi.org/10.1371/journal.pone.0101919 wickramasinghe et al. narrow brown leaf spot in rice cultivation in sri lanka ruhuna journal of science vol 12 (2): 155 -166, december 2021 166 knowledge bank, international rice research institute. [accessed on 25.05.2019]. available at http://www.knowledgebank.irri.org/training/fact-sheets/pestmanagement/diseases/item/narrowbrown-spot. krupinsky jm, bailey kl, mcmullen mp, gossen bd, turkington tk. 2002. managing plant disease risk in diversified cropping systems. agronomy journal 94 (2): 198-209. doi:10.2134/agronj2002.1980 lin cs, poushinsky g, mauer m. 1979. an examination of five sampling methods under random and clustered disease distributions using simulation. canadian journal of plant science 59 (1): 121-130. doi:10.4141/cjps79-017 long dh, lee fn, te beest do. 2000. effect of nitrogen fertilization on disease progress of rice blast on susceptible and resistant cultivars. plant disease 84: 403-409. doi:10.1094/pdis.2000.84.4.403 makheti mutebi c, atieno ondede d. 2021. effect of nitrogen nutrition on the intensity of cercospora leaf spot of mulberry. international journal of horticultural science and technology 8 (4): 335-342. doi:10.22059/ijhst.2021.309763.393 mitchell ce, reich pb, tilman d, groth jv. 2003. effects of elevated co2, nitrogen deposition, and decreased species diversity on foliar fungal plant disease. global change biology 9: 438–51. doi:10.1046/j.1365-2486.2003. 00602.x panique e, kelling ka, schulte ee, hero de, stevenson wr, james rv. 1997. potassium rate and source effects on potato yield quality, and disease interaction. american journal of potato research 74: 379–98. doi:10.1007/bf02852777 pool vw, mckay mb. 1916. relation of stomatal movement to infection by cercospora beticola. journal of agricultural research 5: 1011–1038. rabbani n, bajwa r, javaid a. 2011. interference of five problematic weed species with rice growth and yield. african journal of biotechnology 10: 1854–1862. reganold jp, wachter jm. 2016. organic agriculture in the twenty-first century. nat plants 2: 15221. doi:10.1038/nplants.2015.221 rice cultivation, department of agriculture, sri lanka. [accessed on 25.05.2019]. available at https://www.doa.gov.lk/rrdi /index.php/en/crop/42-crop-rice-cultivation. sall ma, 1981. relationship between plant disease and weed management. [accessed on 25.05.2019]. available at https://ucanr.edu/repository/fileaccess.cfm? article=161344&p=llfvja. schreiber f, scherner a, andres a, concenço g, goulart f. 2018. competitive ability of rice cultivars in the era of weeds resistance. plant competition in cropping systems. doi: 10.5772/intechopen.78342. shane ww, teng ps. 1983. cercospora beticola infection prediction model. sugar beet research and extension report 15: 129-138. simanjuntak fas, safni i, bakti d. 2020. distribution of narrow brown leaf spot disease of rice (cercospora oryzae miyake) in north sumatra. ecology and environmental science 454 (1): 012163. doi:10.1088/1755-1315/454/1/012163 talukder zi, mcdonald aj, price ah. 2005. loci controlling partial resistance to rice blast do not show marked qtl x environment interaction when plant nitrogen status alters disease severity. the new phytologist 168 (2): 455-464. doi:10.1111/j.1469-8137.2005. 01507.x veresogloua sd, bartoab ek, menexesc g, rilliga mc. 2013. fertilization affects severity of disease caused by fungal plant pathogens. plant pathology journal 62: 961–969. doi:10.1111/ppa.12014 vergara bs. 1991. rice plant growth and development. in: rice. springer, boston: ma 13-22. doi:10.1007/978-1-4899-3754-4_2 weerakoon sr, somaratne s, ekanayaka emsi, munasighe s. 2018. natural resistance of sri lankan rice (oryza sativa l.) varieties to broad-spectrum herbicides (glyphosate and glufosinate). 10.5772/intechopen.76991. doi:10.5772/intechopen.76991 wisler cg. 2009. interactions between weeds and cultivated plants as related to management of plant pathogens. weed science 53: 914-917. doi:10.1614/ws-04-051r.1 https://dx.doi.org/10.22059/ijhst.2021.309763.393 sv-lncs ruhuna journal of science vol 6: 50-62, december 2015 issn: 1800-279x  faculty of science university of ruhuna  faculty of science, university of ruhuna 50 printed in sri lanka effects of improved fish production technology on the output of fish farmers in ilorin, kwara state, nigeria j.a. akangbe, g.e. ajiboye, and s.e. komolafe* department of agricultural extension and rural development, university of ilorin, nigeria *correspondence: kemmas04@yahoo.com received: 30 th september 2015, revised: 18 th december 2015, accepted: 21 st december 2015 abstract. the study examined the effects of improved fish production technology on the output of fish farmers in ilorin kwara state, nigeria. a total of 125 respondents were selected. primary data was collected with the use of structured questionnaire. both descriptive and inferential statistics were employed for the study. mean age of respondents was 39.8 years. also majority were male (83.2%), had tertiary education (84.0%) and kept catfish (90.4%). the mean income earned was n 1,012,320.00. commonly used/ adopted improved fish technologies were floating feeds (84.8%), standard feeding regimes (84.0%), improved breeds of fingerlings (80.0%) and provision of inlet and outlet devices in pond (78.4). farmers’ perceptions were that the use of improved fish production technology saves time (mean=1.34), increases profit/improved income (mean=1.12) and conserves fish farmers’ energy (mean=1.08). this study found average increase in output: input ratio as 0.77 kg of harvest per fish fingerlings stocked due to improved fish technology when compared to that before adoption of improved technology. most indicated constraints faced by fish farmers were inadequate capital (88.0%), high cost of feed (79.2%) and high cost of fingerlings (71.2%). profit increase/improved income of catfish farmers via higher yield/harvest due to adoption of new technology was found to be positively correlated with age (p<0.01), gender (p<0.05), and experience (p<0.05). the study concluded that the use of fish improved technology had positively influenced harvest of catfish farmers in the study area. the study recommend the need for training, workshops and seminars for catfish farmers on how they could have easy access to land, feeds at affordable rate and sources of fund at minimal interest rate. keywords. catfish farmers, fish production, improved technology, socio-economics. akangbe et al effects of improved fish production technology ruhuna journal of science 51 vol 6: 50-62, december 2015 1 introduction aquaculture is the breeding and rearing of fish, shellfish, or plants in ponds, or any enclosure for direct harvest of the product, which is an area of activity growing rapidly (fao, 2004). fish is acclaimed to be the principal source of animal protein for over one billion people globally as it is the cheapest source of animal protein, providing many important nutritional and health benefits (fao, 2006). in nigeria, fish demand as estimated by ruma (2008) was 2.1 million metric tons at 11.5 kg per capita consumption. the poor performance of the fishery sub sector in nigeria is most clearly evidenced by low standard of living of the small scale rural fish farmers (fao, 2006). to revamp this sub sector, government of nigeria has introduced and implemented numerous policies and programmes aimed at empowering the small scale fish farmers to get out of the ‘poverty trap’. this include, dissemination of improved catfish production technology such as adequate pond construction, water management, adequate stocking rate, use of nutritious and floating feed, and improved fish feed to the farmers (ike et al., 2009). ekong (2003) defined technology as ways of applying scientific and organized knowledge into practical task. in view of this, lakra and ayyappan (2003) had noted that increased application of advance technological tools could certainly revolutionise fish farming. dissemination of proven technologies is a communication process of sharing and distributing information to a target audience to bridge the gap in knowledge and bring about changes in attitude and skill of the end users. despite the efforts of government in research and extension services, there is still a deficit in the supply and demand for fish (fdf, 2005). other problems facing domestic fish production in nigeria are, innovation adoption, inadequate research and extension, high cost of fisheries inputs, preservation and storage technologies, fish seed, lack of credit and insurance cover for fisheries enterprises (olaoye 2010). despite the fact that fish farming account for the highest percentage of the nigerian’s annual fish production output (fdf, 2005), fish workers are often among the poorest people and they generally operate small scale fishing units using traditional fishing practices. however, new technology and environmental requirements favour large scale capital intensive operation at the expense of traditional and small scale commercial fishing (delgado et al, 2003). thus, it is of paramount importance to study the effect of improved production technologies on the yields and income of fish farmers. the main objective of the study is to examine the effects of improved fish production technology on the output of fish farmers in ilorin kwara state, nigeria. the specific objectives of the study are to: i. describe the socio-economic characteristics of fish farmers in the study area, akangbe et al effects of improved fish production technology ruhuna journal of science 52 vol 6: 50-62, december 2015 ii. identify sources of information and types of fish production technology used by fish farmers, iii. examine fish farmers perception and constraints faced to improve fish production technology. the study aimed at testing the following null hypothesis (h0): there is no significant relationship between socio-economic characteristics of fish farmers and the effects of improved fish production technology. 2 materials and methods 2.1 study area the study was carried out in ilorin, kwara state nigeria. ilorin is the capital of kwara state. the state is geographically located between latitude 7° 20' and 11° 05' north of the equator longitude 2° 05' and 6° 45' east of the prime meridian. the state is bounded in the north by sokoto and niger states, and the federal capital territory, and in the south by oyo, osun, ekiti and edo states. the western boundary is republic of benin, while the eastern boundary consists of plateau and benue states. kwara state has a population of 1.57 million and a land area of about 32,500 km 2 with three main ethnic groups; yoruba, nupe and baruba. the climate is intermediate, varying between the extremes of dryness, coolness and hotness. the mean monthly rainfall ranges between 50 mm during the wettest months and 24 mm during the driest period. the driest months are from january to march, while the rains last from may to september with occasional drizzles in october. the minimum average temperature throughout the state ranges between 21°c while maximum average temperature ranges approximately between 30°c and 35°c (kwadp 2000). 2.2 sampling procedure and sample size the population for this study comprised of all the 163 member of the cat fish farmers association of nigeria (caffan) and 247 member of the association of fish farmers association of nigeria (affan) in ilorin kwara state, nigeria. forty percent of the population from the two fish farmer associations were randomly selected. a total of 165 respondents were selected for the study. akangbe et al effects of improved fish production technology ruhuna journal of science 53 vol 6: 50-62, december 2015 2.3 instrument for data collection the data collected for this study were obtained from primary and secondary sources. primary data was collected from the field survey through the administration of structured questionnaire which was used to solicit information from the respondents on issues related to objectives of the study. the validity of the questionnaire was ensured by lecturers of the department of agricultural extension and rural development, university of ilorin. secondary data on the other hand were collected from relevant literature, textbooks etc. 2.4 data analysis of the 165 questionnaires administered, 130 were retrieved and 5 were rejected for incompleteness. a total of 125 questionnaires were therefore available for analysis. data obtained from the field were subjected to descriptive and inferential statistics. the descriptive statistics tools used were frequency counts, percentage, mean score and standard deviation. pearson product moment correlation analysis was used to test the null hypothesis of the study. 3 results and discussion 3.1 socio-economic characteristics of respondents results presented in table 1 illustrate the socio-economic characteristics of fish farmers in ilorin kwara state, nigeria. the results showed that the mean age of respondents was 39.8 years. the implication is that the fish farmers in the study area are fairly young, and are expected to be active in fisheries activities, and constitute potential labour force for the fisheries enterprise in the study area. the mean age of farmers in nigeria is usually between 45-48 years (ogunwale 2000; ezedinma and otti 2001). the reason for this particular age composition could be attributed to the fact that aquaculture is relatively new in the country. however, the result contradicts the findings of ofuoku et al. (2008) that very few young people are involved in fish farming from a study in dalta state of nigeria. as revealed by table 1, the majority (83.2%) of the farmers were male while only16.8% was female. this implies that fish farming business is male dominated in the study area. this may be due to the high degree of human energy and physical exertion associated with farming activities, as it is evident from significantly low frequency of involvement of women in fish farming. table 1. socio-economic characteristics of fish farmers in ilorin (source: field survey, 2015). akangbe et al effects of improved fish production technology ruhuna journal of science 54 vol 6: 50-62, december 2015 variables frequency a percentage b mean age (years) less than 30 38 30.4 39.8 31 – 40 43 34.4 41 – 50 18 14.4 51 – 60 12 9.6 61 and above 14 11.2 gender male 104 83.2 female 21 16.8 marital status single 45 36.0 married 78 62.4 widow/ widower 2 1.6 divorced 0 0.0 educational status no formal education 1 0.8 adult education 3 2.4 primary education 4 3.2 secondary education 12 9.6 tertiary education 105 84.0 fish farming experience (years) 1 – 5 79 63.2 6.5 6 – 10 22 17.6 11 – 15 16 12.8 16 – 20 6 4.8 21 and above 2 1.6 fish farm size (no. of fingerlings) less than 1000 3 2.4 4,008 1000 – 5,000 108 86.4 5,001 – 10,000 8 6.4 10,001 and above 6 4.8 income per annum (naira) 200,000 and below 23 18.4 1,012,320 201,000 – 400,000 14 11.2 401,000 – 600,000 16 12.8 601,000 – 800,000 14 11.2 801,000 – 100,000 18 14.4 1,000,000 – 2,000,000 32 25.6 2,001,000 and above 8 6.4 extension visit (per annum) no visit 73 58.4 1.75 1 – 5 40 32.0 6 – 10 10 8.0 11 – 15 2 1.6 number of fish training/ seminar attended no attendance 29 23.2 3.0 1 – 5 17 57.6 6 – 10 22 17.6 11 – 15 2 1.6 type of fish kept catfish 113 90.4 cat and tilapia 12 9.6 a total number of fish farmers under survey is 125 for all variables. akangbe et al effects of improved fish production technology ruhuna journal of science 55 vol 6: 50-62, december 2015 similar results of higher percentage of male involvement in fish farming were reported by abiona et al. (2012). this is in consistence with the result of chioma and adebayo (2012) that although women in ilorin have shown willingness to participate in fish farming, but are yet to involve in commercial fish farming and invest in it due to lack of funds. as regards the marital status of respondents, the majority (62.4%) was married. this shows that most of the fish farmers are men and women with household responsibilities. these responsibilities are likely to make them willing to seek innovations so as to increase their income earning capacity and improve their standard of living (raufu et al., 2009). this ascertain was further confirmed by the report of oladoja et al. (2008) who assert that marriage confer some level of responsibility and commitment on an individual who is married. information furnished in table 1 also revealed that the majority (84.0%) had attended tertiary education. this implies that fish farmers in the study area are educated and could be trusted to adopt any innovation that could enhance fish farming practices. okunlola (2010) stated that educational level is one of the factors that influence adoption of new technology by farmers. results of this study are consistent with previous studies by adefalu et al. (2013) and ofuoku et al. (2008) who reported a high percentage of fish farmers having tertiary education in kwara state and delta state respectively. the majority (63.2%) of respondents, being young fish farmers, had between 1–5 years in fish farming experience. the mean fish farming experience of respondents was 6.49 years. the mean score of fish farm size was 4,008 fingerlings. the mean annual income earned by respondents was estimated at naira 1,012,320. the annual mean number of extension visits to fish farmers in the study area was 1.75, which was very low considering their importance for frequent update of improved technologies in enhancing fish production in the study area. low involvement of extension workers could be as a result of the research institutes and extension organizations inadequate planning in nigeria (oladele, et al., 2006). fasakin (2008) also stated that poor agricultural extension services are a serious constraint to fish production in nigeria. of the 125 respondents, mean number of participants in training/seminars for fish farming was 2.99. majority of fish farmers (90.4%) kept catfish while others stocked both catfish and tilapia. this implies that catfish farming was predominant in the study area. in his studies in cross river state, ideba et al. (2013) too reported that the majority of fish farmers stocked clarias (catfish). akangbe et al effects of improved fish production technology ruhuna journal of science 56 vol 6: 50-62, december 2015 3.2 types of improved fish technology used by respondents table 2 presents information on the types of improved fish farming technology used by respondents, ranked in ascending order of the percentage of respondents using such technology. the results revealed that 84.8% used floating feeds (84.8%) was the top rank; followed by standard feeding regimes (84.0%), improved breeds of fingerlings (80.0%), provision of inlet and outlet devices in pond and frequent change of water (78.4%), etc. the lowest ranks were soil testing before site selection (40.8%), use of water testing kits (34.4%) and construction of modern fishing gears (32.8%). table 2. types of improved fish technology used by respondents improved fish technology used frequency* percentage percentage rank floating feeds 106 84.8 1 standard feeding regimes 105 84.0 2 improved breeds of fingerlings 100 80.0 3 provision of inlet and outlet devices in pond 98 78.4 4 frequent change of water 98 78.4 4 regular sampling/sorting of fish 97 77.6 6 daily sanitation and record-keeping practices 87 69.6 7 prevention and control of fish diseases 86 68.8 8 optimum stocking rate 83 66.4 9 improved techniques in pond construction and maintenance 74 59.2 10 fertilization and liming of fish pond 70 56.0 11 techniques of hatchery and fingerling production 64 51.2 12 fish preservation and storage techniques 60 48.0 13 techniques of improving water quality in fish culture 57 45.6 14 integrated fish farming for increased fish production 57 45.6 14 aerated containers for transporting fingerlings to reduce stress and mortality 55 44.0 16 soil testing before site selection 51 40.8 17 water testing kits for oxygen, acidity and fertility 43 34.4 18 construction of modern fishing gears 41 32.8 19 * multiple responses 3.3 sources of information on improved fish production technology of the major source of information on improved fish production technology for fish farmers (table 3) was information obtained from their fellow fish farmers (92.0%). this finding is consistent with ofuoku et al. (2008) and okunlola et al. (2011) that the most popular sources of information available to fish farmers in nigeria was interaction with other fish farmers. also, the second important source of information (54.4%) was from cooperative meetings. akangbe et al effects of improved fish production technology ruhuna journal of science 57 vol 6: 50-62, december 2015 table 3. sources of information on improved fish production technology information sources frequency* percentage agricultural extension agent 68 37.6 internet 65 52.0 television 35 28.0 radio 27 21.6 print media 40 32.0 cooperative meetings 47 54.4 private consultants 58 46.4 input dealers 40 32.0 fellow fish farmers 115 92.0 * multiple responses 3.4 perception to use and effects of improved fish production technology perception of farmers on improved fish production technology (table 4) determined by likert scale revealed that statement that use of improved fisheries technology saves time had the highest mean score of 1.34. table 4. perception to use and effects of improved fish production technology perception statements a mean score std. dev. mean rank using improved fisheries technologies saves time 1.34* 0.975 1 fish farmers’ uses of improved technologies increases profit/improved income via higher yield/harvest 1.12* 0.912 2 improved fisheries technologies conserves fish farmers energy – human, material & finance 1.08* 0.858 3 the require regular contact with extension workers 1.02* 1.020 4 improved technologies are only for the educated 0.97 1.150 5 they are not culturally suitable 0.82 1.001 6 have high labour requirement 0.72 1.052 7 large family size dis-encourages fish farmers’ use of improved technologies. 0.72 0.955 7 improve technologies are too complex for my linking 0.63 1.074 9 lazy fish farmers use improved technologies -0.15 1.115 10 using improved fisheries technologies lead to high product losses. -0.64 1.328 11 a likert scale: strongly agree= 2, agree= 1, undecided= 0, disagree= -1, strongly disagree= -2. 1-11 implies highest to lowest rank; * high level of perception decision rule: score mean 0.99 and below indicates low level of perception; score mean between 1.00 and 2.00 indicates high level of perception. akangbe et al effects of improved fish production technology ruhuna journal of science 58 vol 6: 50-62, december 2015 the second highest perception that use of improved technology increases profit/improved income via higher yield/harvest (mean score: 1.12), that the improved fisheries technology conserve fish farmers labour, material and finance (mean score: 1.08), and that the requirement for regular contact with extension workers (mean score: 1.02). 3.5 effects of improved fish production technology on farmers output the difference in stocking (input) and output (harvest) before and after the use of improved fish technology (table 5) shows that the mean number of fingerlings stocked before improved fish technology was 835.6 and the mean output (harvest) without improved technology was 692.5 kg. hence, the output: input ratio before application of improved technology was 0.83 kg per fish fingerling. table 5. fish production stock (number of fish) and output (kg weight) with and without improved technology (source: field survey, 2015) stock *frequency (%) mean stock input (number of fish) output (kg) *frequency (%) mean output (kg) before improved technologies 835.60 692.53 not applicable 61 (48.8) 0 (0.00) 1 – 1000 35 (28.0) 40 (32.0) 1001 2000 17 (13.6) 16 (12.8) > 2000 12 (9.6) 8 (6.4) total 125 (100.0) 64 (51.2) after improved technologies 3644.72 5828.00 1 – 2,000 61(48.8) 36(28.8) 2,001 – 4,000 42 (33.6) 35(28.0) 4,001 – 8,000 12(9.6) 37(29.6) 8,001 – 10,000 2(1.6) 4(3.2) 10,001 – 12,000 2(1.6) 2(1.6) > 12,000 6(4.8) 11(8.8) total 125(100.0) 125(100.0) t-value (paired sample test) -6.138 -6.642 df 124 124 significance level (2-tailed) <0.0001 <0.0001 * of a total number of 125 farmers akangbe et al effects of improved fish production technology ruhuna journal of science 59 vol 6: 50-62, december 2015 after the adoption of improved technology, the mean stocking was 3644.7 fingerlings and the mean output (harvest) 5828 kg. here, the output: input ratio was 1.60 kg per fish fingerling. this indicates that there is an appreciable increase in the output: input ratio (nearly 2-fold increase; 0.77 kg per fish fingerling after adoption of improved fish technology. the implication is that the use and adoption of improved technology has positive influence on production output of fish farmers in the study area. this result is consistent with the finding of ashaolu et al. (2006) who observed that fish farming is profitable. 3.6 constraints faced by fish farmers in using improved fish production technology of the perceived constraints to adopt improved fish production technology (table 6), inadequate capital was the most severe (88.0%), which confirmed the findings of adefalu et al. (2013) in kwara state and of issa et al. (2014) in kaduna state. table 6. constraints faced by fish farmers in using improved fish production technology (source: field survey, 2015) perceived constraints frequency percentage percentage rank inadequate capital 110 88.0 1 high cost of feed 99 79.2 2 high cost of fingerlings 89 71.2 3 poor marketing structure 78 62.4 4 high cost of land 75 60.0 5 lack of technical skill 73 58.4 6 poor extension service 71 56.8 7 water scarcity 71 56.8 7 poor managerial skill 64 51.2 9 poor transport facility 61 48.8 10 disease outbreak 61 48.8 10 lack of commercial hatchery 58 46.4 12 illiteracy 46 36.8 13 about 79.2% respondents felt that high cost of feed was a major constraint. nwachukwu and onuegbu (2007) had noted that feeding was always a problem because farmers were not always able to afford the cost of the feed. this could be due to sudden change in the price of feed as a result of inflation. akangbe et al effects of improved fish production technology ruhuna journal of science 60 vol 6: 50-62, december 2015 due to the low quality of locally produced fish feed, farmers are often dependent on imported feed, which are more expensive and may be scarce because of import policies. olaoye et al. (2013) also found similar results among catfish farmers in oyo state, nigeria. high cost of fingerlings and poor market structure are also other major constraints faced by catfish farmers in the study area (table 6). 3.7 pearson correlation analysis of socio-economic characteristics of respondents and effects of improved fish technology on fish production increase of profit/income via higher yield/harvest due to improved technology had significant positive correlation with age (r=0.230; p<0.01), gender (r=0.214; p<0.05), experience (r= 0.194; p<0.05), and income (r=0.228; p<0.05) (table 7). these are in agreement with findings of langy and mekura (2005) who reported that older farmers have higher accumulated capital, more contacts with extension workers, better preferred by credit institutions and larger family size, all of which may make them more prepared to adopt technology more than younger ones, but not in agreement with ume, et al. (2009), who suggested that older farmers are less amendable to change and hence reluctant to adopt new technologies. according to langy and mekura (2005), experience of farmers is generally relevant to adopt new technology through interaction with their neighbours and the outside world. table 7. results of correlation between socio-economic characteristics of respondents and effects of improved fish technology on fish production variables r-value significant level. (2-tailed) age 0.230** 0.010 gender 0.214* 0.017 marital status 0.159 0.077 educational status 0.169 0.059 fish farming experience 0.194* 0.030 fish size 0.063 0.488 income 0.228* 0.011 **correlation is significant at the 0.01 level (2-tailed) *correlation is significant at the 0.05 level (2-tailed) 4 conclusion based on findings in the study, it was concluded that use of fish improved technology had positively influenced fish production output of fish farmers in the study area. secondly, fish farmers had high perception that farmers’ uses/adoption of improved technology increased profit/ income via higher yield/harvest. thirdly, the major determinants of farmers output through improved technology were: age, marital status, educational status, fish farmers akangbe et al effects of improved fish production technology ruhuna journal of science 61 vol 6: 50-62, december 2015 experience, fish size, and income of respondents. however, inadequate capital and high cost of feeds were found to be the major factors affecting the output of fish farming in the study area. based on the findings of this study, following recommendations are proposed. associations of catfish farmers in the study area should focus its training, workshops and seminars more on how members could have easy access to land, feeds at affordable rate and sources of fund at minimal interest rate. the extension agent to farmer coverage needs to be improved upon, so that regular and prompt visits could be made to fish farmers in order to intensify the adoption of improved fish technology. government should encourage the fish farmers by providing for them credit facilities at minimal interest rate, subsidized costs of improved fish feeds and fingerlings, and other improved fish production technology, thereby increasing their productivity. fish farming in the area is male dominated. females need to be encouraged to participate in fish farming in the area as a means of augmenting their income and improving their standard of living. references abiona bg, fakoya eo, apantaku so, alegbeleye wo, ajayi mt, obasa so, arowolo k. 2012. assessment of farmer’s technologies on integrated fish farming and non – integrated fish farming in ogun state, nigeria. global journal of science frontier research agriculture & biology 12 (2): 1-8. adefalu ll, aderinoye-abdulwahab sa, bello og, olorunfemi od, oba sa. 2013. information needs of fish farmers in ilorin metropolis, kwara state nigeria. nigerian journal of agriculture, food and environment 9(2): 1-5. agbamu ju. 2006. essentials of agricultural communication in nigeria. malthouse ltd. pp. 65-73. ashaolu of, akinyemi aa, nzekwe lso. 2006. economic viability of homestead fish production in abeokuta metropolis of ogun state nigeria. asset series a 6(2): 209-220. chioma gn, adebayo z. 2012. constraints to fish farming among women in nigeria. advances in environmental biology 6(4): 1421-1425. delgado cl, wada n, rosegrant mw, meijer s, ahmed m. 2003. the future of fish: issues and trends to 2020. international food policy research institute, washington dc, usa, pp 19: 1-22. ezedinma ci, oti nn. 2001. socio-economic issues in the development of cassava processing technology in nigeria. journal of sustainable agriculture and the environment 3(1):120 126. ekong e.e. 2003. an introduction to rural sociology (2nd edition). uyo. nigeria dove educational publishers. pp 259-285. fao. 2004. the state of world fisheries and aquaculture. 2003. rome: fao fisheries. fao. 2006. fisheries management in the federal republic of nigeria. www.fao.org/fi/fcp/en/nga/ body, retrieved on april 13, 2015. fakoya eo. 2000. farmers use of sustainable land management practices in ondo state, nigeria. unpublished phd. thesis, department of agricultural extension and rural development, university of ibadan. pp 160. fasakin ea. 2008. fish as food: yesterday, today and forever. inaugural lecture series 48; the federal university of technology, akure. pp 6-7. fdf. 2005. report of presidential committee on fisheries and aquaculture development, volume one: consolidated report. september, 2005. 63pp. http://www.fao.org/fi/fcp/en/nga/ akangbe et al effects of improved fish production technology ruhuna journal of science 62 vol 6: 50-62, december 2015 ideba ee, otu wi, essien aa, iniobong eo, ekaette su. 2013. economic analysis of fish farming in calabar, cross river state, nigeria. greener journal of agricultural sciences 3(7): 542-549. ike se, onyishi gc, ukwuaba gc, oluason va. 2009. farm evaluation of profitability and farmer’s acceptability of bambara nut waste as major protein supplement in fish diet in heterobranchus and tilapia spp polyculture. proceedings of national farming system research network held at umudike. pp. 127-129. issa fo, abdulazeez mo, kezi dm, dare js, umar r. 2014. profitability analysis of small scale catfish farming in kaduna state nigeria, journal of agricultural extension and rural development 6(8): 267-273. kwadp. 2000. kwara state agricultural development project annual report, pp. 3-7. lakra ws, ayyappans af. 2003. recent advances on biotechnology applications to aquaculture. central institute of fisheries education, (indian council of agricultural research), versova, mumbai. langy a, mekura m. 2005. modelling agricultural technology adoption using the software stat. a paper presented at a training course organised by cimmutalp for its nars partners in southern africa on econometric application of modelling technologies. nwachukwu i, onuegbu r. 2007. adoption of aquaculture technology by fish farmers in imo state of nigeria. the journal of technology studies 33(1): 57-64. ofuoku an, emah gn, itedjere be. 2008. information utilization among rural fish farmers in central agricultural zone of delta state nigeria. world journal of agricultural science 4 (5): 558 – 564. oladoja m a, adedoyin se, adeokun oa. 2008. training needs of fisher folks on fishing technologies. journal of food agriculture and environment science and technology 6(1): 95-198. ogunwale b. 2000. communication channels for information dissemination on poverty alleviation among small scale farmers in oyo state. paper presented at aeson conference, ui ibadan, nigeria. oladele oi, sakagami j, toriyama k. 2006. research-extension-farmer linkage system in southwestern nigeria, journal of food agriculture and environment 4(1): 197-200. olaoye oj. 2010. dynamics of the adoption process of improved fisheries technologies in lagos and ogun states nigeria. ph. d thesis, federal university of agriculture, abeokuta, ogun state, nigeria. pp. 337. olaoye oj, ashley-dejo ss, fakoya eo, ikeweinwe nb, alegbeleye wo, ashaolu fo, adelaja oa. 2013. assessment of socio-economic analysis of fish farming in oyo state, nigeria. global journal of science frontier research agriculture and veterinary 13(9): 45-55. okunlola jo. 2010. factors influencing adoption of rubber based technologies among small holder farmers in delta state nigeria. journal of food agriculture and environment 8(2): 391-394. okunlola, jo, oludare ao, akinwalere bo. 2011. adoption of new technologies by fish farmers in akure, ondo state nigeria. journal of agricultural technology 7(6): 1539-1548. raufu mo, adepoju mo, salau as, adebiyi oa. 2009. determinants of yield performance in small scale fish farming in alimosho local government area of lagos state. international journal of agricultural economics and rural development 2(1): 9-14. ruma ya. 2008. fish production systems. a workshop held at enugu state university, fisheries project. ume si, uloh ve, okoronkwo mo. 2009. adoption of improved rice production technologies by farmers in anambra state. ebonyi technology and vocational education journal 3(1): 1 7. vol 12 (1): 14-25, june 2021 eissn: 2536-8400 © faculty of science http://doi.org/10.4038/rjs.v12i1.97 university of ruhuna sri lanka © faculty of science, university of ruhuna sri lanka 14 mass production of the nematode acrobeloides longiuterus using tribolium castaneum and artificial solid media *n. thiruchchelvan, g. thirukkumaran and g. mikunthan department of agricultural biology, faculty of agriculture, university of jaffna, ariviyal nagar, kilinochchi 44000, sri lanka *correspondence: thiruchchelvann@univ.jfn.ac.lk; orcid: https://orcid.org/0000-0003-3800-7104 received: 12th february, 2021, revised: 30th may, 2021, accepted: 11th june, 2020 abstract. free-living nematode acrobeloides longiuterus (rhabditida: cephalobidae) exhibits a potential to kill some insect pests. mass production of this species is a requirement for use it in pest management programs. tribolium castaneum has been used as a primary host for this nematode as an alternative for galleria mellonella. use of artificial media is another option for mass culturing and such recipes based on soy flour are available. production of a. longiuterus using cost effective method and easily available insect host is important in setting up of small-scale production unit. therefore, this study has the objectives of evaluating the production feasibility of a. longiuterus on t. castaneum larvae, pupae and adults as in vivo production method. further, feasibility of using different solid media such as soy flour, palmyra tuber flour, corn flour, black gram flour and dhal flour with other basic ingredients as in vitro conditions system was evaluated. results revealed that pupa of t. castaneum yielded the highest number of infective juveniles (36112 ijs/ pupa) compared to other life stages tested. in vitro production of a. longiuterus on soy flour and black gram flour media yielded 21530 and 16538 ijs/20g, respectively. pathogenicity against t. castaneum was shown up to 93% by the infective juveniles produced from the in vitro cultures. in conclusion, t. castaneum is an alternative insect that can be used as a host to produce the a. longiuterus. in addition, soy flour and black gram flour can be used as the sources for this nematode production without losing their entomopathogenicity. keywords: entomopathogenic nematodes, free-living nematode, insect host, flour media, red flour beetle. 1 introduction application of entomopathogenic nematodes (epns) in pest management is ecologically safe and has a quick response. epns have been formulated as biohttps://rjs.ruh.ac.lk/index.php/rjs/index https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v12i1.97 https://creativecommons.org/licenses/by-nc/4.0/ mailto:thiruchchelvann@univ.jfn.ac.lk https://orcid.org/0000-0003-3800-7104 n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 15 pesticides and are commercially available to be used in pest management programs (ehlers and shapiro-ilan 2005). free-living nematode acrobeloides longiuterus (rhabditida: cephalobidae) which is not a true member of entomopathogenic genera such as steinernema and heterorhabditis has shown a potential to kill economically important insect pests (thiruchchelvan et al. 2021). this nematode species causes mortality of the insect host within 48-72 h of application (thiruchchelvan et al. 2019, 2021), which is a quick knock down effect similar to the classical epns. therefore, a. longiuterus could be used as an additional insect management strategy in integrated pest management. this strategy is environmentally friendly as well as safe for non-target organisms. in order to use this nematode as a biological control agent, it is necessary to produce it in bulk at a reasonable cost. this can be achieved through in vivo and in vitro production using insect hosts and artificial solid media, respectively. glaser (1940) introduced a method of large-scale in vitro production of epns using solid media. then, friedman (1990) took it into the next level of in vitro liquid fermentation production method from the three-dimensional in vitro production by bedding (1981, 1984). meanwhile, farmers are benefited from both in vivo and in vitro production of nematodes if the production process utilizes substrates generated from their farms (ehlers and shapiro-ilan 2005). in vivo production is a simple process of culturing epns in live insect hosts. in vivo production is based on the white’s trap method, which involves the natural migration of infective juveniles (ijs) from the infected host cadaver into a surrounding water layer, from where it can be harvested. white (1927) invented a method, later it was modified by several researchers (dutky et al. 1964, poinar, 1979, woodring and kaya 1988, lindegren et al. 1993, abdel-razek and abd-elgawad 2007). the selection of the insect host totally depends on the susceptibility of a particular host for the nematode infestation. in addition, the host must be easily culturable using cost effective materials (shapiro-ilan and mccoy 2000, ehlers and shapiro-ilan 2005). diets for the insect host rearing should be carefully selected, as the diet may influence on juvenile yields (nunchanart 2002). generally, last instar larva of the wax moth (galleria mellonella) is the conventional host used for in vivo multiplication of epns (shapiro-ilan and mccoy 2000, shapiro-ilan et al. 2002, ehlers and shapiro-ilan 2005, costa et al. 2007). galleria mellonella is naturally found in beehives and can be reared using artificial diets containing cereals, wax, yeast and glycerol; however, these ingredients are relatively expensive (costa et al. 2007). use of other insect hosts is an option for in vivo production to g. mellonella larva. such hosts are tenebrio molitor (blinova and ivanova 1987, shapiro-ilan et al. 2002, shapiro-ilan and gaugler 2002), diaprepes abbreviates (shapiro-ilan and mccoy 2000), corcyra cephalonica (blinova and ivanova 1987, karunakar et al. 1999, ganguly and singh 2000, shapiro-ilan and gaugler 2002, raj kumar et al. 2003, singh and gupta 2006, khan et al. 2007, ali et al. 2008, shapiro-ilan et al. 2012), diatraea saccharalis (folegatti et al. 1988), achroia grisella, bombyx mori (saenz and luque 2000, zaki et al. 2000), chilo sacchariphagus indicas (karunakar n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 16 et al. 1999), hellicovepa armigera (subramanian 2003, ali et al. 2008, rishi and prasad 2012), spodoptera litura (ali et al. 2008, gupta et al. 2008), plutella xylostella (rishi and prasad 2012), odontotermes obesus (devi et al. 2018) and capnodis tenebrionis (morton and gracia-del-pino 2009). in vitro epn production is the practice of culturing nematodes on a non-living, nutritive medium containing pre-cultured symbiotic bacteria. culture media should be sterilized to avoid any microbial contamination and should facilitate the epnspecific bacterial growth. sterilized medium is inoculated with bacteria followed by the inoculation of epns. infective juveniles could be collected from the second week after inoculation of epns and collected ijs could be stored in water (devi 2018). house et al. (1965) developed a dog food based medium for commercial scale production of epns. hara et al. (1981) reported the production of 125 million ijs/100 g of dog food agar within a week. animal protein-fat based media were developed by bedding (1981, 1984) viz, pig’s kidney-beef fat homogenate on shredded polyether polyurethane sponge (bedding 1981). then, bedding (1984) developed another animal protein-fat based media for epn production, such as homogenate of chicken offal for steinernematids, homogenate of chicken offal with 10% beef fats for heterorhabditds, absorbed on shredded polyether polyurethane sponge. similar chicken offal medium has been used for epn production by tabassum and shahina (2004) in pakistan. solid culture method is economically feasible up to a production level of approximately 10×1012 nematodes/month (friedman et al. 1989, ramakuwela et al. 2016). however, in in vitro production of epns, solid media compositions and culturing environment of the culture conditions should be considered. in addition, proper selection of the nutrients should be considered because it has an impact on the nematode virulence (ehlers and shapiro-ilan 2005). therefore, the objective of this study was to evaluate the production efficiency of acrobeloides longiuterus using larva, pupa and adult stages of tribolium castaneum and artificial solid nutrient media under in vitro conditions. 2 materials and methods the research was carried out at the laboratory of department of agricultural biology, faculty of agriculture (foa), university of jaffna (uoj), sri lanka during 2017 to 2018. the coleopteran insect species t. castaneum and the nematode a. longiuterus obtained from the entomology laboratory, department of agricultural biology, (foa, uoj, sri lanka) were used for this study. all the experiments were conducted under the room temperature (27 ± 2 oc). n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 17 2.1 in vivo production of acrobeloides longiuterus using tribolium castaneum red flour beetle tribolium castaneum (herbst) larvae, pupae and adults were assessed for their suitability to be used in mass culturing of nematodes following the procedure described by shapiro-ilan et al. (2002). for each life stage, ten specimens were put in a moisture chamber (9 cm diameter petri dish containing moistened filter paper). infective juveniles (ijs) were pipetted directly onto the insects at the concentrations of 50, 100 and 150 ijs/1 ml water in moisture chamber. there were four replicate dishes per each concentration and hence, 40 insects were in total per treatment. the moisture chambers were maintained at 27 ± 2 oc in darkness for 10 days. insects were deemed dead if they did not move following gentle prodding with a needle. dead insects were transferred into the white’s trap separately and ijs were collected 10 days after inoculation (dai) in water, and the final volume of the ijs suspension was adjusted to 100 ml. nematodes were counted using a counting dish under a stereo microscope (x 40 magnification). this was done by taking 1 ml nematode sample using a pipette and placing it on a counting dish. total nematode counts were taken five times. 2.2 in vitro production of acrobeloides longiuterus using flour media acrobeloides longiuterus culture was produced in vitro using artificial nutrient media containing different flour with other common ingredients. five culture media were prepared using following ratios. soy flour medium was prepared using 75 g of soy flour, 4.5 g of nutrient broth, 1.5 g of yeast extract, and 49.5 ml of corn oil mixed with 100 ml of distilled water (salma and shahina 2012). four other media were prepared replacing soy flour by 75 g of either palmyra tuber flour, corn flour, black gram flour or dhal flour. all the ingredients of a medium were mixed thoroughly and the 20 g of each was made absorbed to 5 pieces of polyurethane sponge (1.5 mm3). five sponge pieces were added into a polyethylene bag (10x15 cm2), plugged with cotton, and was replicated five times. all the bags were sterilized using an autoclave at 121oc, 1.054 kg/cm3 for 20 min. thereafter, 2 ml of bacterial suspension isolated from the nematodes was added per bag and incubated over three days under the room temperature at 27±2 oc. subsequently, 100 ijs were released into each bag. produced ijs were collected using white’s trap technique 21 dai. bacterial isolation was done as described by upadhyay et al. (2015); a. longiutreus (400-500 ijs/ml) were pipetted directly onto the larvae of t. castaneum in a moisture chamber and incubated at 27±2 oc for 72 hours in the dark. then, infected dead larva were surface sterilized by dipping them in 70% ethanol for 1-2 seconds and the cadavers of dead larvae were aseptically dissected and a loop full of haemolymph was streaked on nutrient agar plates. bacterial culture plates were incubated in an n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 18 incubator at the temperature of 27±1 oc for 48 hours. bacterial colony was washed off into a 100 ml volumetric flask and final volume of bacterial suspension was adjusted to a 100 ml suspension. 2.3 quality testing of acrobeloides longiuterus produced from the in vitro culture media against tribolium castaneum larvae and pupae a moisture chamber assay was used to test the quality of a. longiuterus from in vitro production method against t. castaneum larvae and pupae as described by kaya and stock (1997). for each life stage, 10 specimens were put into a moisture chamber. infective juveniles were pipetted directly onto the insects at concentrations of 50, 100, 150 and 200 ijs/dish in 1 ml water. control insects were only treated with 1 ml of double distilled water. there were four dishes per treatment; hence, there were 40 insects in total per treatment. the dishes were maintained at 27±2oc in darkness. mortality of pupae and larvae stages was recorded 48 and 72 hours after inoculation, respectively. lc50 and lc90 values were calculated based on the mortality data obtained by this study. the experiment was arranged in a completely randomized design. 2.4 statistical analysis all the experiments were arranged in a complete randomized design (crd) and the data of in vivo experiment were analyzed using the anovageneral linear model (glm) with two factor and comparison among the means were done using the fisher test at 95% confidence interval (ci). data from the in vitro production and quality testing were analyzed using one way-anova and mean separation were done as per the fisher test at 95% ci and the probit analysis was used to calculate the lc50 and lc90 using minitab ver. 17. 3 results 3.1 in vivo production of acrobeloides longiuterus using tribolium castaneum life stages figure 1 illustrates in vivo production of a. longiuterus using different life stages of t. castaneum at three different concentrations of infective juveniles (ijs). number of ijs was significantly different among the concentrations (f(2,54) = 422.48, p<0.05), different life stages (f(5,54) = 87.52; p < 0.05) and the interaction of concentration x life stage (f(10,54) =10.30, p<0.05). significantly highest yield of ij (36,112.5 ±14.5 ijs/pupa) was obtained with the pupal stage inoculated with 150 ijs/ml. this was followed by seventh and sixth instars larvae of t. castaneum with 150 ijs/ml n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 19 inoculation producing 31,031±9.1 and 25,425±19.2 ijs/larva, respectively. the use of t. castaneum adults as the host yielded 21,031±47 ijs/adult. generally, pupa yielded the highest number of ijs/ml at each concentration compared to the other insect stages tested. fig.1: production of acrobeloides longiuterus with respect to six life stages of tribolium castaneum at three different concentrations of ijs. 3.2 in vitro production of acrobeloides longiuterus using flour media production of a. longiuterus from the different types of flour-based media is shown in table 1. nematode production was significantly different among the media (f(4,20) = 111.51, p<0.05). soy flour medium yielded the highest ijs as 21530/20 g of medium whereas the lowest yield was observed in corn flour medium (977/20 g of medium). table 1: mean number of acrobeloides longiuterus produced under in vitro conditions. medium number of ijs ±sd/20 g* soy flour 75 g + ing. 21530 ± 143 a corn flour 75 g + ing. 977 ± 51 e black gram flour 75 g + ing. 16538 ± 380 b dhal flour 75 g + ing. 12531 ± 148 c palmyra flour 75 g + ing. 1533.8 ± 122 d ing. nutrient broth 4.5 g, yeast extract 1.5g, corn oil 49.5 ml and distilled water 100 ml *values having the same letter were not significantly different (fisher test at 95% confidence level) n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 20 3.3 quality testing of acrobeloides longiuterus produced from the in vitro culture media against tribolium castaneum larvae and pupae mortality of the larvae and pupae stages of t. castaneum at different concentrations are given in table 2. mortality of larvae at all concentrations of ijs/ml were significantly different from the untreated control (f(4,15) = 149.25, p<0.05). the highest mortalities of larvae and pupae were recorded as 92.5 % (9.25 ± 0.50 mean mortality) at the concentration of 200 ijs/ml. lc50 and lc90 of larvae were calculated as 48.55 and 210.42 ijs/ml, respectively. mortalities of t. castaneum pupa were significantly different from the untreated control at all four concentrations (ijs/ml) tested (f(4,15) = 203.2, p<0.05). mortalities of pupa were recorded as 65, 75 and 90 % at the concentrations of 50, 100, 150 ijs/ml, respectively. lc50 and lc90 of the pupae were 32. 94 and 173.97 ijs/ml, respectively. table 2: bio-efficacy of acrobeloides longiuterus produced from the in vitro culture media against tribolium castaneum larvae and pupae. concentrations ijs/ ml mean ±se mortality* larva pupa 0 0.25 ± 0.500 d 0.5 ± 0.577 d 50 5.25 ± 0.500 c 6.5 ± 0.577 c 100 7.25 ± 0.500 b 7.5 ± 0.577 b 150 8.00 ± 0.816 b 9.0 ± 0.0 a 200 9.25 ± 0.500 a 9.25 ± 0.5 a lc50 48.55 32.94 lc90 210.42 173.97 * values having the same letter in a column were not significantly different according to the fisher test at 95% confidence level 4. discussion determination of multiplication and production potential of any new nematode isolate using in vivo production in live insect hosts, as well as using in vitro production in solid culture media are the most important steps before either initiating mass production or formulation and commercialization of epns (shapiro-ilan and ehlers 2002, ehlers and shapiro-ilan 2005). thiruchchelvan et al. (2021) reported first that acrobeloides longiutreus isolated from sri lanka showed entomopathogenic characteristics. therefore, it is important to study the mass production ability for the future studies and experiments. shapiro-ilan et al. (2002) stated that in vivo production of epns could be used for the laboratory use and small-scale field experiments or applications. in vivo production of epns appears to be the suitable n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 21 method for niche markets and small growers. in addition, it requires a minimum investment cost and technical skills for the initial start-up. however, labour cost and availability on production of the insect hosts are the difficulties for in vivo production (ehlers and shapiro-ilan 2005). there are many lepidopteron, coleopteran and dipteran insect hosts that have been used for in vivo production of epns (devi 2018). the host used in this study was t. castaneum, a member of the order coleoptera and the same family of tenebrio molitor (tenibrionidae), which were used by many researchers previously (ehlers and shapiro-ilan 2005, devi 2018) for in vivo epns production. the size of the insect host plays a major role in epns production (flanders et al. 1996, kaya and stock 1997). this has been confirmed with the use of g. mellonella in epn production. larger larvae give a better yield. for instance, g. mellonella larvae (2022 mm length) yielded the highest of epn/larva (flanders et al. 1996). in this study, availability of insect host, cost effective and convenient rearing of the host insect are considerations in selecting the host insect, t. castaneum. the efficiency of in vivo production correspondingly depends on the quality of insect hosts and their life stages used for the production (morales-ramos et al. 2011). however, the quality of insect hosts and their nutritional characteristics depend on the media where they are reared and the life stages of insects (shapiroilan 2008). in this study t. castaneum was reared in wheat flour, where they are normally crowding and completing their life cycle. therefore, the ijs production differences with respect to different stages mainly depend on the nutritional properties in the particular life stage, anatomical features and the movement of the insect host. another factor that influences on ijs yield is the inoculum doses (shapiro-ilan et al. 2002). nematode concentrations in this study had a positive effect on the ijs production, and boff et al. (2000) and shapiro-ilan et al. (2002) obtained similar results. moreover, we selected intermediate ijs doses (50, 100 and 150 ij/ml) since the best yields are achieved with intermediate inoculum dosage because higher doses create lower yield due to competition for nutrients (shapiro-ilan et al. 2002). environmental factors including temperature, aeration, and moisture can affect yield of epns (grewal et al. 1994, dolinski et al. 2007, shapiro-ilan et al. 2012). optimum production temperatures lie between 18 and 28°c for different epn species (karagoz et al. 2009, morton and gracía-del-pino 2009). production of epns for the commercial use or for international markets, in vitro liquid culture is considered the most cost-effective process while in vitro solid culture is generally considered intermediate between in vivo and liquid culture (shapiro-ilan et al. 2012). soya flour-based medium yielded the highest production of ijs (21500 ijs/20 g) compared to the other four flour-based media. similar results were reported by banu and meena (2015) who recorded the yield of the nematodes ranging from 3.22 to 3.87 x 103 with highest multiplication rate of 3.87 x 103 in medium-i which has soya flour and corn oil as important ingredients. in addition, cao et al. (2013) stated that soya flour has high protein content which is important to build new tissue n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 22 in epns population that were cultivated in various in vitro media. results of this study also showed a higher nematode production with increasing protein and fat contents. soybean powder contained 40.5% of protein, 20.5% of fat, and 22.2% of carbohydrate (indriyanti and muharromah 2016). while other flours contain lower amount of protein and fat contents, such as black gram and dhal (protein 25% fat 6% carbohydrate 53%) (mazmanyan 2020), palmyra flour (protein 4% and 1% fat) (vengaiah et al. 2013) and corn flour (14% protein and 5% fat) (anonymous 2021). thus, nematode production has a positive correlation with protein and fat contents of the media or the nutrition content of ingredients. however further experiments with other flour types along with their proximate analysis will provide more information. 5 conclusions acrobeloides longiuterus was successfully cultured in vivo using different life stages: larva, pupa and adult of t. castaneum. of them, t. castaneum pupa was found as the best stage to produce ijs. among the in vitro media tested, the soy flour-based medium produced the highest yield of ijs compared to black gram, dhal, palmyra and corn flour media. the results show that mass production of a. longiuterus is feasible using the tested in vivo and in vitro methods. acknowledgements authors wish to acknowledge sri lanka council for agriculture research policy (slcarp), ministry of agriculture funded the project of national agricultural research policy [narp/12/uj/ag/01] for the funding of this research work. further, we extend our heart-felt gratitude to the reviewers who have made the valuable comments on the initial draft of the manuscript. author contributions: all authors equally contributed to the paper. nt and gm conceived and designed the study; nt conducted the research work and wrote the initial draft; gm and gtk supervised the work, reviewed and edited the manuscript. references abdel-razek as, abd-elgawad mm. 2007. investigation and efficacy of entomopathogenic nematodes against spodoptera littoralis (biosd.) and galleria mellonella (l.). archives of phytopathology and plant protection 40 (6): 414-422; doi: 10.1080/03235400600627874 ali ss, pervez r, hussain ma, ahmad r. 2008. susceptibility of three lepidopteran pests to five entomopathogenic nematodes and in vivo mass production of these nematodes. archives of phytopathology and plant protection 41 (4): 300-304; doi: 10.1080/03235400600759396 anonymous. 2021. corn flour nutritional value. (https://www.nutritionvalue.org) (accessed 02/05/2021) https://www.nutritionvalue.org/ n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 23 banu jg, meena ks. 2015. effect of different media and temperature on the multiplication and virulence of heterorhabditis indica poinar, karunakar and david, 1992. current biotica 9 (3): 239246. bedding ra. 1981. low cost in vitro mass production of neoaplectana and heterorhabditis species (nematoda) for field control of insect pests. nematologica 27 (1): 109-114; doi: 10.1163/187529281x00115 bedding ra.1984. large-scale production, storage and transport of the insect-parasitic nematodes neoaplectana spp. and heterorhabditis spp. annals of applied biology 104 (1): 117-120; doi: 10.1111/j.1744-7348.1984.tb05593.x blinova sl, ivanova es. 1987. culturing the nematode bacterial complex of neoaplectana carpocapsae in insects. in m. d. sonin, ed. helminths of insects. new delhi: american publishing co. 22–26 pp. boff m, wiegers gl, gerritsen ljm, smits ph. 2000. development of the entomopathogenic nematode heterorhabditis megidis strain nlh-e 87.3 in galleria mellonella. nematology 2 (3): 303-308; doi: 10.1163/156854100509178 cao c, jian h, zhao a, jiang x, liu q. 2013. insect protein digestion improves purity of steinernema carpocapsae in vitro culture and reduces culture period. applied microbiology and biotechnology 97 (19): 8705-8710; doi: 10.1007/s00253-013-5153-x costa jcr, dias rjp, morenz mjf. 2007. determining the adaptation potential of entomopathogenic nematode multiplication of heterorhabditis riobravus and steinernema carpocapsae (rhabditida: heterorhabditidae, steinernematidae) in larvae of alphitobius diaperinus (coleoptera: tenebrionidae) and galleria mellonella (lepidoptera: pyralidae). parasitology research 102 (1): 139-144; doi: 10.1007/s00436-007-0747-9 devi g, bhattacharyya b, mishra h, nath d, das p. 2018. rearing of entomopathogenic nematodes in termite (odontotermes obesus, ramb.). applied biological research 20 (1): 77-81; doi: 10.5958/0974-4517.2018.00009.5 devi g. 2018. mass production of entomopathogenic nematodesa review. international journal of environment, agriculture and biotechnology 3 (3): 1032-1043; doi: 10.22161/ijeab/3.3.41 dolinski c, del valle ee, burla rs, machado ir. 2007. biological traits of two native brazilian entomopathogenic nematodes (heterorhabditidae: rhabditida). nematologia brasileira 31 (3): 180185. dutky sr, thompson jv, cantwell ge. 1964. a technique for the mass propagation of the of the dd136 nematode. journal of insect pathology 6 (4): 417-422. ehlers ru, shapiro-ilan di. 2005. mass production. in: grewal ps, ehlers r-u, shapiro-ilan di, editors. nematodes as biocontrol agents. wallingford, uk: cabi publishing. flanders kl, miller jm, shields ej. 1996. in vivo production of heterorhabditis bacteriophora ‘oswego’ (rhabditida: heterorhabditidae), a potential biological control agent for soil inhabiting insects in temperate regions. journal of economic entomology 89 (2): 373-380; doi: 10.1093/jee/89.2.373 folegatti meg, alves sb, kawai prc, botelho psm. 1988. nova metodologia para produção in vivo de neoaplectana carpocapsae weiser. nematologica brasileira 12:76-83. friedman mj. 1990. commercial production and development. in r. gaugler, and h.k. kaya, eds. in entomopathogenic nematodes in biological control. crc press, boca raton, fl. p. 153-172. friedman mj, langston sl, pollit s. 1989. mass production in liquid culture of insect-killing nematodes. international patent wo89/04602. ganguly s, singh lk. 2000. steinernema thermophilum sp. n. (rhabditida: steinernematidae) from india. international journal of nematology 10 (2): 183-191. glaser rw. 1940. continued culture of a nematode parasitic in the japanese beetle. journal of experimental zoology 84 (1): 1-12; doi: 10.1002/jez.1400840102 n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 24 grewal ps, selvan s, gaugler r. 1994. thermal adaptation of entomopathogenic nematode –niche breadth for infection, establishment and reproduction. journal of thermal biology 19 (4): 245-253; doi: 10.1016/0306-4565(94)90047-7 gupta s, kaul v, shankar u, rai s. 2008. efficacy of local isolate of steinernema carpocapsae against plutella xylostella (l.). vegetable science 35 (2): 148-151. hara ah, lindegren je, kaya hk. 1981. monoxenic mass production of the entomopathogenic nematode, neoaplectana carpocapsae weiser, on dog food / agar medium. advances in agriculture technology 16: 1-8. house hl, welch he, cleugh tr. 1965. food medium of prepared dog biscuit for the mass production of the nematode dd-136 (nematoda: steinernematidae). nature 206: 847; doi: 10.1038/206847a0 indriyanti dr, muharromah nl. 2016. mass cultivation of entomopathogenic nematode in artificial media. journal of biology and biology education 8 (1): 111-118. doi: 10.15294/biosaintifika. v8i1.5579 karagoz m, gulcu b, hazir s, kaya hk. 2009. laboratory evaluation of turkis entomopathogenic nematodes for suppression of the chestnut pests, curculio elephas (coleoptera: curculionidae) and cydia splendana (lepidoptera: tortricidae). biocontrol science and technology 19 (7): 755-768; doi: 10.1080/09583150903100831 karunakar g, easwaramoorthy s, david h. 1999. influence of temperature on infectivity, penetration and multiplication of steinernema feltiae, steinernema glaseri and heterorhabditis indicus. international journal of nematology 9: 126-129. kaya hk, stock sp. 1997. techniques in insect nematology. in manual of techniques in insect pathology 281-324; doi: 10.1016/b978-012432555-5/50016-6 khan mr, khan u, askary th, mohiddin, fa, khan mm. 2007. pathogenicity and host suitability for in vivo mass production of steinernema masoodi amu epn-1. international journal of nematology 17 (2): 151-157. lindegren je, valero ka, mackey be. 1993. simple in vivo production and storage methods for steinernema carpocapsae infective juveniles. journal of nematology 25 (2): 193-197 mazmanyan v. 2020. black gram and beans nutrients. (https://foodstruct.com) (accessed 01/05/2021) morales-ramos ja, rojas mg, shapiro-ilan di, tedders wl. 2011. self-selection of two diet components by tenebrio molitor (coleoptera: tenebrionidae) larvae and its impact on fitness. environmental entomology 40 (5): 1285-1294; doi: 10.1603/en10239 morton a, gracía-del-pino f. 2009. ecological characterization of entomopathogenic nematodes isolated in stone fruit orchard soils of mediterranean areas. journal of invertebrate pathology 102 (3): 203-213; doi: 10.1016/j.jip.2009.08.002 nunchanart t. 2002. mass rearing of the entomopathogenic nematode heterorhabditis sp. on an artificial diet. food and fertilizer technology center publication, sf 14 wencho str. taipei, thailand. poinar gojr. 1979. nematodes for biological control of insects. p. 277. boca raton, fl: crc press raj kumar m, parihar a, siddiqui au. 2003. in vivo culturing of indigenous entomopathogenic nematodes from udaipur. indian journal of nematology 33 (2): 171-196. ramakuwela t, hatting j, laing md, hazir s, thiebaut n. 2016. in vitro solid-state production of steinernema innovationi with cost analysis. biocontrol science and technology 26 (6): 792-808; doi: 10.1080/09583157.2016.1159284 rishi p, prasad cs. 2012. efficacy of entomopathogenic nematode, heterorhabditis indica (meerut strain) against lepidopteran insect pest of agriculture importance. trends in biosciences 5: 321-325. saenz a, luque je. 2000. cultivation in vivo and method of storage for infective juveniles of steinernema feltiae (rhabdita: steinernematidae). agronomía colombiana 17 (1-3): 37-42. salma j, shahina f. 2012. mass production of eight pakistani strains of entomopathogenic nematodes (steinernematidae and heterorhabditidae). pakistan journal of nematology 30 (1): 1-20. shapiro-ilan di, gaugler r, tedders wl, brown i, lewis ee. 2002. optimization of inoculation for in vivo production of entomopathogenic nematodes. journal of nematology 34 (4): 343–350. n. thiruchchelvan et al. mass production of the nematode acrobeloides longiuterus ruhuna journal of science vol 12 (1): 14-25, june 2021 25 shapiro-ilan di, guadalupe rojas m, morales-ramos ja, lewis ee, tedders wl. 2008. effects of host nutrition on virulence and fitness of entomopathogenic nematodes: lipid and protein based supplements in tenebrio molitor diets. journal of nematology 40 (1): 13-19. shapiro-ilan di, han r, dolinski c. 2012. entomopathogenic nematode production and application technology. journal of nematology 44 (2): 206-217 shapiro-ilan di, mccoy cw. 2000. virulence of entomopathogenic nematodes to diaprepes abbreviates (coleoptera: curculionidae) in the laboratory. journal of economic entomology 93 (4): 1090–1095; doi: 10.1603/0022-0493-93.4.1090 singh m, gupta pr. 2006. occurrence of entomopathogenic nematodes in himachal pradesh, india and their pathogenicity against various insect species. pest management and economic zoology 14: 179189. subramanian s. 2003. in vivo production of entomopathogenic nematodes. insect environment 9: 33. tabassum ka, shahina f. 2004. in vitro mass rearing of different species of entomopathogenic nematodes in monoxenic solid culture. pakistan journal of nematology 22 (2): 167175 thiruchchelvan n, thirukkumaran g, edgington s, buddie a, mikunthan g. 2021. morphological characteristics and insect killing potential of a soil dwelling nematode, acrobeloides cf. longiuterus from sri lanka. plant protection 05 (1): 1–11; doi: 10.33804/pp.005.01.3512 thiruchchelvan n, thirukkumaran g, mikunthan g. 2019. production of an entomopathogenic nematode, acrobeloides longiuterus using artificial solid media. proceeding of 8th young scientist forum symposium, national science and technology commission 8: 208-211. upadhyay d, mandjiny s, bullard-dillard r, storms m, menefee m, holmes ld. 2015. lab-scale in vitro mass production of the entomopathogenic nematode heterorhabditis bacteriophora using liquid culture fermentation technology. american journal of bioscience and bioengineering 3 (6): 203-207; doi: 10.11648/j.bio.20150306.19 vengaiah pc, murthy gn, prasad kr, kumari ku, arul raj s. 2013. physico-chemical and functional characteristics of palmyrah (borassus flabellifer l) tuber flour. journal of plantation crops 41 (3): 437-440. white gf. 1927. a method for obtaining infective nematode larvae from cultures. science 66 (1709): 302-303; doi: 10.1126/science.66.1709.302-a woodring lj, kaya kh. 1988. steinernematid and heterorhabditid nematodes. a handbook of biology and techniques. southern cooperative series bulletin. p:28. a publication of the nematode subcommittee of the southern research project s135entomopathogens for use in pest management systems. arkansas agricultural experimental station, fayetteville arkansas. cooperatives series bulletin no.331. zaki fa, mantoo ma, gul s. 2000. in vivo culturing of entomopathogenic nematodes heterorhabditis bacteriophora and steinernema carpocapsae on silkworm (bombyx mori) and their effect on some lepidopterous insects. indian journal of nematology 30 (1): 1-4. sv-lncs ruhuna journal of science vol 7: 111, june 2016 issn: 1800-279x  faculty of science university of ruhuna  faculty of science, university of ruhuna printed in sri lanka 1 preparation and characterization of a polyacrylonitrile based gel polymer electrolyte for redox capacitors c.m. bandaranayake, w.a.d.s.s. weerasinghe, k.p. vidanapathirana and k.s. perera1 department of electronics, wayamba university of sri lanka, kuliyapitiya, sri lanka correspondence: 1kumudu31966@gmail.com received: 17th february 2016, revised: 23rd june 2016, accepted: 24th june 2016 abstract. in this study, a gel polymer electrolyte (gpe) consisting with polyacrylonitrile (pan), ethylene carbonate (ec), propylene carbonate (pc) and magnesium trifluromethane sulfonate (mg(cf3so3)2) was prepared using the hot pressed method. the starting materials were heated at 130 c for 2 hours and the resulting hot viscous mixture was pressed in between two well cleaned glass plates. the composition was fine-tuned by varying the salt and the polymer concentration in order to obtain a mechanically stable, thin and flexible film with a high ionic conductivity. it was found that the composition, 105 pan : 150 mgtf : 400 ec : 400 pc gives the maximum conductivity of 1.06 x 10-2 scm-1. dc polarization test done with blocking electrodes confirmed the ionic nature of the sample while the results obtained with non-blocking electrodes proved that the anionic contribution for the conductivity is dominant. the sample was used in redox capacitors having two identical polypyrrole electrodes doped with dodecylbenzesulfonate. cyclic voltammetry, galvanostatic charge discharge and electrochemical impedance spectroscopy techniques were used to evaluate the performance of the redox capacitors. the specific capacitance was high at low scan rates. the electrolyte was quite stable when use in the redox capacitors. further, redox capacitor was having a good cycleability which is one of the important key issues to be considered for practical applications. keywords. gel polymer electrolyte, redox capacitor, polyacrylonitrile, polypyrrole, eis. 1 introduction basically, a gel polymer electrolyte (gpe) consists of a polymer network, a salt and solvent/s. so, they exhibit higher ambient temperature ionic conductivities comparable to liquid electrolytes and good mechanical bandaranayake et al. polyacrylonitrile based gel polymer electrolyte ruhuna journal of science 2 vol 7: 1-11, june 2016 properties similar to solid electrolytes. in addition, they possess some several unique properties like ease of preparation, wide range of composition with control properties, good adhesive properties and satisfactory thermal and electrical stability. the role of the polymer network is very important as it envelopes the liquid part containing the salt and the solvent/s and thereby prevents their escape. due to that, gpes do not undergo leakage problems suffer by liquid electrolytes. hence, gpes have been recognized as better alternatives for liquid electrolytes in a wide range of applications such as batteries, electrochromic devices, dye sensitized solar cells, artificial muscles and super capacitors (liu et al. 2014; perera et al. 2007; wu et al. 2013). today, energy storage devices have received a global attention due to predicted power crisis that may come up in the next few decades. super capacitors are thought as one class of potential energy storage devices to be employed in hybrid automobiles, digital telecommunication etc. there are two types of super capacitors namely electrochemical double layer capacitors (edlcs) and redox capacitors (wang et al. 2015). edlcs use carbon based materials as electrodes. in redox capacitors, electro active materials such as conducting polymers and metal oxides are used as electrodes and they exhibit capacitive behaviour due to redox (faradaic) reactions when assembled with a suitable electrolyte. at the moment, many of the reports about redox capacitors are based on liquid electrolytes (ingram et al. 2004; hashmi 2004). but, they have undergone different disadvantages such as leakage, corrosion and bulky design. hence, use of gel polymer electrolytes in place of liquid electrolytes has emerged as a novel concept in the field of energy storage devices. in this study, a gpe consisted with polyacrylonitrile (pan), ethylene carbonate (ec), propylene carbonate (pc) and magnesium trifluromethane sulfonate (mg(cf3so3)2) was prepared by fine tuning the composition to optimize the ionic conductivity. the properties of the sample having the highest room temperature conductivity and good mechanical stability were investigated and then, its potential candidacy in conducting polymer based redox capacitors was evaluated. 2 materials and methods 2.1 preparation of gel polymer electrolytes starting materials polyacrylonitrile (pan) (aldrich,mw 150000), ethylene carbonate (ec)(aldrich, 98%), propylene carbonate (pc)(aldrich, 99%) and magnesium trifluromethane sulfonate (mg(cf3so3)2, mgtf)(aldrich, 99%) were used as received. to prepare bandaranayake et al. polyacrylonitrile based gel polymer electrolyte ruhuna journal of science 3 vol 7: 1-11, june 2016 the samples, hot-pressed method was used. first, the required amount of pan was dissolved in a mixture of ec/pc. the weight ratio of ec and pc was kept at 1:1. after performing magnetic stirring for about 2 hours, appropriate amount of mgtf was mixed in. heating was done at 130 c for about 2 hours. the resulting hot viscous mixture was pressed in between two well cleaned glass plates and vacuum dried at room temperature overnight. this procedure was repeated by varying salt concentration as well as polymer concentration. preparation steps were carried out in side an argon filled glove box. 2.2 ac conductivity measurements electrochemical impedance spectroscopy (eis) is a powerful analytical technique to gather information about the resistive and capacitive properties of materials. in this method, properties of materials are analysed using the impedance data gathered by varying the frequency. a circular shape electrolyte sample having a diameter of 14 mm was sandwiched in between two stainless steel (ss) electrodes inside a spring loaded brass sample holder which is sealed by means of an o ring. ac impedance data collection was accomplished in the frequency range 0.01 hz to 0.4 mhz at room temperature using metrohm m101 impedance analyser. this was repeated for each gpe sample having different salt and polymer concentrations. the room temperature conductivity,  was calculated using the equation,  = (1/rb)(t/a) where rb is the bulk resistance, t is the thickness and a is the area of cross section of the gpe. rb was determined by analysing the impedance data. t and diameter of the sample were measured using a micrometer screw gauge. for the composition having the maximum conductivity at room temperature, impedance measurements were taken from room temperature up to 55 c in the frequency range 0.01 hz to 0.4 mhz. conductivity at each temperature was calculated using the above equation the same composition was used for further investigations. 2.3 transference number measurements first, a circular shape electrolyte sample was loaded inside a brass sample holder in between two ss electrodes. a dc bias voltage of 1 v was applied and the current variation with time was monitored. ionic transference number, ti was calculated using the equation, ti = (ii-is) / ii where ii is the initial current and is is the steady state current. same procedure was repeated using two magnesium electrodes. the cationic transference number was evaluated using the equation, t+ = is / ii. here, ii is the initial current and is is the steady state current. bandaranayake et al. polyacrylonitrile based gel polymer electrolyte ruhuna journal of science 4 vol 7: 1-11, june 2016 2.4 preparation of conducting polymer electrodes pyrrole (aldrich) which has been stored in the dark in a refrigerator was electrochemically polymerized in the presence of sodium dodecylbenzenesulfonate (nadbs)( aldrich, 98%) on fluorine tin oxide (fto) glass plates by electrochemical polymerization using a computer controlled potentiostat (metrohm autolab 101). the concentration of the monomer, nadbs was 0.1 m. the three electrode electrochemical cell used for electrochemical polymerization was consisted of a working electrode (fto), a ag/agcl reference electrode and a pt counter electrode. polypyrrole (ppy) films of the thickness of 1 x 10-6 m were prepared by applying a current density of 1 ma cm-2. 2.5 fabrication and characterization of redox capacitors symmetrical redox capacitors of the configuration, ppy:dbs / gpe / ppy:dbs was assembled inside a glove box purged with ar. cross sectional area of the assembly was 1 cm2. cyclic voltammetry technique was used to study the withstanding of the redox capacitor for continuous cycling at a constant scan rate of 5 mvs-1. in the three electrode setup, one ppy : dbs electrode was taken as working electrode whilst the other electrode was the (reference + counter) electrodes. the redox capacitor was cycled in the potential window -2.4 v to 2.4 v. the specific capacitance was calculated using the equation, cs = 2(idv) / m(v)s (bandaranayake et al. 2015). here idv is the area of integral under the cyclic volatmmogram, m is the single electrode mass, v is the potential window and s is the scan rate. redox capacitor was subjected to galvanostatic charging and discharging in the potential window 0.0 v – 1.5 v. the constant current was set to 2 x 10-4 a. the discharge capacitance was calculated using the equation, cd = (2i) / m(dv/dt) [wang et al. 2013]. here, i is the constant current, m is the single electrode mass, dv is the discharge potential and dt is the discharge time. impedance data were collected in the frequency range from 0.01 hz to 0.4 mhz at the room temperature for the redox capacitor to investigate the resistive and capacitive properties as well as the specific capacitance. for the low frequency region, there exists a relationship between specific capacitance, cs, z” at low frequency region and f as z” = 1/ (2fcs) [pandey et al. 2013] which was used to calculate cs. 3 results and discussion figure 1 shows the variation of room temperature conductivity with the salt concentration for the gpe. the conductivity first increases until it reaches the bandaranayake et al. polyacrylonitrile based gel polymer electrolyte ruhuna journal of science 5 vol 7: 1-11, june 2016 maximum when the concentration is 150 mg and then decreases with the increase of salt concentration further. fig. 1. variation of room temperature conductivity with salt concentration the initial increment can be attributed essentially to an increase of free ions which are contributed for conductivity (jayathilake et al. 2015). when the salt concentration increases further (after the maximum), there is a tendency towards formation of ion aggregates. this reduces the availability of free mobile ions which subsequently reduces conductivity (ravindran et al. 2012). room temperature conductivity was measured by varying the polymer concentration and it is given in fig.2 fig. 2. variation of room temperature conductivity with polymer concentration. in figure 2, it is seen an increment of conductivity initially with increasing pan concentration followed by a decrement of conductivity with further increment of pan concentration. this is an indication of a simultaneous occurrence of two competing processes. they are free ion concentration at low pan concentration and low mobility of ions at high pan concentration (sharma et al. 2006; perera et al. 2011). it was found out that the composition that exhibits the maximum conductivity at room temperature was 105 pan : 150 mgtf : 400 ec : 400 pc (by weight basis). the value of the conductivity at the maximum is 1.06 x 10-2 scm-1. bandaranayake et al. polyacrylonitrile based gel polymer electrolyte ruhuna journal of science 6 vol 7: 1-11, june 2016 fig. 3 shows one of the resulting nyquist plots of the optimized composition at the room temperature. for a cell assembly with the configuration, ss / gpe / ss, the nyquist plot should contain two semicircles at high and intermediate frequencies while a spike appears at low frequency region. the semicircle at high frequency region corresponds to the bulk electrolyte and the other semicircle represents the charge transfer process at the electrode/electrolyte interface (ayathilake et al. 2014). the spike appears due to the capacitive behaviour of electrodes. in the resulting plot (fig.3), the high frequency semicircle is absent. this may be due to the unavailability of the required high frequency range in the study (>0.4 mhz) (ramesh et al. 2012). the first intercept on the real axis was used to calculate the conductivity, . fig. 3. the resulting nyquist plot obtained at the room temperature for the gpe of the composition. 105 pan : 150 mgtf : 400 ec : 400 pc . fig. 4. variation of conductivity with reciprocal temperature for the gpe of the composition, 105 pan : 150 mgtf : 400 ec : 400 pc figure 4 shows the ln  variation with reciprocal temperature. the data could be fitted with vogel-tamman-fulcher (vtf) equation,  = at-1/2 exp(-ea / (t-t0) where a is a pre-exponential factor, t is the absolute temperature and to is related to the glass transition temperature. this clues that conductivity bandaranayake et al. polyacrylonitrile based gel polymer electrolyte ruhuna journal of science 7 vol 7: 1-11, june 2016 behaviour of the system with the temperature takes places via the free volume formed inside the polymer network (jayathilake et al. 2014). figure 5 exhibits the current variation with time recorded with ss electrodes that are acting as blocking electrodes for mobile ions. fig. 5. variation of current with time for the configuration ss / 105 pan : 150 mgtf : 400 ec : 400 pc / ss at room temperature an abrupt current decrement could be observed initially followed by a steady state current (fig. 5). that initial sudden drop is due to polarization of ionic species at the electrolyte/electrode interface. the steady state current arises due to electrons. the calculated value of ti = 0.96 ascertains the dominance of ionic charge transport over electron transport in the electrolyte (kumar et al. 2010). this is highly essential to use the gpe as an electrolyte. the results of the dc polarization test done with two mg electrodes are shown in fig. 6. fig. 6. variation of current with time for the configuration mg / 105 pan : 150 mgtf : 400 ec : 400 pc / mg at room temperature the cationic and anionic contribution for the conductivity was evaluated using the resultant polarization graph shown in fig.6. the initial current drop is not bandaranayake et al. polyacrylonitrile based gel polymer electrolyte ruhuna journal of science 8 vol 7: 1-11, june 2016 very steep. due to mg ions for which the electrodes are non-blocking, current has reached the steady state soon. the value of t+, which is 0.30, implies the fact that anion contribution is dominant than cation contribution for the conductivity. fig. 7. cyclic voltammogram of the redox capacitor in the configuration, ppy : dbs / 105 pan : 150 mgtf : 400 ec : 400 pc / ppy : dbs at room temperature in fig.7, the resulting peaks are corresponding for reduction/oxidation redox reactions. the cyclic volatmmogram with such peaks also reflects the potential dependence of capacitance (hashmi et al. 2005). the mirror image around 0 v potential is a characteristic of better reversibility. the average specific capacitance calculated from the cvs is 151.67 fg-1. cvs obtained for each cycle shows similar shape as well as equal area. this is an indication for the presence of a stable specific capacitance over continuous cycling. in other words, results prove the distinctive cycleability of the redox capacitor. current values in a wide potential window demonstrate the behaviour of the gpe/ppy as charge-storing electrodes in the gpe medium for capacitor application. fig. 8. galvanostatic charge-discharge curves of a redox capacitor, ppy : dbs / 105 pan : 150 mgtf : 400 ec : 400 pc / ppy : dbs at room temperature bandaranayake et al. polyacrylonitrile based gel polymer electrolyte ruhuna journal of science 9 vol 7: 1-11, june 2016 the redox capacitors were subjected to charge and discharge under a constant current of 2 x 10-4 a. this value was extracted from the maximum value of current in the resulting cv. figure 8 shows the continuous charge discharge behaviour for several cycles. though the investigation was carried out for 1000 cycles, only few cycles are shown in the figure. the charge-discharge curves are nearly linear. the average specific capacitance was 8 fg-1 which is lower than the value obtained from the cyclic voltammetry technique. this deviation may be due to the difference in scan rates used in the two techniques. the cyclic voltammetry technique was carried out at a scan rate of 5 mvs-1. but, for charge discharge test, the calculated value of the scan rate is about 100 mvs-1. when the scan rate increases, the specific capacitance reduces as the ions can only partially penetrate into the inner surface of electrodes (sun et al. 2012 ; uppugalla et al. 2014). with that explanation, the value obtained at 5 mvs-1 should be nearly 20 times higher than the value obtained at 100 mvs-1. accordingly, the two capacitance values (151.67 fg-1 and 8 fg-1) are agreed subject to the difference of scan rate. fig. 9. a resulting nyqusit plot of a redox capacitor eis is a good technique to learn about the characteristic frequency responses of redox capacitors. figure 9 shows the nyquist plot obtained for the redox capacitor investigated in the present study. in general, a nyquist plot of a conventional capacitor should consist of a vertical line parallel to the imaginary impedance axis. but, this is not so under normal situations. hence, it consists with numerous features at different frequency regions. at high frequency region, a semi-circle appears representing the bulk electrolyte. another semicircle appears at intermediate frequency region due to the charge transfer resistance and associating double layer capacitance. at lower frequencies the capacitive behaviours become dominant. a tilted line can be seen due to the capacitance related to warburg diffusion and it will be followed by another tilted line having a steep rise representing the capacitive behaviour of the electrodes. the line having the steep rising behaviour is not perfectly parallel to imaginary axis of impedance as for a typical capacitive bandaranayake et al. polyacrylonitrile based gel polymer electrolyte ruhuna journal of science 10 vol 7: 1-11, june 2016 behaviour. it may be due to the surface roughness as well as non-uniform active layer thickness (prabhaharan et al. 2006). in the resulting nyquist plot, the semi-circle corresponding to the bulk electrolyte is absent. it may be due to the unavailability of required high frequency values. the observed bulk electrolyte resistance from the nyquist plot was used to calculate the bulk electrolyte conductivity and it was in the order of 10-2 scm-1 which is comparable with the value obtained with ss electrodes. this is an indication for the stable nature of gpe in redox capacitors. the depressed semicircle in the intermediate frequency range can be attributed for irregularities of electrode surfaces. the two tilted lines with different slopes are present and the calculated specific capacity value was about 2 fg-1 which is quite closer to the value obtained with continuous charge discharge test. 4 conclusions polymer and salt concentration affects the conductivity and the mechanical properties of the gpe. the highest conductivity at room temperature, 1.06 x 10-2 scm-1, could be obtained with the composition, 105 pan : 150 mgtf : 400 ec : 400 pc. furthermore, it was possible to obtain a bubble free, thin and flexible film from that composition. the gpe is predominantly an anionic conductor. the results obtained for redox capacitors confirm the followings. i. conducting properties of the gpe remained unchanged when employed in redox capacitors ii. gpe can withstand for continuous cycling maintaining a constant specific capacitance iii. specific capacity varies proportionally with the scan rate iv. gpe considered in the present study is a suitable candidate to be used in redox capacitors. further studies are being carried out to improve the performance. acknowledgments. authors wish to acknowledge national research council, sri lanka (nrc 12-109), university grants commission, sri lanka (ugc/vc/dric/irg-2014/wusl) and national science foundation, sri lanka (rg/2014/bs/01) for their financial assistance. references bandaranayake cm, weerasinghe wadss, vidanapathirana kp, perera ks 2015. a cyclic voltammetry study of a gel polymer electrolyte based redox-capacitor. sri lankan journal of physics. 16, 19-27 hashmi s 2004. super capacitor: an emerging power source. national academic sci. lett. 27, 27-46 bandaranayake et al. polyacrylonitrile based gel polymer electrolyte ruhuna journal of science 11 vol 7: 1-11, june 2016 hashmi sa, kumar a. tripathi sk 2005. investigations on electrochemical supercapacitors using polypyrrole redox electrodes and pmma based gel electrolytes. european polymer journal. 41, 1373-1379 ingram md, staesche h, ryder ks 2004. ladder doped polypyrrole: a possible electrode material for inclusion in electrochemical super capacitors. journal of power sources. 129, 107-112 jayathilake ymcd, perera ks, vidanapathirana kp, bandara lrak 2014. a novel gel polymer electrolyte based on polymethylmethacrylate and copper trifluromethanesulfonate. journal of electroanalytical chemistry. 724, 125-129 jayathilake ymcd, perera ks, vidanapathirana kp 2015. preparation and characterization of a polyacrylonitrile-based gel polymer electrolyte complexed with 1 methyl-3 propyl immidazolium iodide. journal of solid state electrochemistry. 19 (8), 2199-2203 kumar d, hashmi sa 2010. ionic liquid based sodium ion conducting gel polymer electrolytes. solid state ionics. 181, 416-423 liu z, abedin sze, endres f 2014. electrodeposition and stripping of zinc from an ionic liquid polymer gel electrolyte for rechargeable zinc based batteries. journal of solid state electrochemistry. 18, 2581-2587 pandey gp, rastogi ac, westgate cr 2013. polyacrylonitrile and 1ethyl 3 methyl imidazolium thiocyanate based gel polymer electrolyte for solid state supercapacitors with grapheme electrodes. ecs transactions. 50(43) 145-151 perera k, vidanapathirana kp, dissanayake makl 2007. artificial muscles based on polyacrylonitrile based electrolytes. sri lankan journal of physics. 8, 39-45 perera k, vidanapathirana kp, dissanayake makl 2011. effect of the polymer host, polyacrylonitrile on the performance of li rechargeable cells. sri lankan journal of physics. 12, 25-31 prabhaharan srs, vimala r, zainal z 2006. nano structured mesoporous carbon as electrodes for super capacitor. journal of power sources. 161, 730-736 ramesh s, lu sc, morris e 2012. towards magnesium ion conducting poly(vinylidenefluoride-hexafluoropropylene) based solid polymer electrolytes with great prospects : ionic and dielectric behaviour. journal of taiwan institute of chemical engineers 43, 806-812 ravindran d, vickraman p 2012. magnesium ionic conductivity behaviour of mixed salt system in pva-peg blend matrix. international journal of science and research publications. 2112, 1-4 sharma jp, sekhon ss 2006. pmma based polymer gel electrolytes containing nh4pf6: role of molecular weight of polymer. materials science and engineering b. 129, 104-108 sun a, xu y, wang j 2012. electropolymerized composite film of polypyrrole and functionalized multi walled carbon naotubes: effect of functionalization time on capacitve performance. journal solid state electrochemistry. 16, 1781-1789 uppugalla s, male u, srinivasan p 2014. design and synthesis of heteroatoms doped carbon / polyaniline hybrid material for high performance electrode in super capacitor application. electrochimica acta. 146, 242-248 wang w, guo s, penchev m, ruiz i, bozhilov kn, yan d, ozkan m, ozkan cs 2013. 3d few layer grapheme and carbon nano tube foam architectures for high fidelity super capacitors. nano energy, 2, 294-303 wang y, yang y, zhang x, liu c, hao x 2015. one step electrodeposition of polyaniline / nicle hexacynoferrate sulfonated carbon nano tubes interconnected composite films for super capacitors, journal of solid state electrochemistry, 19, 3157-3168 wu ty, li wb, kuo cw, chou cf, liao jw, chen hr, tseng cg 2013. study of pmma based gel electrolyte for ec devices. international journal of electrochem science, 8, 10720-10732 ruhuna journal of science vol 12 (2): 84-96, december 2021 eissn: 2536-8400 © faculty of science http://doi.org/10.4038/rjs.v12i2.104 university of ruhuna © faculty of science, university of ruhuna sri lanka 84 formulation and quality assessment of tomato (lycopersicon esculentum) cordial with xanthan stabilizer m.r. roshana*, musthapha mufeeth and u.l. abdul majeed department of biosystems technology, faculty of technology, south eastern university of sri lanka *correspondence: mrroshana30@gmail.com; orcid: https://orcid.org/0000-0002-8634-6699 received: 12th december 2020, revised: 13th september 2021, accepted: 16th december 2021 abstract tomato fruit is a nutritious part of the human diet providing a variety of health benefits. searching for other means of utilization than direct consumption of tomatoes seems a timely need to reduce post-harvest losses caused by the rapid perishability of ripened tomatoes. in the present study, three tomato cordial formulations with xanthan gum as a stabilizer were tested for sensory properties, and the best formula was tested for its physico-chemical properties and stability over one month period. well-ripened and healthy fruits were chosen to prepare the cordial. according to the sri lankan standards (sls 730), three different cordial samples were prepared only changing the concentration of xanthan gum stabilizer as 0.3%, 0.4% and 0.5% (w/v). using 30 semi-trained panellists, prepared samples were subjected to sensory testing on a 5-point hedonic scale, and 0.5% of xanthan gum containing cordial was selected as the best formula. the shelf-life evaluation of the best sample was conducted weekly for one month through physicochemical and microbiological analysis. significant changes (p<0.05) were observed only in vitamin c content and total soluble solids (tss) during the shelf-life evaluation. at end of the storage period, the cordial with 0.5% xanthan gum possessed ph 3.26, titratable acidity 1.35%, vitamin c 31.9 mg/100ml, tss 44.9 °brix, total sugar 19.8% and polyphenol content 0.0021 mg/ml gae. changes in the total plate counts during the one month of storage are within the acceptable limit (less than 50 per ml) according to the sri lankan standards (sls 516). it can be concluded that tomato cordial can be formulated by using 5% of xanthan gum stabilizer and that was found to be stable for one month without any deterioration. keywords: preservation, physiochemical analysis, shelf-life, stabilizer, value addition. 1 introduction tomato (lycopersicon esculentum l.) belongs to the family solanaceae. it originated in peru and was considered a weed due to its spreadable nature in south and central https://rjs.ruh.ac.lk/index.php/rjs/index https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v12i2.104 https://creativecommons.org/licenses/by-nc/4.0/ m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 85 america (dhaliwal 2017). at present, it is cultivated on large scale all over the world to maintain the availability of tomatoes for indigenous consumption or as an export commodity. it is an annual plant; therefore, it is considered to be a warm-season crop. in 2014, with a production of 171 million tons, the global tomato-cultivated area was 5 million ha, with china and india being the significant tomato-producing countries (faostat 2017). tomato can be filled in an assortment of geological zones in open fields or nurseries, and the organic product can be gathered by manual or mechanical methods. in the past, the tomato was not perceived as an important natural product, yet its acceptability is currently all around the world acknowledged as a nutritious food product. because of its high vitamin, malic acid, and citric acid content (serio et al. 2006), it is often referred to as the poor man's orange. in several ways, tomatoes are eaten, such as raw in salads, fried in soups, jams, ketchup, pickled sauces, and other ways. tomato is a nutrient-dense and super-food which is consumed by most people. rich with beneficial phytochemicals such as lycopene (agarwal and rao 2000), tomatoes play a vital role in preventing chronic diseases and delivering other health benefits (nasir et al. 2015). consumption of tomatoes can decrease the risk of diabetes and cancer, especially prostate cancer in men (shahar et al. 2011). increasing consumption of fruits and vegetables significantly reduces the risk of obesity and overall mortality (dias 2012). lycopene is an important antioxidant that helps to combat the development of cancerous cells and other forms of health problems and diseases. the presence of lycopene accumulation in the human body flushes out free radicals in the body (basu and imrhan 2007). among the carotenoids, lycopene is the most effective antioxidant (mortensen and skibsted 1997). from the perspective of consumers, the colour of the fruits is a significant feature. lycopene is the pigment that gives ripe tomatoes their rich red colour. the generation of lycopene is prevented when the temperature of the fruits reach 30°c. when the temperature is between 12 and 21°c, lycopene develops optimally. low light intensity reduces lycopene accumulation, resulting in inconsistent fruit colour (brandt et al. 2006). tomato fruit quality is determined by the skin colour of the fruits when the redness increases on the skin that will resemble the good flavour attributes on the skin (tigist et al. 2013). in tomato fruits, when the sugar content is at its highest, a linear relationship has been observed between the sugar content and taste, at which point the skin acquires its richest colour (samaila et al. 2011). the satisfactory texture of tomato results only when adequate pectinase, calcium and pectin are available in tomatoes. based on the maturity of tomato pectin concentration noticed that initial increment and subsequent declined (freeman and reimers 2010). tomato fruit can be divided into five portions such as; outer and inner wall, inner locule tissue, gelatinous pulp, skin, and seed. tomato’s outer and inner wall regions assume a significant function in the tomato quality because the highest contents of dry matter, insoluble solids, and decreasing sugar exist in those locales of the tomato (bhowmik et al. 2012). m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 86 tomatoes are considered a healthy source of ascorbic acid (vitamin c), with about 25 mg/100 g. in tomato-based products, acidity plays an essential role in taste. the presence of acids in tomatoes helps the food processor because ph below 4.3 can control the anaerobic microorganisms in the food while ph over 5 makes it difficult to control the spores of microorganisms (nasir et al. 2015). soluble solids of tomato mainly come from sugar which is an important contributor to the flavour. more than 60% of solids in tomatoes are free sugars such as d-glucose and d-fructose and a lower amount of sucrose, ketoheptose and raffinose (abdu 2016). starch content in tomatoes depends on the fruit maturity, cultivar, and ripening conditions. overripe fruits have low starch content than matured fruits. adding value to a product increases the shelf-life of most perishable fruits and is a way of mitigating post-harvest loss. tomatoes are well known as quickly perishable fruits, and the post-harvest loss is 54% in sri lanka (institute of post-harvest technology 2016). being a nutritionally rich commodity, tomatoes are used in different forms of food preparations which will improve human nutritional status by supplying micronutrients and bioactive phytochemicals. cordials are concentrated fruit beverages. making juice from fruits is the key stage in cordial preparation, which is then purified through a fine cloth or special juice filters to make it pure. a filtered 50-60% sugar syrup solution is heated to 90°c and the right proportion is combined with the fruit juice. the addition of hot sugar syrup to the juice decreases the boiling time and retains the colour and taste of the juice (helali et al. 2008). several hydrocolloids have been widely utilized in the food industry to produce gel structure, to improve viscosity, to function as an encapsulating agent in the coating process, to regulate crystallization, to prevent syneresis, and to improve product physical stability (dickson 2003). therefore, they can have a direct impact on food characteristics including appearance and texture. xanthan gum is a commonly used stabilizer to provide a satisfying texture, odour and taste in citrus and fruit-flavoured beverages. these beverages include xanthan gum in the range of 0.001-0.5% and facilitate the suspension of insoluble components by dissolving fast and fully at low ph. xanthan gum has branching hydrocolloid with more and lengthier branches than other forms of gum, allowing it to create more hydrogen bonds and enhance viscosity significantly (akkarachaneeyakorn and tinrat 2015). godoy (1997) used 0.175% xanthan to study the stability of guava nectar over 180 days of storage and found 99% stability. garruti (1989) reported 100% stability in passion fruit juice after adding 0.2 % xanthan gum, while souza (2009) discovered that adding 0.2% xanthan gum was one of the best treatments for stabilizing peach nectar (94.7%). commercial production of tomato-based cordial in sri lanka is not available, and this is the first attempt of research to produce and introduce it into the market. in addition, no attempt has been made to study the effect of xanthan gum in the formulation of tomato-based cordial as a stabilizer. therefore, this study was conducted with two objectives; firstly, to find the best formula for a cordial from tomato (lycopersicon esculentum) by changing the amount of xanthan gum and m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 87 testing for sensory properties, and secondly, to analyze the best formula for its physico-chemical properties and stability over one month period. 2 material and methods 2.1 tomato pulp extraction well ripened and healthy tomato fruits (variety: thilina) were procured locally from the home garden of a villager from kaluthawali, batticaloa after harvesting. heat treatment was conducted at 60°c in a 500 ppm sodium meta-bisulphite (sms) (bdh chemicals, uk) for 2 min to inactivate the enzymatic activity that decreases the browning and increases the shelf-life of the fruits (abeywickrama and jayasooriya 2010). the fruit's outer skin was stripped off and the fruits were placed in the juice extractor (model: kn-sb056, japan) to extract the pulp. to remove the seeds, the pulp was purified through a 750 μm sieve. finally, the extracted dense pulp was stored in a glass bottle and kept in a refrigerator at 4°c until further use. 2.2 preparation of tomato cordial the tomato cordial was prepared according to the method described by abeywickrama and jayasooriya (2010) with some modifications. table 1: different formulations of tomato cordial developed. the cordial was prepared as shown in table 1 according to the sri lanka standards (sls 730) guidelines. three cordial formulations were established with three concentrations of xanthan gum, i.e., 0.3%, 0.4%, and 0.5%. first, sugar was blended with water and heated to 70°c for 10 min. while stirring, xanthan gum (sigmaaldrich, usa) was added to the mixture. after dissolving the xanthan gum correctly, the tomato pulp was added to the sugar mixture and then heated using a gas burner up to 85°c for 20 min to change the brix value. the cordial was allowed to cool and sodium meta-bisulphite (smbs) (bdh chemicals, uk) was applied as a ingredients treatments with different amounts of ingredients treatment 1 treatment 2 treatment 3 water (ml) 49.17 49.07 48.97 tomato pulp (g) 25 25 25 sugar (g) 25 25 25 xanthan gum (g) 0.3 0.4 0.5 sms (g) 0.035 0.035 0.035 citric acid (g) 0.5 0.5 0.5 m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 88 preservative. the cordial was eventually filled into bottles that were pre-sterilized using a hot air oven at 160°c for 45 min and immediately sealed. 2.3 sensory evaluation to find the best formula among the cordials produced, a sensory evaluation was carried out. it was conducted using a 5-point hedonic scale. thirty semi-trained panellists were served with 1:4 (cordial: water) diluted samples of cordial. therefore, thirty replications were prepared per each treatment. only one questionnaire per person was provided to mention their responses for colour, odour, taste, mouthfeel, and overall acceptability of three cordials. friedman rank test was done to analyze sensory data using minitab 19.2 version. 2.4 shelf-life evaluation of the final product the shelf-life evaluation was carried out for one month by determining the organoleptic properties and physicochemical changes of the final product (treatment 3). a microbiological test also was carried out. all experiments were done weekly for a month. 2.5 analysis of physico-chemical properties of the final product only the best sample chosen from the sensory assessment was tested for physicochemical properties. three replications were prepared per each treatment for the analysis. by using a digital ph meter (starter 3100, ohaus, usa), the ph was measured. by titrating juices with standard naoh using an auto titrator, the titratable acidity was calculated as a percentage of citric acid. the vitamin c content was measured using 2,6-dichlorophenol indophenol dye process. using recommended aoac (2019) techniques, the total soluble solids (tss) was calculated. total sugar was measured according to the lane-eynon equation. total phenol content was spectrophotometrically calculated (siddiqui et al. 2017) based on the folin-ciocalteu process. 2.6 microbiological analysis of the product microbial analysis (total plate count) of the final product was carried out to determine the safety of the product. nutrient agar media and dilution series were prepared as follows. sterilized peptone water (9 ml) was filled into three test tubes. one milliliter of tomato cordial was added to the tube-1 and mixed (10-1 dilution). one milliliter from the tube-1 mixture was pipetted into the tube-2 and mixed (10-2 dilution). one m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 89 milliliter from the tube-2 mixture was added to the tube-3 and mixed (10-3 dilution). one milliliter from each dilution sample was pipetted out and introduced aseptically into sterilized petri dishes in two replicates. then 15 ml of media was poured into petri dishes and kept in the incubator for 24 hours. after an incubation period, colony counts were taken. 2.7 data analysis the experiment was carried out in a complete randomized design. the significance of the treatment effects on measured parameters was analyzed by analysis of variance (anova) (α = 0.05). the difference between means of nutritional parameters was determined by duncan’s multiple range test (dmrt). nutritional data were analyzed using statistical analysis system (sas) software. friedman test was done to analyze the sensory data using minitab 19.2 version software package. 3 results and discussion 3.1 selection of the best recipe from the sensory evaluation the findings of the sensory assessment of the cordial with three levels of xanthan gum are given in table 2. table 2: sensory evaluation of freshly prepared tomato cordial with three different xanthan gum levels. treatment (% xanthan gum) colour taste odour appearance mouthfeel overall acceptability t1 (0.3%) 3.65±0.09c 3.33±0.15b 3.19±0.05a 3.75±0.04a 4.05±0.02a 3.45±0.09b t2 (0.4%) 4.05±0.12b 3.05±0.02c 3.38±0.06 a 3.45±0.02 a 4.25±0.04 a 3.15±0.03c t3 (0.5%) 4.48±0.05 a 4.25±0.08 a 3.48±0.04 a 3.93±0.03 a 4.35±0.03 a 4.45±0.12 a the values are means of 30 replicates ± standard error, the means with the same letters for a given parameter are not significantly different from each other at 5% significant level based on friedman’s rank test. (t1: cordial having 0.3% xanthan gum; t2: cordial having 0.4% xanthan gum; t3: cordial having 0.5% xanthan gum) three tomato cordial samples had significant differences (p<0.05) in colour, flavour, and overall acceptability, but not (p>0.05) in the odour, appearance, and mouthfeel. treatment 3 received the maximum volume of rank for colour, taste, odour, appearance, mouth sensation, and overall acceptability with 0.5% xanthan gum. therefore, treatment 3 was selected as the best sample for the shelf-life evaluation. dogan et al. (2013) has reported that 0.5% xanthan gum as the stabilizer was the most accepted one for cloudy mulberry juice among panellists (average acceptance was 6.90 ± 1.37 points). m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 90 3.2 evaluation of the shelf-life of tomato cordial with 0.5% xanthan gum by physicochemical properties changes in the ph of the cordial during one month of storage are shown in figure 1. the initial ph of the product was 3.43 which indicated microbial stability of the product, but it decreased within one month. between the initial and end of the 4th week, there were no significant differences (p>0.05) of the ph of tomato cordial having 0.5% xanthan gum. titrable acidity and ph are inversely proportional to each other (hirdyani 2015). therefore, the product is safe for consumption within a month. the addition of sugar to the mixture increases the efficiency of the fermentation reaction of microorganisms. therefore, release of organic acid in the final product is the reason for declining the ph with time (koh et al. 2010). fig 1. changes in the ph of the cordial having 0.5% xanthan gum during the one month of storage. values represent the means of triplicate readings (n=3). error bars represent the standard error (se) of the mean. the titratable acidity of cordial having 0.5% xanthan gum increased during the storage period as shown in figure 2. there were no significant differences (p>0.05) among titratable acidity between the initial and end of the 4th week. acidity increased with time due to the slow rate of diffusion of organic acids in the product by the fermentation reaction of lactic acid bacteria (you et al. 2017). there is a positive relationship between ph and titratable acidity (islam et al. 2013). normally tomatoes are a good source of citric acid, but a small amount of malic acid and glutamic acid are responsible for the titratable acidity (anthon et al. 2011). these findings were consistent with the results of kesavanath et al. (2015) for the star fruit and sweet orange juices blend fruit cordial. 3.1 3.15 3.2 3.25 3.3 3.35 3.4 3.45 3.5 initial 1 2 3 4 p h storage duration (week) m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 91 fig 2. changes in the titratable acidity of the cordial having 0.5% xanthan gum during the one month of storage. values represent the means of triplicate readings (n=3). error bars represent the standard error (se) of the mean. the changes in vitamin c content of the cordial having 0.5% xanthan gum was slightly decreasing along with the storage period (figure 3). there was a significant difference (p<0.05) of vitamin c between the initial and end of the 4th week. fig 3. changes in the vitamin c of the cordial having 0.5% xanthan gum during the one month of storage. values represent the means of triplicate readings (n=3). error bars represent the standard error (se) of the mean. the symbol (*) indicates significant differences between the initial and corresponding week of the cordial (p< 0.05). breakdown of vitamin c into dehydro-ascorbic acid may be attributed to it (hamid et al. 2017). due to the presence of oxygen, ascorbic acid is particularly susceptible in * 27 28 29 30 31 32 33 34 35 36 initial 1 2 3 4 v it a m in c ( m g /1 0 0 m l) storage duration (week) 1.05 1.1 1.15 1.2 1.25 1.3 1.35 1.4 initial 1 2 3 4 t it ra ta b le a c id it y ( % ) storage duration (week) m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 92 its present state. higher storage temperature causes a more rapid loss of ascorbic acid (el-ishaq and obirinakem 2016), and thus ascorbic acid of tomato-based products may also be heavily affected by temperature variation. total soluble solid (tss) was increasing throughout the storage period (figure 4). during the storage time, there was a significant difference (p<0.05) in the total soluble solid between the initial and end of the 4th week. but, the final product after one month can be considered as in acceptable range, as the tss value was above 40 brix as per the cordial specification of sri lankan standards (sls 730). the slower rate of hydrolysis of carbohydrates into soluble sugar, excessive moisture losses from the final product can be the reasons for increasing the brix value of the final product (tigist et al. 2013). fig 4. changes in the total soluble solid of the cordial having 0.5% xanthan gum during the one month of storage. values represent the means of triplicate readings (n=3). error bars represent the standard error (se) of the mean. the symbol (*) indicates significant differences between the initial and corresponding week of the cordial (p-value < 0.05) fig 5. changes in total sugar content of the cordial having 0.5% xanthan gum during the one month of storage. values represent the means of triplicate readings (n=3). error bars represent the standard error (se) of the mean. 0 5 10 15 20 25 30 initial 1 2 3 4 t o ta l s u g a r (% ) storage duration (week) * 36 37 38 39 40 41 42 43 44 45 46 47 initial 1 2 3 4 t o ta l s o lu b le s o li d ( b ri x ) storage duration (week) m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 93 total sugar indicated a diminishing pattern with the storage time due to polysaccharide hydrolysis and sugar oxidation (figure 5). however, no significant change was observed (p>0.05) in the total sugar content of the finished product between the beginning and last week of the tested shelf life. kesavanath et al. (2015) found that star fruit and sweet orange fruit blended cordial had a substantial reduction in total sugar over the entire storage cycle at ambient temperature. fig 6. changes in polyphenol content of the cordial having 0.5% xanthan gum during the one month of storage. values represent the means of triplicate readings (n=3). error bars represent the standard error (se) of the mean. the polyphenol content of the final product decreased during the storage period (figure 6). there were no significant differences (p>0.05) among the polyphenol content between the initial and end of the 4th week. thermal treatment of tomatobased products will result in a loss of the total phenol content (abushita et al. 2000). in the breakdown of cellular constituents, thermal treatment such as conventional pasteurization applied to tomato related products such as ketchup and tomato juices could release more bound phenolic compounds (dewanto et al. 2002). cellular destruction, though, can unleash oxidative enzymes that could demolish phenolic compounds, therefore, continuous loss of polyphenol content was observed (vallverdú-queralt et al. 2016). 3.3 evaluation of the shelf-life of tomato cordial having 0.5% xanthan gum by microbiological tests there were no total plate counts observed in the freshly prepared cordial having 5% xanthan gum. carter et al. (2007) stated that numerous products with the addition of sodium metabisulphite alone could securely be kept sterile via the pasteurization process. 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 initial 1 2 3 4 p o ly p h e n o l c o n te n t (m g /m l) storage duration (week) m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 94 table 3: changes of the total plate count test for the final product of tomato cordial with 0.5% xanthan gum over four weeks. days mean plate count in cfu/ml 7 18 14 25 21 32 28 38 changes in the total plate counts (table 3) during the one month of storage are within the acceptable limit (less than 50 per ml) according to the sri lankan standards (sls 516). from the results, the final product had met the requirement for the shelf-life evaluation. therefore, the product is acceptable for human consumption for one month and the cost of one bottle (100 ml) of cordial was rs. 24.84 based on the raw material cost at the market. 4 conclusions the present study was conducted to formulate the tomato cordial by changing the concentration of the xanthan gum stabilizer. according to the sensory evaluation, cordial having 0.5% xanthan gum was selected as the best sample among the panelists. there were no significant changes (p>0.05) in the physicochemical properties of the 0.5% xanthan gum cordial during one month of storage period. but, vitamin c content significantly (p<0.05) decreased throughout the month where the tss was rising significantly. after a month of storage, the selected cordial sample (0.5% xanthan gum) had ph 3.26, titratable acidity 1.35%, vitamin c 31.9 mg/100ml, total soluble solids 44.9 °brix, total sugar content 19.8% and polyphenol content 0.0021 mg/ml gae. the product was found to be stable for one month, and suitable for human consumption. references abdu j. 2016. a comparative study on selected tomato varieties for manufacturing paste. master of science degree in food engineering. addis ababa university, ethiopia. abeywickrama wss, jayasooriya mcn. 2010. formulation and quality evaluation of cordial based on kirala (sonneratia caseolaris) fruit. tropical agricultural research & extension 13(1): 16-18. http://dx.doi.org/10.4038/tare.v13i1.3132. abushita aa, daood hg, biacs pa. 2000. changes in carotenoids and antioxidant vitamins in tomato as a function of varietal and technological factors. journal of agricultural and food chemistry 48: 2075-2081. https://doi.org/10.1021/jf990715p agarwal s, rao av. 2000. tomato lycopene and its role in human health and chronic diseases. canadian medical association journal 163(6): 739-744. http://dx.doi.org/10.4038/tare.v13i1.3132 m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 95 akkarachaneeyakorn s, tinrat s. 2015. effects of types and amounts of stabilizers on physical and sensory characteristics of cloudy ready-to-drink mulberry fruit juice. food science and nutrition 3(3): 213-220. doi: 10.1002/fsn3.206. anthon ge, lestrange m, barrett dm. 2011. changes in ph, acids, sugars and other quality parameters during extended vine holding of ripe processing tomatoes. journal of the science of food and agriculture 91(7): 1175-1181. https://doi.org/10.1002/jsfa.4312. aoac. 2019. official methods of analysis. (21st edition). association of official analytical chemists. washington, usa. basu a, imrhan v. 2007. tomatoes versus lycopene in oxidative stress and carcinogenesis: conclusions from clinical trials. european journal of clinical nutrition 61(3): 295. doi: 10.1038/sj.ejcn.1602510. bhowmik d, kumar ks, paswan s, srivastava s. 2012. tomato-a natural medicine and its health benefits. journal of pharmacognosy and phytochemistry 1(1): 33-43. brandt s, pe´ k z, barna e, lugasi a, helyes l. (2006). lycopene content and colour of ripening tomatoes as affected by environmental conditions journal of the science of food and agriculture 86: 568–572. https://doi.org/10.1002/jsfa.2390. carter hw, charley vls, bristol c. 2007. the preservation of fruit juice products with special reference to nutritional value. journal of cambridge 8: 12-14. doi: 10.1079/bjn19470036. dewanto v, wu x, adom kk, liu rh. 2002. thermal processing enhances the nutritional value of tomatoes by increasing total antioxidant activity. journal of agricultural and food chemistry 50: 3010-3014. https://doi.org/10.1021/jf0115589. dhaliwal s. 2017. solanaceous vegetables. handbook of vegetable crops, 3rd edition, kalyani publishers, india. 34-76 pp. dias js. 2012. nutritional quality and health benefits of vegetables: a review. food and nutrition sciences 3(10):1354–1374. doi: 10.4236/fns.2012.310179. dickson e. 2003. hydrocolloids at interfaces and the influence on the properties of dispersed systems. food hydrocolloids 17(1): 25-39. http://dx.doi.org/10.1016/s0268-005x(01)00120-5. dogan m, kayacier a, toker ös, yilmaz mt, karaman s. 2013. steady, dynamic, creep, and recovery analysis of ice cream mixes added with different concentrations of xanthan gum. food and bioprocess technology 6(6): 1420–1433. https://doi.org/10.1007/s11947-012-0872-z el-ishaq a, obirinakem s. 2015. effect of temperature and storage on vitamin c content in fruits juice. international journal of chemical and biomolecular science 1(2): 17-21. faostat. 2017. “production-crops-area harvested/ production quantity-tomatoes-2014”, fao statistics online database, food and agriculture organization, rome, www.fao.org/faostat/en (accessed 22nd sept. 2020). freeman bb, reimers k. 2010. tomato consumption and health: emerging benefits. american journal of lifestyle medicine 5(2): 182-191. doi: 10.1177/1559827610387488 garruti ds. 1989. contribuição ao estudo da estabilização física do suco de maracujá integral (passiflora edulis f. flavicarpa deg) (tese de doutorado). universidade estadual de campinas, campinas. godoy rb. 1997. gomas na estabilidade do néctar e do suco de goiaba (psidium guayava l.). (dissertação de mestrado). universidade federal de pelotas, pelotas. hamid, thakur ns, kumar p, thakur a. 2017. studies on preparation and preservation of ready-toserve (rts) beverage from underutilized mulberry (morus alba l.) fruits and its quality evaluation during storage. international journal of current microbiology and applied sciences 6(9): 1067– 1079.https://doi.org/10.20546/ijcmas.2017.609.128 helali moh, ibrahim m, shafique mz, rahman mm, biswas sk, and islam ms. 2008. formulation, preparation and preservation of lemon (citrus limon l.) cordial. journal of bio-science 16: 125127. https://doi.org/10.3329/jbs.v16i0.3755 hirdyani h. 2015. development and quality evaluation of rts (ready to serve) beverages made from traditional indian medicinal plants. journal of nutrition and food science s13: 1–4. doi: 10.4172/2155-9600.s13-004 https://www.researchgate.net/deref/http%3a%2f%2fdx.doi.org%2f10.1038%2fsj.ejcn.1602510 https://doi.org/10.1079/bjn19470036 http://dx.doi.org/10.4236/fns.2012.310179 http://dx.doi.org/10.1016/s0268-005x(01)00120-5 doi:%2010.1177/1559827610387488 https://doi.org/10.20546/ijcmas.2017.609.128 https://doi.org/10.3329/jbs.v16i0.3755 doi:%2010.4172/2155-9600.s13-004 doi:%2010.4172/2155-9600.s13-004 m.r. roshana et al. formulation and quality assessment of tomato cordial ruhuna journal of science vol 12 (2): 84-96, december 2021 96 institute of post-harvest technology. 2016. annual report-2016. retrieved from https://www.parliament.lk/uploads/documents/paperspresented/annual-report-institute-of-postharvest-technology-2016.pdf islam m, khan mzh, sarkar mar, absar n, sarkar sk. 2013. changes in acidity, tss, and sugar content at different storage periods of the postharvest mango (mangifera indica l.) influenced by bavistin df. international journal of food science 2013: 1-8. https://doi.org/10.1155/2013/939385 koh jh, kim y, oh jh. 2010. chemical characterization of tomato juice fermented with bifidobacteria. journal of food science 75(5): 428-432. doi: 10.1111/j.1750-3841.2010.01632.x. kesavanath j, premakumar k, inthujaa, y. 2015. nutritional and sensory characteristics of star fruit and sweet orange juices blend fruit cordial. international journal of multidisciplinary studies (ijms) 2(2): 1-7. https://doi.org/10.31357/ijms.v2i2.2834 mortensen a, skibsted lh. 1997. importance of carotenoid structure in radical-scavenging reactions. journal agricultural food chemistry 45: 2970-2977. https://doi.org/10.1021/jf970010s. nasir mu, hussain s, jabbar s. 2015. tomato processing, lycopene and health benefits: a review. science letters 3(1): 1-5. samaila a, amans eb, abubakar iu, babaji ba. 2011. nutritional quality of tomato (lycopersicon esculentum mill) as influenced by mulching, nitrogen and irrigation interval. journal of agricultural science 3(1): 266. doi: 10.5539/jas.v3n1p266 serio f, ayala o, bonasia a, santamaria, p. 2006. antioxidant properties and health benefits of tomato. recent progress in medicinal plants. editors: govil j.n., singh v.k., arunachalam cstudium press, llc, houston, texas (usa). 159-179 pp. shahar s, shafurah s, hasan shaari ns, rajikan r, rajab nf, golkhalkhali b, zainuddin zm. (2011). roles of diet, lifetime physical activity and oxidative dna damage in the occurrence of prostate cancer among men in klang valley. malaysia. asian pacific journal of cancer prevention 12(3): 605-611. souza jll. 2009. hidrocolóides nas características físicoquímicase sensoriais do néctar de pêssego (prunus pérsica l. batsch). (dissertação de mestrado). universidade federal de pelotas, pelotas. siddiqui n, rauf a, latif a, mahmood z. 2017. spectrophotometric determination of the total phenolic content, spectral and fluorescence study of the herbal unani drug gul-e-zoofa (nepeta bracteata benth). journal of taibah university medical sciences 12(4): 360-3. doi: 10.1016/j.jtumed.2016.11.006. sri lanka standard institute. 2010. methods of test for the microbiology of food and animal feeding stuffs horizontal method for the enumeration of microorganisms colony count at 30 oc by the pour plate technique. sls 516: 2010 sri lanka standard institute. 2010. fruit cordial concentrates, fruit squash concentrates and fruit syrup concentrates. sls 730: 2010. tigist m, workneh ts, woldetsadik k. 2013. effects of variety on the quality of tomato stored under ambient conditions. journal of food science and technology 50(3): 477-486. doi: 10.1007/s13197011-0378-0 vallverdú-queralt a, arranz s, medina-remón a, casals-ribes i, lamuela-raventós rm. 2011. changes in phenolic content of tomato products during storage. journal of agricultural and food chemistry 59(17): 9358-9365. https://doi.org/10.1021/jf202140j you sy, yang js, kim sh, hwang im. 2017. changes in the physicochemical quality characteristics of cabbage kimchi with respect to storage conditions. journal of food quality 2017: 1-7. https://doi.org/10.1155/2017/9562981. https://www.parliament.lk/uploads/documents/paperspresented/annual-report-institute-of-post-harvest-technology-2016.pdf https://www.parliament.lk/uploads/documents/paperspresented/annual-report-institute-of-post-harvest-technology-2016.pdf https://doi.org/10.1155/2013/939385 doi:%2010.1111/j.1750-3841.2010.01632.x. https://doi.org/10.5539/jas.v3n1p266 doi:%2010.1016/j.jtumed.2016.11.006. doi:%2010.1016/j.jtumed.2016.11.006. doi:%2010.1007/s13197-011-0378-0 doi:%2010.1007/s13197-011-0378-0 https://doi.org/10.1021/jf202140j https://doi.org/10.1155/2017/9562981 rjs-2007-jayasekara-yanagawa.dvi ruhuna journal of science vol. 2, september 2007, pp. 18–29 http://www.ruh.ac.lk/rjs/ i s s n 1800-279x © 2007 faculty of science university of ruhuna. testing non-linear ordinal responses in l2 × k tables leslie jayasekara department of mathematics, univerrsity of ruhuna, matara 81000, sri lanka. leslie@maths.ruh.ac.lk takashi yanagawa center for biostatistics, kurume university, japan. yanagawa.takashi@kurume-u.ac.jp abstract. in comparative studies responses of two populations are often summarized in stratified 2 × k tables with ordinal categories. a test, called qet test, is proposed for testing the homogenuity of the populations against non-linear alternatives in such tables. the asymptotic distributions of proposed test are obtained both under the null and alternative hypothesis. the powers of the qet test and extended mantal test are compared by simulation. key words : ordinal data, l2 × k tables, homogeatiy, nonlinear response, asymptotic distribution, chi-square tests, confounding variables 1. introduction data are often summarized in ordinal 2 × k tables in comparative medical studies, and statistical tests such as pearson’s chi-squared test (pearson, 1900), wilcoxon test(wilcoxon, 1945), nair’s test (nair, 1986), cumulative chi-squared test (takeuchi and hirotsu,1982), and max χ2 test (hirotsu,1983) are applied to those data for detecting the difference of two distributions. it is well known that pearson’s chi-squared test has no good powers against ordered alternatives. the wilcoxon test is specifically designed for testing location difference of two samples; also the tests are asymptotically uniformly most powerful unbiased tests for logistic linear alternatives. whereas nair’s test is designed for detecting dispersion alternatives. the cumulative chi-squared test and max χ2 test are ominibus tests developed for a wider class of alternatives including linear and non-linear responses. here we call the response patterns like a, b and c in table 1 the linear and the other patterns the non-linear; more specifically, the pattern d, e, ··· , and i respectively called the ∩ pattern, ∪ pattern, ··· , and �� � � �� pattern. we developed the qt test (jayasekara and yanagawa, 1995; jayasekara, nishiyama and yanagawa, 1999) for non-linear responses in 2 × k tables. the qt test is shown to have higher powers than those tests just described when the control and treatment groups show the combination of the patterns of non-linear responses. now confounding variables such as sex, age, blood pressure and others are involved in medical data and it is important to block their effects on testing. the above statistical tests lack this function and logistic models are conventionally employed. however, as is well known, the result of logistic models depend on the goodness of fit of the models to the data, 18 jayasekara and yanagawa: testing non-linear ordinal responses ... ruhuna journal of science 2, pp. 18–29, (2007) 19 table 1 response probabilities and patterns. ordered categories pattern 1 2 3 4 5 a 0.2 0.2 0.2 0.2 0.2 b � 0.1 0.15 0.2 0.25 0.3 c q 0.3 0.25 0.2 0.15 0.1 d ∩ 0.15 0.2 0.3 0.25 0.1 e ∪ 0.25 0.2 0.1 0.15 0.3 f �� � � 0.25 0.1 0.2 0.3 0.15 g �� � � 0.1 0.25 0.2 0.15 0.3 h �� � � �� 0.2 0.1 0.3 0.15 0.25 i �� � � �� 0.15 0.25 0.1 0.3 0.2 and yet it is not easy to establish the models, in particular, when responses are non-linear and the size of the data is not large. here we may see the raison d’etre of nonparametric tests. as far as we are aware the extended mantel test (mantel 1963, lindis, heyman, and koch, 1978, yanagawa 1986)(called emt test in the sequel) is the only test that has been developed in the sprit. the emt test adjusts for the effect of the confounding variables by stratification. in this paper we consider the same framework as the emt test and develop a test for testing the homogenuity against non-linear alternatives. more specifically, considering 2 × k tables such as those given in table 2 which have been constructed in the l-th stratum, l = 1, 2, ··· , l, to block the effect of confounding variables, we extended the qt test. it is shown that the extended qt test has higher power in most cases than emt test when the alternatives are non-linear. 2. the test statistics we suppose in table 2 that yl1 = (yl11,yl12, ··· ,yl1k)′ and yl2 = (yl21,yl22, ··· ,yl2k)′ are multinomial random vectors independently distributed with parameters nl1, (pl11, pl12, ··· , pl1k) ′ and nl2, (pl21, pl22, ··· , pl2k) ′ respectively (l = 1, 2, ··· , l). suppose that categories b1, b2, ··· , bk are ordinal (b1 < b2 < ··· < bk ), and define the odds-ratio of category bk relative to category b1 by ψlk = pl11 pl2k/pl21 pl1k (k = 1, 2, ··· , k). the homogenuity of the distributions of the control and treatment groups in the table may be represented by ψlk = 1 for all k = 1, 2, ··· , k and l = 1, 2, ··· , l, which we simply denote by ψ ≡ 1. thus the problemma is testing h0 : ψ ≡ 1 against h1 : ψlk 6= 1 for some k = 2, ··· k and l = 1, 2, ··· , l. in particular, considered under the alternatives are the odds ratios derived from the combinations of those linear and non-linear response patterns presented in table 1. we extend the qt test(jayasekara and yanagawa (1995), jayasekara and nishiyama(1996) ) for testing h0 vs. h1. let clk be the wilcoxon scorollarye in the l-th table defined by cl1 = (τl1 − nl )/2 and clk = σk−1j=1τl j + (τli − nl )/2 for k = 2, 3, ··· , k, where τlk is the marginal total in table 2. note that it is normalized to satisfy σkk=1τlk clk = 0 for l = 1, ··· , l. now for two k dimensional vectors al and bl in l-th stratum we define the inner product of al and bl by (al , bl ) = σkk=1τlkalkblk and the norm of al by ‖al‖ = (al, al ) 1/2. jayasekara and yanagawa: testing non-linear ordinal responses ... 20 ruhuna journal of science 2, pp. 18–29, (2007) table 2 2 × k table in stratum l, l = 1, ··· , l. ordered categories stratum l b1 b2 ··· bk total control yl11 yl12 ··· yl1k nl1 treatment yl21 yl22 ··· yl2k nl2 total τl1 τl2 ··· τlk nl let crlk be the r-th power of clk, and put clr = (c r l1, c r l2, ··· , c r lk ) ′, r = 0, 1, ··· , k − 1. furthermore let al0 = cl0/‖clo‖ and alr = dlr/‖dlr‖, where dlr = clr − σr−1j=0(clr, al j)al j, r = 1, 2, ··· , k − 1. note that (alr, alr′) = { 1 if r = r′, 0 if r 6= r′, for r, r′ = 0, 1, ··· , k − 1. (1) putting for given tε{1, 2, ··· , k − 1} a = (alr),l=1,2,··· ,l;r=1,··· ,t (kl × t matrix), y2 = (y ′12, ··· ,y ′ l2) ′ (kl dimensional vector), s2 = ∑ll=1 nl1nl2/nl(nl − 1), and uet = a ′y2/s we propose the following qet as a test statistic for testing h0 vs. h1: qet = u ′ et uet for each tε{1, 2, ··· , k − 1}. let ur be the r-th elemmaent of uet , then we have ur = l ∑ l=1 a′lryl2/s, (2) and the qet may represented as follows: qet = u 2 1 + u 2 2 + ··· + u 2 t . remark: qet is identical to the test statistic of emt test when t = 1, and to the wilcoxon test statistic (wilcoxon, 1945) when t = 1 and l = 1. now under h0, the conditional distribution of yl2 given cl = {nl1, nl2, τl1, ··· , τlk} is multiple hypergeometric with e[yl2k|cl ] = nl2τlk/nl cov[yl2k,yl2k′|cl ] = nl1nl2 n 2l (nl − 1) τlk(δ jk′ nl − τlk′ ), for k, k′ = 1, ··· , k, where δkk′ = 1 if k = k′ and 0 otherwise. theorem 1. under h0, the elemmaents of uet , i.e., ur, r = 1, 2, ··· , t, are uncorollaryrelated with zero mean and unit variance when conditioned on c = {cl , l = 1, ··· , l}. proof. we first show e[uet|c] = 0. putting τl = (τl1, ··· , τlk )′, we have from (1) a′lr τl = 0. (3) jayasekara and yanagawa: testing non-linear ordinal responses ... ruhuna journal of science 2, pp. 18–29, (2007) 21 thus e[uet|c] = a ′ t e[y2|c]/s = a′t (n12τ ′ 1/n1, ··· , nl2τ ′ l/nl) ′/s = ( l ∑ l=1 nl2a ′ l1τl /nl , ··· , l ∑ l=1 nl2a ′ lt τl /nl )′ /s = 0. we next compute the conditional covariance matrix of uet . since yl2, l = 1, ··· , l, are independent, the conditional covariance matrix v (uet|c) can be expressed as, v (uet|c) = a ′ tv (y2|c)at /s 2 = a′t     v (y′l2|c) . 0 0 . v (y′l2|c)     at /s 2 v (uet|c) = ( l ∑ l=1 a′lrv (y ′ l2|c)alr′ ) /s2 for r, r′ = 1, ··· , t. (4) since l ∑ l=1 a′lrv (y ′ l2|c)alr′ = l ∑ l=1 nl1nl2 n 2l (nl − 1) [nl a ′ lr     τl1 . 0 0 . τlk     alr′ − a ′ lr τl τ ′ l alr′ ], it follows from (3) that l ∑ l=1 a′lrv (y ′ l2|c)alr′ = l ∑ l=1 nl1nl2 nl (nl − 1) (alr, alr′ ). thus from (1) l ∑ l=1 a′lrv (y ′ l2|c)alr′ /s 2 = { 1 if r = r′, 0 if r 6= r′, r, r′ = 0, 1, ··· , k − 1. therefore from (4), we have v (uet|c) = it . 3. asymptotic distributions theorem 1 shows that the elemmaents of qet are uncorollaryrelated and furthermore from (2) they are linear combinations of yl2 = (yl21, ··· ,yl2k ). however, their weight vectors, alr’s, depends on nl , which makes the asymptotic theory not straightforward. we assume that when nl → ∞ the marginal totals nli and τlk for l = 1, ··· l, satisfy: (a1) nli/nl → γli, 0 < γli < 1, for i = 1, 2, and τlk/nl → ρlk, 0 < ρlk < 1, for k = 1, 2, ··· , k. to begin with we review the normal approximation of a multiple hypergeometric distribution. jayasekara and yanagawa: testing non-linear ordinal responses ... 22 ruhuna journal of science 2, pp. 18–29, (2007) 3.1. normal approximation of a multiple hypergeometric distribution plackett (1981) showed that when assumption (a1) is satisfied the asymptotic conditional distribution of xl = (yl22, ··· ,yl2k )′ given cl = {nl1, nl2, τl1, ··· , τlk}, is a k −1 dimensional normal with mean ml2 and covariance matrix vl , where ml2 = (ml22, ··· , ml2k)′ and v −1 l = (σl jk) with σlkk′ = m−1l11 + m −1 l21 + (m −1 l1k + m −1 l2k)δkk′ , for k, k ′ = 2, ··· , k and l = 1, ··· , l. here the sequence {mlik}, i = 1, 2; k = 1, 2, ··· , k, is determined uniquely by equations ∑kk=1 mlik = nli, ∑ 2 i=1 mlik = τlk , and ml11ml2k/ml21ml1k = ψlk, for i = 1, 2; k = 1, 2, ··· , k and l = 1, ··· , l. it is known (sinkhorn, 1967) that the sequence may be obtained by the following iterative scaling procedure: m (1) l1k = nl1 k , k = 1, 2, ··· , k m (1) l21 = nl2 k[1 + ∑kj=2(ψl j − 1)/k] m (1) l2k = nl2ψlk k[1 + ∑kj=2(ψl j − 1)/k] , k = 2, ··· , k m (2) lik = m (1) lik τlk m (1) l.k , m (3) lik = m (2) lik nli m (2) li. , . . . m (2h) lik = m (2h−1) lik τlk m (2h−1) l.k , m (2h+1) lik = m (2h) lik nli m (2h) li. , h = 1, 2, ··· , and l = 1, ··· , l. 3.2. asymptotic distributions under h0 we first evaluate the weight, alrk. we write n 1/2 l alrk = o(1) if and only if n 1/2 l alrk tends to a constant as n → ∞. lemma 1. if (a1) is satisfied, then (i) n−1l clrk = o(1), where clrk = c r lk, is the r-th power of the k-th wilcoxon scorollarye in the l-th table, for r = 1, 2, ··· , k − 1, k = 1, 2, ··· , k and l = 1, ··· , l. (ii) let al0k be the k-th elemmaent of al0. then n −r l (clr, al0)al0k = o(1), for r = 1, 2, ··· , k − 1, k = 1, 2, ··· , k and l = 1, ··· , l. (iii) let dlvk be the k-th component of dlv. if n −v l dlvk = o(1), k = 1, 2, ··· , k, then for any v = 1, 2, ··· , we have (a) n−2v−1l ‖dlv‖ 2 = o(1), (b) n−rl (clr, dlv)dlvk/‖dlv‖ 2 = o(1), l = 1, ··· , l. (iv) n−rl dlrk = o(1) for r = 1, 2, ··· , k − 1, k = 1, 2, ··· , k and l = 1, ··· , l. (v) n 1/2 l alrk = o(1) for r = 1, 2, ··· , k − 1, k = 1, 2, ··· , k and l = 1, ··· , l. proof. (i) by the definition of clk, and from (a1), we may get n −1 l clk = o(1) for l = 1, ··· , l. thus it is obvious that n−rl c r lk = o(1). (ii) by the definition of al0 we have al0k = jayasekara and yanagawa: testing non-linear ordinal responses ... ruhuna journal of science 2, pp. 18–29, (2007) 23 1/n 1/2 l for all k. so from (i) we obtain n −(r+1/2) l (clr, al0) = o(1). thus we have (ii). (iii) (a) the result may be obtained by the definition of dlv. (b) expanding the inner product (clr, dlv) and applying (i) we may show n −(r+l+1) l (clr, dlv) = o(1). now using (a), the result follows. (iv) to prove this result we use induction on r. in case of r = 1, dl1k = cl1k − (cl1, al0)al0k, for k = 1, 2, ··· , k. applying (i) and (ii), it follows that n−1l dl1k = o(1) for k = 1, 2, ··· , k. suppose that the result is true for r = 1, 2, ··· , m − 1. since dlm = clm − m−1 ∑ j=0 (clm, al j)al j, = clm − (clm, al0)al0 − m−1 ∑ j=1 (clm, dl j) dl j ‖dl j‖2 , it follows that n−ml dlmk = o(1) from (i), (ii) and (iii). so the result is true for r = m. thus by the induction the result follows. (v) from the definition of alr and also by (iv) the result is straightforward. next, we consider the asymptotic distribution of the test statistics under h0. to apply the normal approximation in section 3.1 we represent the t dimensional vector uet by: uet = b ′w/s, (5) where b = (blr), blr = (alr2 − alr1, ··· , alrk − alr1)′n 1/2 l , l = 1, ··· , l; r = 1, 2, ··· , t, w = (w ′1 ,w ′ 2 , ··· ,w ′ l) ′, wl = n −1/2 l (xl − nl2τl /nl ). theorem 2. under h0, qet is asymptotically distributed as a chi-squared distribution with t degrees of freedom as n → ∞, l = 1, 2, ··· , l. proof. from section 3.1 we have mlik = nliτlk/nl , under h0. thus the conditional distribution of wi given cl = {nl1, nl2, τl1, ··· , τlk} converges in distribution to nk−1(0, ∑l0) as nl → ∞, where ∑−1l0 = (σl jk0), j, k = 2, ··· , k, with σl jk0 = [ρ −1 l1 + δ jkρ −1 lk ]/(γl1γl2). furthermore, since n 1/2 l alrk = o(1) from lemmama 1(v), we have blr = o(1). thus as nl → ∞, l = 1, ··· , l, it will be easy to show that uet = b ′w/s converges in distribution to a t dimensional normal distribution with mean zero and the covariance matrix v [uet ]∞ = b ′     ∑10 . 0 0 . ∑l0     b/s2 (6) now putting ml =     ρl2(nl − ρl2) −ρl2ρl3 ··· −ρl2ρlk −ρl3ρl2 ρl3(nl − ρl3) ··· −ρl3ρlk . . ··· . −ρlk ρl2 −ρlk ρl3 ··· ρlk (nl − ρlk )     γl1γl2, jayasekara and yanagawa: testing non-linear ordinal responses ... 24 ruhuna journal of science 2, pp. 18–29, (2007) we may show ml −1 ∑ l0 = ik−1. furthermore, from (1) b′     m1 . 0 0 . ml     b/s2 ∼ it , where ∼ means that the ratio of the both hands side tends to one as nl → ∞, l = 1, 2, ··· , l. thus from (6) v [uet ]∞ ∼ it , and qet = u ′ et uet follows asymptotically a chi-squared distribution with t degrees of freedom. 3.3. asymptotic distribution under contiguous alternatives in this section we obtain the asymptotic distribution of qet under alternative hypothesis h1 : ψlk = 1 + alk/n 1/2 l , for k = 2, 3, ··· , k, where alk is a constant. lemma 2. under h1, we may represent mlik = m 0 lik + n 1/2 l ηlik + o(n 1/2 l ) for i = 1, 2, k = 1, 2, ··· , k and l = 1, ··· , l, where m0lik = nliτk/nl is the asymptotic mean under h0, and ηi1 = (−1)i+1n 1/2 l γl1γl2ρl1 k ∑ j=2 (ψl j − 1)ψl j ηik = (−1)in 1/2 l γl1γl2ρlk[ψlk − 1 − k ∑ j=2 (ψl j − 1)ψl j] k = 2, 3, ··· , k. proof. adopting the iterative scaling algoritheorem in section 3.1, we have the following expressions for m (1) l1k, m (1) l21m (1) l2k, m (2) li1 and m (2) lik , under h1. m (1) l1k = nl1 k m (1) l21 = nl2 k [1 − k ∑ j=2 (ψl j − 1) k + o(n −1/2 l )] m (1) l2k = nl2 k [ψlk − k ∑ j=2 (ψl j − 1) k + o(n −1/2 l )], k = 2, 3, ··· , k, m (2) li1 = m 0 li1 + (−1) i+1nl γl1γl2ρl1 k ∑ j=2 (ψl j − 1) k + o(n −1/2 l ), m (2) lik = m 0 lik + (−1) inl γl1γl2ρlk[ψlk − 1 − k ∑ j=2 (ψl j − 1) k ] + o(n −1/2 l ). jayasekara and yanagawa: testing non-linear ordinal responses ... ruhuna journal of science 2, pp. 18–29, (2007) 25 using mathematical induction on v, we may show m (v) lik = m 0 lik + n 1/2 l ηlik + o(n 1/2 l ), k = 1, 2, ··· , k, and v = 3, 4, ··· . thus we have the desired results. theorem 3. under h1, qet is asymptotically distributed as a non-central chi-squared distribution with t degrees of freedom. the noncentrality parameter is given by λ = ∑tr=1 δ 2 r , where δr = ∑ll=1 nl γl1γl2 ∑ k k=2 alrkρlk(ψlk − 1)/s. proof. from section 3.1 and lemmama 2 it follows that under h1, the conditional distribution of wl given cl = {nl1, nl2, τl1, ··· , τlk} converges in distribution to nk−1(ηl2, ∑l0), where ηl2 = (ηl22, ··· , ηl2k )′, and ∑l0 is that given in the proof of theorem 2. thus under h1, uet = b ′w/s converges in distribution to t dimentional normal distribution with mean δet = b′(η′12, ··· , η ′ l2) ′/s and covariance matrix v [uet ]∞, which is shown to be it in the proof of theorem 2. the r-th elemmaent of δet , say δr , is obtained as: δr = l ∑ l=1 k ∑ k=2 n 1/2 l (alrk − alr1)ηl2k/s. from (7) and ψl1 = 1, we have δr = l ∑ l=1 nl γl1γl2 k ∑ k=2 alrkρlk(ψlk − 1)/s. the theorem is immediately obtained from these results. corollary 1. the power of u 2r is approximately maximized when ln ψlk = βl alrk, k = 1, 2, ··· , k, for some constant βl , l = 1, 2, ··· , l. proof. from the proof of theorem 3 it follows that u 2r follows asymptotically a noncentral chi-squared distribution with one degree of freedom with noncentral parameter δ2r . thus the asymptotic power of u 2r for testing h0 vs. h1 may be approximated by p(u 2r ≥ χ 2 1(α)|h0) ≈ φ(δr − χ(α)), where φ is the cdf of a standard normal distribution. since δr may be represented by δr = ∑ll=1 γl1γl2(alr, ψl − 1)/s, this power is maximized when ψ − 1 = βl alr, that is when ln ψlk ≈ βl alrk for some constant βl . from the corollaryollary the statistic qet = u 2 1 + u 2 2 + ··· + u 2 t is viewed as a sum of the statistics that are asymptotically optimum against the alternatives which are expressed as log linearities of the odds ratios with scorollarye alrk, the standardized r-th power of the wilcoxon scorollarye. jayasekara and yanagawa: testing non-linear ordinal responses ... 26 ruhuna journal of science 2, pp. 18–29, (2007) 4. simulation studies simulation was conducted to compare the qet , t = 1, 2, 3, 4, test with the emt test (mantel 1963, landis, heyman and koch, 1978, yanagawa 1986). because the emt test with wilcoxon scorollarye is equivalent to the qe 1 test, we herein considered the emt test with scorollaryes 0, 1, 2, ··· , and k − 1 assigned to categories b1, b2, ··· , and bk , respectively. first we assessed type i error of the qet , t = 1, 2, 3, 4 and emt tests at the significance level α = 0.05. the response probabilities employed are those listed in table 1. we considered four strata and combinations of response patterns shown in the first column of table 3. for example, (� , ∩, � � �� , � � �� � �) in the table means that the response probabilities in the 1st stratum are p111 = p121 = 0.1, p112 = p122 = 0.15, p113 = p123 = 0.2, p114 = p124 = 0.25, p115 = p125 = 0.3; 2nd stratum are p211 = p221 = 0.1, p212 = p222 = 0.15, p213 = p223 = 0.2, p214 = p224 = 0.25, p215 = p225 = 0.3; and so on. we generated 10,000, four 2 × 5 tables for each combination of patterns and computed empirical significance levels when nl1 = nl2 = 60, 80, and 100. the results are listed in table 3. the table shows that type i error of the qet and emt tests are close to the nominal level for all combinations of patterns. second we assessed the powers of the qet , t = 1, 2, 3, 4, and emt tests. we conducted similar simulation as above by using again the response probabilities listed in table 1. considering the combinations of pattern of distribution of y1 from {( , , , ), (� , � , � , � ), ( q, q, q, q), ··· ,(�� � � ��, �� � � ��, �� � � ��, �� � � ��),(� , ∩, � � �� , �� � � ��),( q, ∪, �� � �, �� � � �� ) ,(∩, �� �� , � � � � ��, �� � � �� ), (∪, �� �� , � � � �, � � � � �� )} we computed the powers of the tests for all combinations of patterns of each distribution, 48 all together, when nl1 = nl2 = 100, l = 1, 2, 3 and 4. the tests which give the largest and second largest powers are listed in table 4a, 4b, and 4c. for example, the entry of the 2nd row and 3rd column in table 4a means that when the pattern of y1 is and that of y2 is the test with the largest power is qe 4 followed by qe 3; and the entry of the 2nd row and 4th column in table 4c means that when the pattern of y1 is and that of y2 then the test with the largest power is qe 4 followed by qe 3. the tests in the tables show that those tests have equal powers. the tables show that in most combinations, 45 among 48, the powers of the class of the qet test are larger or equal to than those of the emt test. table 5 lists the maximum, mean and minimum values of the powers of each test for 48 combinations of response patterns considered in table 4. inspection of the table shows that the mean and minimum powers of the qet test dominates the corollaryresponding values of the other tests, and that the maximum powers of the tests are almost equal. 5. discussion the qet test is proposed for testing the homogeneity against non-linear responses in l2 × k tables. we took into account the combinations of patterns of linear and non-linear responses summarized in table 1, and shown that the class of qet test is superior to the extended mantel test (mantel 1963, landis, heyman and koch 1978, yanagawa 1986). those non-linear patterns we considered often appear, for example, in phase iii randomized clinical trials for proving the efficacy of a new drug against the active control,in which the efficacy is sometimes categorized as excellent, effective slightly effective, not effective and aggravation. we emphasize that in such example, the response probabilities like 0.15, 0.25, 0.1, 0.3 and jayasekara and yanagawa: testing non-linear ordinal responses ... ruhuna journal of science 2, pp. 18–29, (2007) 27 table 3 estimated type i errors of the qet , t = 1, 2, 3, 4, and extended mantel test (emt). pattern sample size estimated type i error levels nl1 = nl2, l = 1, 2, 3, 4 qe 1 qe 2 qe 3 qe 4 emt ( , , , ) 60 0.052 0.052 0.052 0.052 0.052 80 0.05 0.048 0.048 0.047 0.051 100 0.049 0.049 0.0.05 0.051 0.049 (�, �, � , �) 60 0.054 0.052 0.052 0.049 0.054 80 0.051 0.051 0.048 0.047 0.049 100 0.052 0.052 0.051 0.05 0.053 ( q, q, q, q) 60 0.053 0.054 0.053 0.05 0.052 80 0.05 0.049 0.05 0.05 0.053 100 0.052 0.051 0.049 0.048 0.05 (∩, ∩, ∩, ∩) 60 0.052 0.051 0.049 0.049 0.051 80 0.049 0.048 0.05 0.05 0.049 100 0.052 0.047 0.05 0.05 0.052 (∪, ∪, ∪, ∪) 60 0.054 0.056 0.052 0.052 0.054 80 0.051 0.051 0.05 0.048 0.051 100 0.052 0.052 0.051 0.051 0.052 (� � �� ,�� � � ,� � �� ,� � �� ) 60 0.053 0.054 0.052 0.053 0.052 80 0.05 0.051 0.051 0.051 0.049 100 0.052 0.051 0.05 0.048 0.052 ( � � ��, � � ��, � � � �, �� � �) 60 0.052 0.05 0.051 0.052 0.053 80 0.049 0.049 0.05 0.047 0.049 100 0.051 0.052 0.051 0.05 0.052 (� � �� � �,� � �� � �,� � �� � �,� � �� � �) 60 0.054 0.054 0.052 0.051 0.053 80 0.052 0.049 0.051 0.05 0.052 100 0.051 0.052 0.051 0.05 0.052 ( � � �� � � , � � �� � � , � � �� � � , � � �� � � ) 60 0.052 0.05 0.051 0.049 0.053 80 0.05 0.05 0.05 0.049 0.05 100 0.05 0.051 0.053 0.054 0.05 (�, ∩, �� � � , � � �� � �) 60 0.053 0.054 0.051 0.05 0.053 80 0.052 0.049 0.049 0.05 0.052 100 0.053 0.052 0.054 0.05 0.055 ( q, ∪, �� � �, � � �� � � ) 60 0.054 0.052 0.051 0.048 0.053 80 0.05 0.05 0.049 0.047 0.051 100 0.049 0.048 0.049 0.049 0.048 (∩, �� �� , � � � � ��, � � � � �� ) 60 0.053 0.053 0.049 0.051 0.053 80 0.05 0.051 0.051 0.05 0.049 100 0.05 0.049 0.05 0.047 0.05 (∪, �� �� , � � � �, � � � � �� )} 60 0.055 0.051 0.048 0.05 0.054 80 0.05 0.051 0.05 0.05 0.049 100 0.052 0.05 0.049 0.048 0.051 jayasekara and yanagawa: testing non-linear ordinal responses ... 28 ruhuna journal of science 2, pp. 18–29, (2007) table 4 tests which give the largest and second largest powers: y2 y1 (�, ∩, �� � � , � � �� � �) ( q, ∪, �� � �, � � �� � � ) (∩, � � �� , �� � � ��, �� � � �� ) (∪, � � �� , �� � �, � � �� � � ) ( , , , ) qe4, qe3 qe4, qe3 qe3, qe4 qe4, qe2 (�, �, �, �) qe4, qe1 (qe1, qe2, qe3, (qe1, qe2, qe3, emt, qe2 qe4, emt) qe4, emt) ( q, q, q, q) (qe1, qe2, qe3, (qe1, qe2, qe3, (qe1, qe2, qe3, (qe1, qe2, qe3, qe4, emt) qe4, emt) qe4, emt) qe4, emt) (∩, ∩, ∩, ∩) qe2, qe4 (qe2, qe3, qe2, qe3 (qe2, qe3, qe4, emt) qe2, qe1) (∪, ∪, ∪, ∪) (qe2, qe3, qe2, qe3 (qe2, qe3, qe3, qe2 qe4, emt) qe4, qe1) (� � �� ,� � �� ,� � �� ,� � �� ) qe3, qe4 (qe3, qe4, qe2) (qe3, qe4, qe2) qe4, qe3 ( � � ��, � � ��, � � ��, � � �� (qe3, qe4, qe2) qe4, qe3 (qe3, qe4, emt) qe3, qe4 (� � �� � �,�� � � ��,� � �� � �,�� � � ��) qe4, qe3 qe4, qe3 qe4, qe3 qe4, qe3 ( � � �� � � , �� � � �� , � � �� � � , �� � � �� ) qe4, qe3 qe4, qe3 qe4, qe3 qe4, qe2 (�, ∩, �� � � , �� �� � �) qe4, qe3 qe4, qe1 qe4, qe3 ( q, ∪, �� � �, �� �� � � ) qe4, qe3 qe4, qe3 emt, qe3 (∩, �� � � , �� � � ��, �� � � �� ) qe4, qe1 qe1, qe2 qe4, qe3 (∪, �� � � , �� � �, �� � � �� ) qe4, qe3 emt, qe1 qe4, qe3 table 5 the maximum, mean and the minimum powers of the tests for 48 combinations of the patterns in table 4. qe 1 qe 2 qe 3 qe 4 emt max. 1 1 1 1 1 mean 0.343 0.561 0.705 0.832 0.345 min. 0.049 0.076 0.084 0.154 0.048 0.2,i.e. pattern is not unreasonable. it is suggested in the simulation that when all combinations of those response patterns are taken into account the qe 4 test is good choice. the qet is shown to be the sum of u 2r , r = 1, 2, ··· , t, that are asymptotically optimum against the alternatives which are expressed as log linearities of the odds ratios with scorollarye alrk, the standardized r-th power of the wilcoxon scorollarye. references pearson, k. (1900), on criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling, philos. mag., series 5, 50, 157-175. (reprinted 1948 in karl pearsons early statistical papers, ed. by e. s. pearson, cambridge university press.) nair, v. n. (1986), on testing in industrial experiments with ordered categorical data, journal of american statistical association, 58, 690-700. jayasekara and yanagawa: testing non-linear ordinal responses ... ruhuna journal of science 2, pp. 18–29, (2007) 29 takeuchi, k. and hirotsu, c. (1982), the cumulative chi square method against ordered alternative in two-way contingency tables, reports of statistical application research, japanese union of scientists and engineers, 29, 1-13. hirotsu, c. (1983), defining the pattern of association in two-way contingency tables, biometrika, 70, 579-589. jayasekara, l. and yanagawa, t. (1995) testing ordered categorical data in 2 × k tables by the statistics with orthonormal scorollarye vectors, bulletin of informatics and cybernetics, 27, 129-145. jayasekara, l., nishiyama, h. and yanagawa, t. (1999) versatile test for ordinal data in 2 × k tables, lindis, j.r., heyman, e.r. and koch, g.g. (1978), average partial association in three way contingency tables: a review and discussion of alternative tests, international statistical review, 46, 237-254. mantel, n. (1963), chi-square tests with one degree of freedom: extensions of the mantelhaenszel procedure, journal of american statistical association, 58, 690-700. plackett, r.l. (1981). the analysis of categorical data, london: griffin. sinkhorn, r. (1967), diagonal equivalence to matrices with prescribed row and column sums, amer. math. month., 74, 402-5. wilcoxon, frank (1945), individual comparisons by ranking methods, biometrics 1, 8083. ynagawa, t. (1986), analysis of discrete multivariate data, tokyo: kyoritu shuppan. 9 th ruhuna international science and technology conference https://www.sci.ruh.ac.lk/conference/ristcon2022/index.php ristcon 2022 is organized by the faculty of science, university of ruhuna, sri lanka. this conference aims to provide the premier multidisciplinary forum for leading academics, researchers, and research students to present and discuss their innovations, concerns, practical challenges encountered, and the solutions adopted in the fields of science technology. important dates initial submission deadline september 30, 2021 notification of acceptance november 15, 2021 submission of camera-ready abstract december 01, 2021 https://www.sci.ruh.ac.lk/conference/ristcon2022/index.php https://www.sci.ruh.ac.lk/conference/ristcon2022/index.php https://www.sci.ruh.ac.lk/conference/ristcon2022/index.php rjs-vol-1-sept-2006-7.dvi ruhuna journal of science vol. 1, september 2006, pp. 61–66 http://www.ruh.ac.lk/rjs/ issn 1800-279x ©2006 faculty of science university of ruhuna. a simple device to reduce the wastage of cooking gas e.m. ranatunga department of physics, university of ruhuna, matara, sri lanka, ranatung@phy.ruh.ac.lk w.g.d. dharmaratna department of physics, university of ruhuna, matara,sri lanka, dharma@phy.ruh.ac.lk we have invented a simple device to reduce the wastage of cooking gas. the main objective is to reduce the energy loss to the surroundings due to convection, radiation and conduction from pans with spherical bottoms like thachchiya used in cooking traditional food such as hoppers. in particular, the heat loss to the surroundings between the burner and the surface of the cooking pan is considered here. the efficiency of the device is compared with that of the bare pan by measuring the time taken to boil a fixed amount of water while keeping other conditions the same. the result indicates that the device reduces the gas consumption by about 23%. the device can be easily manufactured locally at low cost for practical applications. key words : energy technology, cooking gas savings 1. introduction any energy resource, whether available in abundance or in limited quantity, is to be managed effectively to get the maximum and sustained benefit. traditional sources of energy have to be supported with alternate supply in order to overcome the power crisis. in addition to finding alternate sources of energy, simple techniques have to be invented for efficient use of available resources. most of the developed countries mainly use electricity or gas for cooking purposes whereas a majority of households in some of the developing countries still depend on firewood. in order to successfully introduce energy saving techniques, one has to understand the daily routines of the people such as cooking, eating and working habits as well as their culture. without such knowledge, many organizations and individuals attempt to introduce novel technologies to the society and are often perplexed at lack of acceptance by people. hay stack (cleovoulou 1997) and straw stove (cleovoulou 1998) are two simple, low-cost and easy to understand devices built from locally available materials to introduce in tamil nadu and ghana, respectively. these devices are made to suit needs of the people and their traditional cooking practices in the respective countries rather than introducing unfamiliar new technology. these devices are proven to be successful in reducing use of firewood by 50%. hay box (arnold 1980) is another such example that reduces 58% of cooking time and 44% of the cost. the use of gas and electricity for cooking purposes has been increased significantly in developing countries such as sri lanka. devices such as rice cookers, pressure 61 ranatunga and dharmaratna: a simple device to reduce ... 62 ruhuna journal of science 1, pp. 61–66, (2006) cookers and microwave ovens have also been introduced for efficient cooking. such technologically improved devices also have to be used properly in order to save energy (cureton 1995). these devices are getting popular very slowly in sri lanka because majority of people are either unable to purchase these equipment or have no interest in such devices due to their traditional methods of cooking. therefore, in countries like sri lanka, new devices at low cost have to be invented to suit the needs and the interests of the people, in particular, for cooking traditional food. the use of liquid petroleum gas for cooking purposes in middle class households is becoming more popular in sri lanka, rapidly replacing the traditional use of firewood. however, with frequent increase of fuel prices, use of gas is becoming an additional burden, increasing cost of living by a rather significant amount. therefore, a gas saving device, which can still be used in traditional cooking would be a very useful introduction to our society. we examined several different types of pots and pans used in normal household cooking in sri lanka. in comparing them, it is clear that the traditional pan with a spherical shaped bottom, “thachchiya”, is less efficient in absorbing the heat generated from the flame. flat bottom pans, on the other hand, absorb more effectively, especially, when the size of the flame is smaller than the size of the bottom of the pan. however, the spherical bottom prying pans, thachchiya (it is used to make hoppers as well), are used in every household for traditional cooking and frying which is not replaceable by any pots or pans of other shapes. therefore, we have studied the possibility of reducing energy loss while cooking in this particular shape of pans. we have attempted to design a simple device to minimize the loss of heat when cooking using standard gas cookers. in cookers, the gap between the flame and the bottom of the pan is open, and hence some of the heat is lost to the surroundings. we have studied the possibility of reducing the heat loss through the gap, using proper insulating and reflecting materials in specific shapes and sizes to cover the gap. here we report the results of a study that indicates the possibility of saving at least 23% of gas with a very simple and low cost device that can be manufactured locally and marketable easily. some preliminary results have been submitted for publication elsewhere (ranatunga and dharmaratna 2006) in abstract form. 2. materials and methods in using standard gas cookers, especially with thachchiya, the heat is lost to the surroundings from the region between the burner and the pan as mentioned earlier. it is certain that all three processes, convection, radiation and conduction contribute with different orders of magnitudes to the heat loss in this region. therefore, the amount of gas used in cooking can be reduced by minimizing the heat loss to the surroundings in this region and directing the heat generated towards the pan. in all studies discussed below, a standard burner and a gas cylinder with a standard controller available in the market are used to provide the heat for cooking. aluminum pan with approximately spherical bottom (a thachchiya, diameter 18 cm) filled to about 3/4 of the volume with a known amount of water is used for all tests. a temperature sensor connected to a computer through a data acquisition ranatunga and dharmaratna: a simple device to reduce ... ruhuna journal of science 1, pp. 61–66, (2006) 63 figure 1 a sketch of the device unit is used to measure the variation of the temperature of water with time. the time taken to heat the fixed amount of water from room temperature up to the boiling point is used as a measure of useful heat absorbed by water. the fraction of the heat absorbed by the pan depends on the size of the flame, which depends on the rate of gas flow, and the height between the burner and the pan. the first study was to determine the optimal height between the burner and the bottom of the pan. for a reasonably good flame (almost blue flame) the time taken to boil the water sample is measured for various heights of the pan. the rate of flow of gas was kept the same for all measurements. the procedure was repeated to find the optimal height between the burner and the pan. the optimal height was found to be about 5 mm under the conditions used in this experiment (size of the pan, size of the flame, etc.) and it was kept as a constant for the following tests. in order to reduce the heat loss in the gap between the burner and the bottom of the pan, various devices were tried. figure 1 shows the device constructed after several experiments. this particular device is a result of five experiments, namely, (a) the pan with no cover, (b) the gap covered with a cylindrical shaped aluminum foil with an insulating top, (c) same as (b) with partially covered bottom with aluminum foil on the top of an insulating sheet, (d) same as (b) with the cylindrical foil covered with glass wool (thickness 2.5cm) and (e) same as (c) with cylindrical foil covered with glass wool as in (d). the setup used in experiment (e) is shown in figure 1. for each of the above cases the time taken to boil the same amount of water is measured while keeping the other conditions, such as the optimal height to the bottom of the pan, the rate of gas flow and other environmental conditions, the same. 3. results and discussion for the selected rate of gas flow with a good blue color flame, the time taken to boil the water depends on the height between the burner and the bottom of the pan. the optimum value for the height was found to be 5 mm for the rate of gas flow used. ranatunga and dharmaratna: a simple device to reduce ... 64 ruhuna journal of science 1, pp. 61–66, (2006) figure 2 comparison of the variation of temperature with time the variation of temperature of water from room temperature to the boiling temperature with time for the above five cases, (a) to (e), are shown in fig. 2. reading errors for the points in the plot are smaller than the size of the symbol. as can be seen from the plots, the time taken to boil water decreases from experiment (a) to (e), gradually. the numerical values with percentages of savings are compared in table 1. this clearly indicates that each component added in each step of experimental setup (from (b) to (e)) contributes a significant factor in energy saving and also in reduction of cooking time. a total saving of 23% has been achieved with the experimental set up (e) in fig. 1, which is a quite significant fraction of energy saving. setup (b) blocks the heat loss due to convection, which is the hot air moving away from the pan. this alone reduces the gas used by 10%. it is interesting to see that the setup (c) with the addition of cover on the bottom of the pan reduces 5% more. still the aluminum foil gets warm and some energy is lost to the surroundings due to radiation as well as convection. having the glass wool layer as insulation around the aluminum foil reduces the energy loss by at least another 7%. having insulating layers at the top and bottom also help to reduce energy loss. note that the cylindrical aluminum foil cannot be completely sealed to the pan at the top edge, because the burned out gas has to move out of the covered space. from this study it is clear that trapping the hot air around the pan for some time and reducing the energy loss to the surrounding one can reduce the gas usage by at least 23%. as can be seen here, this is a very simple and less expensive device that can be manufactured by a local company and market very easily for a reasonably low ranatunga and dharmaratna: a simple device to reduce ... ruhuna journal of science 1, pp. 61–66, (2006) 65 table 1 comparison of experiments (a)-(e) price, which can save at least 23% of gas used in cooking. it should be noted that the device constructed here is important for any type of pan with spherical shaped bottom although the study was done specifically for thachchiya. furthermore, the calculated savings in (e) is in compared with (a), which is with a well-controlled flame with optimal height between the burner and the pan. in normal household cooking the wastage of gas could be much higher than our controlled experiment (a). therefore, the proper use of this device could save even a higher percentage of energy than shown in the above table when compared with the normal practice of household cooking. 4. conclusion we have successfully invented a very simple low cost device that can be used with cooking pans with spherical bottoms, which reduces the use of gas by at least 23%. the device could be manufactured locally with locally available materials. it could be marketable easily, since this type of pans are used in almost every household for cooking traditional food and the gas savings is quite significant when compared with the price of gas. the study is in progress to improve the device, because the optimal dimensions of the device depend on the size of the cooking pan. references cleovoulou m. 1997. introducing fuel-saving cooking methods in southern tamil nadu, http://www.cleovoulou.com/fuelsave.htm cleovoulou m. 1998. introducing fuel-saving cooking methods in northern ghana, http://www.cleovoulou.com/fsghana.htm. arnold b. 1980. rediscovery the hay-box cooker, mother earth news, no. 61. ranatunga and dharmaratna: a simple device to reduce ... 66 ruhuna journal of science 1, pp. 61–66, (2006) cureton m. 1995. inefficient home appliances may be stealing your water, robbing your electricity and assaulting your pocketbook. rocky mountain institute, http://www.sdearthtimes.com/ et0595/ et0 95s1.html ranatunga em. dharmaratna wgd. 2006. a simple device to use cooking gas efficiently, (preliminary results in abstract form), accepted for slaas 2006. ruhuna journal of science vol 8: 1-11, june 2017 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v8i1.22 university of ruhuna  faculty of science, university of ruhuna 1 sri lanka farm waste utilization among farmers in irepodun local government area, kwara state, nigeria: implication for extension education service delivery f.o. oladipo1, o.d. olorunfemi1, 2*, o.d. adetoro1 and t.o. oladele1, 2 1department of agricultural extension and rural development, faculty of agriculture, university of ilorin, ilorin, nigeria 2department of agricultural economics and extension, faculty of agriculture, science and technology, north west university, mafikeng campus, south africa *correspondence: davidsoa2003@yahoo.com received: june 10th 2016, revised: april 15th 2017, accepted: april 21st 2017 abstract. the study examined the utilization of farm waste among farmers in irepodun local government area of kwara state, nigeria. structured interview schedule was used to elicit data from 120 farmers in the study area. results revealed that more than half (58.4%) of the respondents were crop farmers who grow majorly maize and cassava on a subsistence scale. the major farm wastes generated in the area were maize cobs, husk and stalk (62.5%) and cassava stalk and peels (60%). majority of the farmers got rid of their farm waste through burning while more than half of the respondents do not utilize the waste they generate from their farms. the major constraints militating against farm waste utilization includes inadequate access to extension services, inadequate awareness of benefits of farm waste, inadequate facilities for processing of farm wastes to other products and low knowledge on usage of farm waste. logistic regression modelling results revealed that farmers who are more likely to utilize their farm waste efficiently are those with higher level of education and more years of experience in farming. it was therefore recommended that extension agencies should embark on enlightenment campaigns and trainings of farmers on various innovative ways of farm wastes utilization in order to facilitate more efficient and environmental friendly farm waste utilization initiatives in the area. keywords. assessment, farmers, farm waste, kwara state, utilization. 1 introduction one of the main features of agriculture today is waste, which is inevitable on farmlands. globally, 140 billion metric tons of waste is generated every year from agriculture. this volume of waste can be converted to an enormous f. o. oladipo et al. farm waste utilization among farmers in irepodun ruhuna journal of science 2 vol 8: 1-11, june 2017 amount of energy and raw materials (unep 2009). as defined by shaban and omaima (2010), farm wastes are residues produced as a result of various agricultural operations. household farms in rural communities generate solid organic wastes such as manure, tree trimmings, grass clippings, and crop residues such as rice husk, rice straws, maize stalk, maize husk, maize cobs, cassava peels and stalk, groundnut shells and straws, soy beans pods, sugarcane bagasse and leaves, and cotton stalk. organic wastes can amount up to 80 percent of the total solid wastes generated in any farm household. also, livestock generate large amounts of wastes. manure production can amount up to 5.27 kg/day/1000 kg live weight, on a wet weight basis (mbam and nwibo 2013). farm waste contains many reusable substances of high value but it seems that they are most of the time taken next to nothing by a majority in nigeria farmers. these large volumes of biomass can be converted to an enormous amount of energy and raw materials depending on the availability of adequate technology. they can be converted into commercial products either as raw material for secondary processes, as operating supplies or as raw materials of new products (gunther et al. 2003). this process can be termed creativity in agriculture, ‘creativity is making marvellous out of the discarded’, i.e. creating “wealth” from “wastes”. wealth can be created from wastes when they are recycled or further utilized (auke and japp 1997). a case of songhai farm in port novo is a show case of the system that produces almost zero waste by recycling waste into biogas used for cooking, electricity, fertilizers etc. the abundant farm residues in the rural communities can also be converted into resources for generating wealth. for instance, cocoa pods are generated on cocoa plantation and it has been estimated that about 150 kg dry pods per hectare are left in the field as it provides a valuable source of potash fertilizer (lim, 1986). also, palm kernel shell, wood chippings that are left in the fields or burnt off in nigeria have the capacity to generate over 750mw of electricity (egun 2012). okey et al. (2014) also reported that plant material and animal wastes used primarily as domestic energy source are naturally abundant in rural communities and present a renewable energy opportunity that could serve as an alternative to fossil fuel. composting agricultural and other types of wastes can be a useful process for recycling nutrients and maintaining or restoring levels of organic matter in the soil (solano et al. 2001). composting of farm and organic waste can be an attractive, low-cost technology for farmers. utilization of the finished product on the farm can help the farmer avoid some of the costs associated with the purchase of organic matter, fertilizer and soil conditioners (romeela 2007). livestock wastes are used as soil conditioner (hermanson 2005), fuel source either by direct combustion or converted to biogas (jones et al. 2005), and livestock and fish feeding (sevilleja et al. 2005). livestock wastes could also constitute nuisance through environmental pollution especially the liquid component which seep into the ground contaminating both surface and ground f. o. oladipo et al. farm waste utilization among farmers in irepodun ruhuna journal of science 3 vol 8: 1-11, june 2017 water, and this is why many legislations have been made to protect the environment while various improved methods of waste management have been prescribed towards ameliorating the impact of wastes on the environment (anon. 2005; nebraska department of environmental quality, 2005). agricultural wastes have been reported to be a large and an underexploited resource, almost always underestimated (rosillo-calle 2007). wastes can be used as fuel, fodder, manure, fibre, feedstock and further uses; this establishes the slogan waste-to-wealth which means moving waste from a platform of exhausted utility to valuable and desirable level. it has been observed that rural farmers have little knowledge about wastes utilization and are not well informed about modern economically viable waste utilization innovation. farm wastes depending on utilization could either be assets in improving the living standards of farmers, if their benefits are maximized, or potential hazards to the environment where they are generated. this study, in a bid to provide information and insight for extension agencies and policy makers in ensuring that better waste utilization strategies are extended to farmers aims to assess farm wastes utilization among farmers in irepodun local government area of kwara state, nigeria. specifically, the study sought to describe the socio-economic characteristics of the farmers in the study area, identify the types of farm wastes generated on the farms, examine the methods of farm waste management on the farms in the study area, determine the utilization of the farm wastes generated, identify the constraints militating against farm waste utilization in the study area and identify the determinants of farm waste utilization in the study area. 2. materials and methods the study was carried out in irepodun local government area in kwara state, nigeria. farmers of the area cultivates food crops such as yam, maize, guinea corn, cocoyam, cassava, rice, locust bean as well as shea butter and cash crops such as cocoa, kola and oil palm. the population for the study consists of both crop and livestock farmers in irepodun local government area of kwara state. structured interview schedule was used to collect information from one hundred and twenty farmers (120) sampled in the area. two-staged sampling procedure was used in the study. first, was a purposive sampling of six wards from the eleven wards in the local government area based on their farming population base, namely; esie, omu-aran, oko, ajaseipo, ipetu and arandun wards. this was then followed by a random selection of 20 farmers from each of the selected wards carried out with the aid of a list of registered farmers/sampling frame gotten from the kwara state agricultural development project (adp). f. o. oladipo et al. farm waste utilization among farmers in irepodun ruhuna journal of science 4 vol 8: 1-11, june 2017 socio-economic characteristics of the respondents, the types of waste generated, their method of waste management and the constraints they face in utilizing waste generated from their farms were the key data groups. these variables were measured as follows. ‘constraints to farm waste utilization’ was measured on a 3-point likerttype scale of very severe (3), severe (2), and not severe (1). ‘farm waste utilization’: respondents were asked to indicate whether they utilize the waste generated from their farms and this was measured on yes (1) or no (0) basis. data analysis was carried out using descriptive statistics such as frequency counts, percentages, means and ranks. logistic regression was used in modelling the determinants of farm waste utilization of the respondents. logistic regression model is widely used to analyze data with dichotomous dependent variables. the binary logistic regression model is stated as: yj = β0 + β1x1 + β2x2… + β6x6 + ε where, yj is the binary variable with value 1 if farmers utilize the waste generated from their farms and 0 if otherwise (if farm waste is not utilized). β0 is the intercept (constant), and β1, β2, to βn are the regression coefficients of the predictor variables, x1, x2, …… x6 respectively, and ε is the independent and normally distributed random error, and x1 = age (in years) x2 = gender (measured as a dummy variable 1 for male, 0 if otherwise) x3 = marital status (measured as a dummy variable 1 for married, 0 if otherwise) x4 = household size (measured in number of persons) x5 = educational level (measured as 1 for formal educated ranging from adult, primary, secondary and tertiary education and 0 if otherwise) x6 = years of experience (measured as number of years in farming profession). f. o. oladipo et al. farm waste utilization among farmers in irepodun ruhuna journal of science 5 vol 8: 1-11, june 2017 3 results and discussion 3.1 socio-economic characteristics of the respondents results from table 1 revealed that majority (72.5%) of the respondents were male, while less than one-third (27.5%) are female, showing that, the women in irepodun lga are less involved in farming activities. this result affirms that, traditionally, women are regarded as homemakers, who oversee and coordinate the affairs and activities at home. however oladejo et al. (2011) revealed that beyond being homemakers, women are still very relevant in agriculture as they engaged in manual processing of food crops and other farm produce in addition to their housekeeping duties. majority (74.2%) of the farmers are within 45-65 years of age, which means that they are still in their economically active years indicating a high degree of prospects to be more receptive to ideas and innovations as regards farm waste utilization and management. majority (85.8%) of the respondents were married, which shows they have family responsibilities ties that will require more financial commitment which may serve as an impetus for them to exploit the prospects in farm waste utilization. this is in consonance with titus et al. (2015) who stated that agriculture is primarily practiced by married people in the rural areas of nigeria. findings in table 1 further revealed that most of the respondents (79.2%) had one form of formal education or the other and this attribute might enhance the farmers’ information seeking behaviour on farm waste management and utilization thus facilitating their high rate of adoption of new technology relating to waste usage and disposal. the mean household size of the respondents was 7 persons. this finding on household size implies that the respondents could draw some level of family labour from their household as regard management of waste generated from their farms. findings also revealed that a little above average (51.7%) of the farmers engaged in trading as a form of secondary occupation while only a handful (11.7%) of them were not involved in any secondary occupation but solely depend on farming. the finding is in agreement with mbam and nwibo (2013) who reported that farmers engaged in various farm and non-farm activities as a way of income diversification which helps in reducing poverty among the farming household. table 1 further showed that more than half (58.4%) of the farmers in the area are crop farmers implying that majority of the waste generated in the study area will be crop related as only few (35.8%) of the farmers are into livestock farming. the mean years of farming experience in the area was 15 years. this indicates that majority of the respondents are highly experienced and this is supposed to enhance the acquisition of their knowledge and skills in the area their farming enterprise and waste management and utilization. f. o. oladipo et al. farm waste utilization among farmers in irepodun ruhuna journal of science 6 vol 8: 1-11, june 2017 table 1. socio-economic characteristics of respondents. characteristics percentage gender male female 72.5 27.5 age < 25 26-45 45-65 >65 3.3 14.2 74.2 8.3 marital status single married widowed 5.9 85.8 8.3 education level no formal education primary education secondary education tertiary education 20.8 45.8 31.7 1.7 household size 1-5 6-10 11-15 49.0 38.3 12.7 secondary occupation teacher trader civil servant craft work none type of farmer crop livestock both years of experience 1-10 11-20 > 21 8.3 51.7 5.0 23.3 11.7 58.4 35.8 5.8 23.3 54.4 22.3 source: field survey; n = 120 3.2 types of farm waste generated table 2 reveals the profile of solid wastes generated in the area. the findings showed that husk, stalk and cobs of maize, and cassava stalks and peels were the major waste generated in the area while animal related waste was only generated in small quantities. this confirms the earlier finding that majority of the respondents in the area were crop farmers and indicates that they were f. o. oladipo et al. farm waste utilization among farmers in irepodun ruhuna journal of science 7 vol 8: 1-11, june 2017 mostly involved in maize and cassava farming as wastes from these crops were the predominant solid waste generated in the area. table 2. distribution of the respondents based on the types of farm wastes generated. type of wastes percentages grasses (weed) leaves cassava stalk & peels 55.8 44.2 60.0 maize cobs, husk & stalk yam peels 62.5 15.0 vegetable wastes soybeans straw & pods poultry droppings 15.8 32.5 36.7 livestock manure poultry feathers 27.5 35.0 egg shells fruits (droppings) 33.3 7.5 livestock mortalities wood shavings 32.5 22.5 oil palm fronds kernel shafts 5.0 5.0 source: field survey; n = 120 3.3 techniques of waste disposal findings from table 3 revealed that majority of the farmers got rid of the waste generated from their farms which they consider not useful to them through burning and only a few of them used compost pit or dump sites. it was further revealed that some of the farmers still got rid of their farm waste by depositing them in nearby streams and rivers. this indicates that majority of the respondents still do not know the implications of improper disposal of farm wastes on the environment and the resultant negative effects the constant waste disposal techniques like burning and stream dumping have on their health and their immediate and global environment. table 3. distribution of respondents based on their technique of farm waste disposal management. percentages waste disposal dump site compost pit burning stream 21.7 19.2 55.8 3.3 source: field survey; n = 120 f. o. oladipo et al. farm waste utilization among farmers in irepodun ruhuna journal of science 8 vol 8: 1-11, june 2017 this implies that there is a need to enlighten the respondents on the environmental hazards that improper waste disposal and burning of farm waste can cause, and to teach them better waste management and utilization techniques that will benefit them socio-economically and be more sustainable for their environment. 3.4 farm waste utilization findings from the study showed that more than half (53.3%) of the respondents were not utilizing the waste generated from their farms while the remaining 46.7% stated that they use their farm waste. this implies that majority of the farmers have little or no knowledge about the benefits and socio-economic potentials they can derive by utilizing the waste generated from their farms. this points out the need for farmers to be enlightened by extension agents on the various farm waste management and utilization initiatives which include utilization as manure, mulching, compost, dried as feed and sale to other farms and recycling agencies they can leverage upon for better economic livelihoods and sustainable agricultural practice in the study area. 3.5 constraints to farm waste utilization from the findings in table 4, it was observed according to the mean scores that the major/severe constraints militating against the utilization of farm waste indicated by the respondents include inadequate access to extension services, inadequate awareness of benefits, inadequate facilities for processing some waste, and low knowledge on usage. the implication of this finding is that the famers need access to more effective extension services that will help improve their knowledge on the benefits and farm waste utilization techniques they can adopt for better waste management and usage in the study area. table 4: constraints to farm wastes utilization. constraints mean score low knowledge on usage 2.37 bad odour from wastes 2.23 inadequate of awareness of benefits 2.53 stress of transportation 2.32 inadequate facilities for processing some wastes 2.53 alternative products 2.03 inadequate access to extension services 2.64 limited labour 2.10 source: field survey; n = 120 f. o. oladipo et al. farm waste utilization among farmers in irepodun ruhuna journal of science 9 vol 8: 1-11, june 2017 3.6 determinants of respondents’ farm waste utilization logistic regression results of determinants of respondents’ farm waste utilization revealed that the coefficient of educational level and years of experience were significant at 5 percent level of significance indicating that these two variables significantly influence the utilization of farm waste by the farmers (table 5). the variables of age, gender, marital status and household size were not significant even at 10 per cent indicating that these variables do not significantly influence the respondents’ farm waste utilization in the study area. the coefficients of educational level (0.321) positively and significantly (p<0.05) determine the utilization of farm waste by the respondents in the study area. this implies that an increase in the educational level of the farmers in the study area is going to lead to an increase in the way they will adopt farm waste utilization initiatives. furthermore, the parameter of the respondents’ years of farming experience (0.523) was also seen to positively and significantly (p<0.05) influence the farm waste utilization of the respondents’. this indicates that an increase in the years of experience of the respondents’ increases their likelihood to adopting and utilizing more farm waste innovation. thus the overall summary implication for the findings from the logistic model reveals that the characteristics of farmers that are more likely utilize the waste generated from their farms efficiently are those with higher level of education and with more years of experience in the farming profession. table 5: logistic regression results of the factors influencing respondents’ farm waste utilization. variables coefficient standard error t-value gender 0.129 0.042 0.81 age 0.423 0.089 0.56 marital status 0.023 0.035 1.25 educational level 0.321** 0.219 2.42** household size 0.457 0.271 0.18 years of experience 0.523** 0.034 2.16** model chi-square = 186.234 -2 log likelihood = 59.233 overall case corrected predicted = 89.5% **significance level at 5%; source: analysis of field survey data 4 conclusion the findings of the study revealed that the majority of the farmers in the study area are not utilizing the waste generated from their farms and they result to f. o. oladipo et al. farm waste utilization among farmers in irepodun ruhuna journal of science 10 vol 8: 1-11, june 2017 disposing them predominantly through burning. the major constraints highlighted by the respondents inhibiting their efficient use of farm waste were inadequate access to extension services which results in them having inadequate awareness of the benefits they can derive from utilization of farm waste and low knowledge of its usage. 5 recommendations the study therefore recommended that the government should look into providing some modern waste processing facilities in the area that will assist the farmers convert the waste from their farms into more durable and useful products. also, extension agencies and other relevant stakeholders should embark on enlightenment campaigns and trainings of farmers in the areas of socio-economic prospects and benefits of waste utilization and the various innovative ways through which they can use them. these will result in the farmers adopting more efficient and environmental friendly farm waste utilization initiatives that will enhance proper farm waste management systems and leveraging on the socio-economic potentials derivable through their use of these innovative farm waste management. references anon. 2005. legislation for management of wastes. hong kong environmental protection department. retrieved january 26th, 2015 from http://www.epd.gov.hk auke k., jaap k. 1997. agricultural and forest residues generation, utilization and availability. paper presented at the regional consultation on modern applications of biomass energy, kuala lumpur, malaysia. egun nk. 2012. the waste to wealth concept; waste market operation in delta state nigeria. greener journal of social sciences, 2(6):206-212. falaye ae. 1993. utilization of agro-industrial wastes as fish feedstuffs in nigeria. in: 10th annual conference of the fisheries society of nigeria (fison). 16-20 november 1992, abeokuta, nigeria, pp. 47-57. gunther l, benno k, marianne n. 2003. transformation of vegetable waste into value added products: (a) the upgrading concept; (b) practical implementations. bioresour technol 87(2): 167-198. hermanson re. 2005. managing livestock manure to protect groundwater. washington state university cooperative extension. retrieved december 17, 2014 from, http://www.whatcomcd.org jones dd, nye jc, dale, ac. 2005. methane generation from livestock waste. retrieved january 15th, 2015 from http://danpatch.ecn.purdue.edu lim ko. 1986. the energy potential and current utilization of agriculture and logging wastes in malaysia , renewable energy review journal, vol. 8, no. 2, december 1986, rericait, bangkok. f. o. oladipo et al. farm waste utilization among farmers in irepodun ruhuna journal of science 11 vol 8: 1-11, june 2017 mbam bn, nwibo su. 2013. entrepreneurship development as a strategy for poverty alleviation among farming households in igbo-eze north local government area of enugu state nigeria: greener journal of agricultural sciences, 3(10):736-742. nebraska department of environmental quality. 2005. fact sheet on 1999 livestock legislation. retrieved january 25, 2015 from http://www.deq.state.ne.us/priority.nsf/ okey fo, busayo sa, nneoma na. 2014. biomass briquetting and rural development in nigeria. published june 2, 2014. oladejo ja, olawuyi so, anjorin td. 2011. analysis of women participation in agricultural production in egbedore local government area of osun state, nigeria: international journal of agricultural economics and rural development, 4(1). romeela m. 2007. composting as an effective waste management strategy for farm households and others. food and agriculture organization of the united nations rome. rosillo-calle f. 2007. overview of bio-energy. in: the biomass assessment handbook: bioenergy for a sustainable environment. f. rosillo-calle, p. de groot, s.l. hemstock and j. woods, (eds.). published by earth scan, uk. sevilleja r, torres j, sollows j, little d. 2005. using animal wastes in fishponds. retrieved august 12, 2015, from http://www.fao.org/documents shaban da, omaima ms. 2010. the utilization of agricultural wastes as one of the environmental issues in eygpt (a case study). journal of applied sciences research, 6(8):1116 1124. solano ml, iriarte f, ciria p, negro mj. 2001. structure and environmental performance characteristics of three aeration systems in composting of sheep manure and straw. journal of agric, engineering research, 79:317-329. titus ob, adebisola oa, adeniji ao. 2015. health-care access and utilization among rural households in nigeria. journal of development and agricultural economics, 7: 195-203. unep 2009. converting waste agricultural biomass into resource: compendium of technologies. united nations environmental programme 2009, division of technology, industry and economics international environmental technology centre osaka/shiga, japan. 441pp. ruhuna journal of science vol 7: 21-31, december 2016 eissn: 2536-8400  faculty of science university of ruhuna  faculty of science, university of ruhuna 21 sri lanka effect of na2sio3 on heavy metal uptake by field grown basella alba l. in matara, sri lanka samanthika r. hettiarachchi 1* and darshani weerahewa 2 1 department of chemistry, the open university of sri lanka, p.o. box 21, nawala, nugegoda 10250 sri lanka. 2 department of botany, the open university of sri lanka, p.o. box 21, nawala, nugegoda 10250 sri lanka. correpondence: 1 srhet@ou.ac.lk received: august 05 th 2016, revised: september 29 th 2016, accepted: october 10 th 2016 abstract. in this study, we investigated heavy metal uptake and the effects of na2sio3 on heavy metal absorption by field grown basella alba l (basellaceae). the concentrations of fe, cr, pb and cd in the field soils were 29755.30 ± 292.02, 32.99 ± 0.97, 26.01 ± 1.02, 0.13 ± 0.004 µg/g, respectively. these concentrations are significantly below the maximum permissible limits reported by fao/who. although fe, cr, pb and cd were present in the soil, only fe was absorbed by b. alba; the tissue concentrations of other heavy metals were below the detection limit. the distribution of fe from soil to different plant parts was investigated by calculating transfer factors. low transfer factors indicated low absorption and translocation of fe from soil to plant tissue. we also investigated the effects of na2sio3 on metal absorption by applying two different concentrations of na2sio3 (si-100 mg/l and si50 mg/l) alongside a control. there was a significant reduction of fe absorption in b. alba treated with si-100mg/l of na2sio3 compared to that of plants treated with si50 mg/l of na2sio3 and the control. keywords. heavy metal uptake, maximum permissible limits, silicon 1 introduction farmers use synthetic and organic fertilizers as well as pesticides at several stages during the cultivation of vegetables. these fertilizers and pesticides enhance the growth of crops by providing essential nutrients and by controlling pests but some of these cause soil contamination via the release of heavy metals (mclaughlin et al. 2000). heavy metals can be toxic even under low concentrations and can have lasting impacts on human and ecosystem health (gall et al. 2015). most farmers apply fertilizers 2-8 times in excess than the dosage recommended by the department of agriculture, sri lanka (jayathilaka et al. 1989). although several studies have been carried out on soil contamination by heavy metals in the up country region (elevation 1000 hettiarachchi and weerahewa heavy metal uptake by field grown basella alba ruhuna journal of science 22 vol 7: 21-31, december 2016 m above the sea level) of sri lanka (jayathilaka et al. 1981, premarathna et al. 2005), limited research has been conducted to date on heavy metal contamination in the low country region of sri lanka (premarathne et al. 2011). farmers in the low country region of sri lanka mainly use animal fertilizers such as cattle and poultry manure during the cultivation of green leafy vegetables, including basella alba l (premarathna et al. 2011). these natural fertilizers contain relatively high concentrations of zn, se, mn, co, as and fe (bolan et al. 2010). consumption of heavy metal contaminated vegetables is one of the direct pathways of heavy metal entry to the food chain (sharma et al. 2009; chen et al. 2014). world health organization (who) has proposed maximum permissible limits for different heavy metals in soil and vegetables (chiroma et al. 2014). leafy vegetables are known to accumulate heavy metals (neilson and rajakaruna 2014); several studies have been carried out to assess heavy metals in vegetables (guptha et al. 2010; abah et al. 2014, kananke et al. 2014; rajapakshe et al. 2011; premarathna et al. 2005), including those grown in sri lanka. results of studies conducted in sri lanka have shown that heavy metals present in several leafy vegetables grown in certain areas of the country are above the maximum permissible limits set by who (guptha et al. 2010; abah et al. 2014; kananke et al. 2014; rajapakshe et al. 2011; premarathna et al. 2005; jayasinghe et al. 2005; rathnayake et al. 2004). silicon (si) is present as silicate minerals in the earth’s crust and these minerals undergo chemical and physical withering, finally getting incorporated in to the soil. in the soil solution, si is present as uncharged monomeric orthosilicic acid (h4sio4) with concentration in the range of 0.10.6 mm (epstein et al. 1994; ma et al. 2002). although si is the second most abundant element on the earth’s crust, plants can absorb si only in the form of orthosilicic acid which is quickly precipitated as amorphous si after absorption (lux et al. 2003). therefore, amorphous si is the only form of si present in plants (ding et al. 2008). amorphous si particles that are precipitated in plant cells are called phytoliths but the locations and the proportions can vary with the plant species as well as the age of plants (ponzi et al. 2003; sangster et al. 2001). number of studies has shown that metal toxicity can be alleviated with the application of small quantities of si (neumann et al. 2001; ma et al. 2002; rogalla et al. 2002; liu et al. 2009; ma et al. 2008). ma and takahashi (2002) have showed that after the application of si, the oxidizing capacity of roots increases so that ferrous ions oxidize to ferric ions, preventing the uptake of fe. manganese toxicity is also reduced with the application of si because mn binds to the cell wall, limiting hettiarachchi and weerahewa heavy metal uptake by field grown basella alba ruhuna journal of science 23 vol 7: 21-31, december 2016 cytoplasmic concentrations (rogalla et al. 2002). neumann and zur nieden (2001) showed that with the application of si, zn can be co-precipitated with si in cell walls, resulting in less soluble zn in plants. it was also shown that silicic acid has the ability to decrease as accumulation (ma et al. 2008). as heavy metals have persistent and accumulative nature, they have the ability to concentrate through the food chain and reach lethal doses to humans (sharma et al. 2009; gall et al. 2005). therefore, it is important to analyze heavy metals present in field grown vegetables such as b. alba to determine whether they comply with the permissible limits proposed by who. basella alba is a green leafy vegetable with important mineral nutrients; people frequently include this leafy vegetable in their diet. we conducted the present study to investigate heavy metal absorption by b. alba and the effect of si on heavy metal absorption by the plant. 2. materials and methods 2.1 study site the study was conducted at sulthanagoda, matara district, southern province of sri lanka. average annual temperature and the annual rain fall of this area are 26.7 0 c and 2327 mm, respectively. 2.2 experimental design experiment was conducted using three different concentrations of na2sio3 (si0 mg/l-control, si50 mg/l and 100 mg/l). each treatment was composed of three replicate beds (1m x 2m × 2m) arranged in a randomized complete block design (rcbd). basal fertilizer (npk) application was done as recommended by the department of agriculture, sri lanka (bolan et al. 2010). cattle and poultry manure were also applied as basal fertilizer. we commenced this work on 15.01.2015. 2.3 plant material seeds of b. alba were from one mother plant of the farmer’s field at sulthanagoda and they were sawn in a nursery and maintained for one month, and then seedlings were transferred into beds. hettiarachchi and weerahewa heavy metal uptake by field grown basella alba ruhuna journal of science 24 vol 7: 21-31, december 2016 2.4 preparation of plant and soil samples prior to the analysis of heavy metals (before the application of na2sio3) after one month of planting, three plants were pulled off randomly from each bed and washed with tap water, followed by three separate washes with deionized water and air dried. then, roots, stems and leaves were separated from each plant, cut into small pieces, freeze dried for 5 days in separately labeled zip lock bags and stored at -4 0 c until analyses were carried out. 10 g of three soil samples were taken from the middle of each bed up to 1 feet depth from the rhizosphere of the harvested plants, using a stainless steel spatula, mixed well and air dried for two days followed by oven drying at 70 o c for three days. soil samples were kept in labeled zip lock bags until further analysis. 2.5 application of na2sio3 and preparation of plant and soil samples prior to the analysis of heavy metals three different concentrations of liquid na2sio3 (si 100 mg/l, 50 mg/l and 0 mg/l) were added to the rooting zone of the plants in the three replicate beds as a spray application on a weekly basis for two months. after one month of na2sio3 application, three plants from each bed were pulled out and washed with tap water followed by three washes with deionized water and air dried. then the roots, stems and leaves were separated and cut into small pieces and freeze dried for five days. labeled samples were stored at -4 o c until analyses were carried out. 10 g of three soil samples from each bed were collected, air dried, and then oven dried for three days at 70 o c. dried soil samples were kept in labeled zip lock bags. same procedure was repeated for the samples collected after second month. 2.6 analysis for heavy metals soil samples: soil samples were crushed, sieved (less than 2 mm pore size) and mixed to obtain homogenized mixtures. approximately 4 g of soil sample was ashed using a muffle furnace for 6-8 h by controlling the temperature within the range of 490-500 0 c. subsequently, the sample was cooled down to room temperature and about 10 ml of analytical grade hcl: hno3 (1:3) mixture was added and the resultant sample, filtered using 0.45µm filter paper, hettiarachchi and weerahewa heavy metal uptake by field grown basella alba ruhuna journal of science 25 vol 7: 21-31, december 2016 transferred in to a 50 ml volumetric flask and diluted up to the mark with deionized water. gf-aas was calibrated by using fischer scientific calibration standards and the results were obtained from graphite furnace atomic absorption spectrophotometer (gf-aas) [model gbc 932+, australia]. plant materials: plant materials were ground and mixed well to get homogenized mixtures. the procedure used for soil samples was followed for acid digestion of plant materials, but, only 10 ml of hno3 was added instead of hcl:hno3 (1:3) mixture. concentration of heavy metal ions were obtained from graphite furnace atomic absorption spectrophotometer after calibration. 2.7 transfer factors in order to understand the translocation of fe into different plant parts, transfer factors among different plant parts:soil were calculated. fe transfer factor of leaves:soil = concentration of fe in leaves ----------------------------------- concentration of fe in soil 3 results and discussion 3.1 heavy metals present in soil as shown in table 1, higher concentrations of fe, cr and pb were detected compared to cd in all the soil samples tested. concentrations of fe, cr, pb and cd in field soils were 29755.30 ± 292.02, 32.99 ± 0.97, 26.01 ± 1.02, 0.13 ± 0.004 µg/g, respectively. all the concentrations were below the maximum permissible limits set by who (chiroma et al. 2014). premarathna et al. (2011) have reported heavy metal concentrations in certain crops and soil in up-country and some parts of low-country regions of sri lanka. they have observed much higher concentrations of cd and pb in soil samples than those observed in the present study. according to their results, concentration of cd in sedawatta in colombo district, sri lanka has been in the range of 0.61-3.28 µg/g whereas those in kandapola in nuwara eliya district, sri lanka has been in the range of 0.39-1.96 µg/g. concentrations of pb were reported to be 39-118 and 27-97 µg/g respectively. some of these reported concentrations in upand low-country soils of sri hettiarachchi and weerahewa heavy metal uptake by field grown basella alba ruhuna journal of science 26 vol 7: 21-31, december 2016 lanka are higher than the maximum permissible limits. in contrast, the soil samples analyzed in the present study showed negligible amount of cd and considerably lower amount of pb when compared to the maximum permissible limits (chiroma et al. 2014). table 1. heavy metals present in soil after the silicon (si) treatments. metal concentration in soil (µg/g) before spraying na2sio3 1 month after spraying na2sio3 2 months after spraying na2sio3 fe si (100 mg/l) 28582.31 ± 290.45 a 27877.96 ± 290.45 d 27240.80 ± 296.86 g si (50 mg/l) 27362.71 ± 287.40 b 25188.91 ± 298.52 e 27171.35 ± 286.38 g si (0 mg/l) 29755.30 ± 292.02 c 29445.89 ± 294.41 f 29523.45 ± 293.61 h cd si 100 mg/l 0.13 ± 0.004 a 0.11 ± 0.004 c 0.11 ± 0.004 d si 50 mg/l 0.00002 ± 0.004 b 0.00012 ± 0.004 b 0.00002 ± 0.004 e si 0 mg/l 0.13 ± 0.004 a 0.11 ± 0.004 c 0.00002 ± 0.004 d cr si 100 mg/l 35.56 ± 1.02 a 30.89 ± 1.03 cd 34.56 ± 1.06 f si 50 mg/l 29.19 ± 1.00 b 29.52 ± 0.98 ce 29.54 ± 1.04 g si 0 mg/l 32.99 ± 0.97 a 30.91 ± 1.01 de 32.56 ± 1.06 fg pb si 100 mg/l 28.31 ± 1.11 ab 27.90 ± 1.15 d 28.10 ± 1.29 f’ si 50 mg/l 25.50 ± 1.13 ac 21.15 ± 1.27 e 23.79 ± 1.53 g silicon 0 mg/l 26.01 ± 1.02 bc 24.43 ± 1.13 de 21.59 ± 1.22 g the values which share the same letter have no significant difference. comparisons were carried out for each metal separately. based on a one-way anova, a significant difference (p<0.05) was noted for soil fe among the treatments (100 mg/l, 50 mg/ l, and the control) during both application stages: before spraying (28,582.31 µg/g, 27,362.71 µg/g, 29,755.30 µg/g, respectively) and one month after spraying (27,877.96 µg/g, 25,188.91 µg/g, 29,445.89 µg/g, respectively). two months after spraying, a significant difference was noted in both treatments, including 100 mg/l (27,240.80 µg/g) and 50 mg/l (27,171.35 µg/g) compared to the control (29,523.45 µg/g). furthermore, there was no significant difference in the fe concentrations in soil among three application stages. hettiarachchi and weerahewa heavy metal uptake by field grown basella alba ruhuna journal of science 27 vol 7: 21-31, december 2016 3.2 fe accumulation in different plant parts of basella alba table 2 summarizes fe transfer factors of different plant parts:soil. all transfer factors are very small thus only small portion of fe present in soil has been translocated to plant tissue. transfer factors decreased in the order of root:soil > stems:soil > leaves:soil. in other words, translocation of fe decreased from the bottom to the top of the plant. about one thousandth of fe in soil was transferred to leaves of b. alba. table 2. fe transfer factors of different plant parts:soil of b. alba. plant part: soil fe transfer factors leaves: soil 1.30 ×10 -3 stems:soil 1.90 ×10 -3 roots:soil 9.70 ×10 -3 although heavy metals were present in the soil, only fe was absorbed by b. alba. all the other heavy metals were below the detection limit. kananke et al. (2014) have reported heavy metal accumulation in some leafy vegetables including b. alba collected from open market sites in piliyandala area in colombo district, sri lanka. according to their report, concentrations of ni, cd, cr and pb in b. alba were above the maximum permissible limits set by fao/who (chiroma et al. 2014). si-mediated heavy metal absorption has been observed in many plants (rogalla et al. 2002; neumann et al. 2001; liu et al. 2009; ma et al.2008; wang et al. 2000). wang et al. have (2000) reported a reduction of cd uptake in rice with the application of si (wang et al. 2000). similarly, si mediated cd uptake has been observed in other plants such as strawberry, cucumber, and maize (nwachukwu et al. 2007; chiroma et al. 2014; wijewardena et al. 2004). table 3 shows that the fe content in different plant tissues are significantly different among two treatments of na2sio3 and the control, indicating differences in the capacity for fe uptake. increase in accumulation of fe was observed during the first month despite the na2sio3 treatment. it may be due to increase in accumulation of fe with time. however, after two months of na2sio3 application, fe accumulation was decreased. fe concentrations of leaves after two months of si applications (both in si-100mg/l and si-50 mg/l) are less than before treatments. for example, fe concentrations in hettiarachchi and weerahewa heavy metal uptake by field grown basella alba ruhuna journal of science 28 vol 7: 21-31, december 2016 leaves treated with si-100 mg/l before and after application are 55.02 ± 2.05 μg/g and 16.27 ± 2.30 μg/g respectively. further, table 3 shows that the application of higher concentration of si lowers the capacity for fe absorption. for example, fe concentrations of leaves treated with si – 100 mg/l before and after the application and the leaves treated with si-50 mg/l before and after the application are 55.02 ± 2.05 μg/g, 16.27 ± 2.30 μg/g respectively, and 24.13 ± 1.99, 13.27 ± 2.09 μg/g, respectively. there is no significant difference in leaves treated with si-0 mg/l (control) before and after the application. table 3. fe absorption capacities of different plant parts of b. alba after treating with different concentrations of na2sio3 at different time periods. fe concentration in plant samples (µg/g) na2sio3 (si-100 mg/l) na2sio3 (si 50 mg/l) control (si -0 mg/l) before spraying na2sio3 leaves 55.02 ± 2.05 a 24.13 ± 1.99 d 32.90 ± 2.26 g stem 67.96 ± 2.01 b 60.95 ± 2.15 e 69.04 ± 2.30 h root 394.73 ± 2.11 c 233.24 ± 2.19 f 198.53 ± 1.87 i 1 month after spraying na2sio3 leaves 114.33 ± 1.77 a 74.21 ± 2.79 d 37.62 ± 2.70 g stem 363.73 ± 1.77 b 133.93 ± 2.82 e 123.56 ± 3.16 h root 472.32 ± 1.77 c 545.04 ± 2.81 f 462.57 ± 2.75 i 2 months after spraying na2sio3 leaves 16.27 ± 2.30 a 13.27 ± 2.09 d 34.54 ± 2.33 g stem 84.89 ± 2.31 b 129.91 ± 2.35 e 82.46 ± 2.53 h root 285.22 ± 2.14 c 741.99 ± 2.16 f 341.49 ± 2.29 i the values which share the same superscript letters have no significant (p<0.05) difference. comparisons were made separately for different spraying stages. our results show a significant difference in fe accumulation among different plant tissues including stem, leaves and roots collected from the plants exposed to different treatments. our results also show that fe accumulation in different plant parts is significantly different (p<0.05) within treatments. when comparing the accumulation of fe in a particular plant part within hettiarachchi and weerahewa heavy metal uptake by field grown basella alba ruhuna journal of science 29 vol 7: 21-31, december 2016 different treatments, significant differences were noted in all the plant parts. also, when comparing the accumulation of fe among different treatments under different time periods a significant difference was shown in all the plant parts. we document that the highest fe concentrations are found in the roots and the lowest concentrations are found in the leaves in all three treatments under all three application stages. 4 conclusion concentrations of fe, cr, pb and cd in the soils tested were 29755.30 ± 292.02, 32.99 ± 0.97, 26.01 ± 1.02, 0.13 ± 0.004 µg/g, respectively. concentration of cd in the tested soil is much less, compared to other heavy metals. all these concentrations are below the maximum permissible limit reported by fao/who. therefore, the soil samples from our study site were not heavily contaminated by the studied heavy metals. among the metals studied, only fe was absorbed by basella alba. transfer factors calculated for fe among different parts of the plant:soil revealed that translocation of fe is different among individual parts of the plant. our experimental results (we had to limit up to three replicates due to high cost of analysis) also showed that application of na2sio3 reduces fe absorption capacity of b. alba. there was a significant difference in fe accumulation in different plant tissue treated with different treatments of na2sio3. fe accumulation in different plant parts was significantly different within treatments and among different treatments from three different application stages. between the two concentrations of na2sio3 used, si-100 mg/l reduced the absorption capacity more than that of si-50 mg/l. therefore, we can conclude that fe absorption in field grown b. alba can be decreased by treating the plants with si-100 mg/l of na2sio3 than treating with si-50mg/l of na2sio3. acknowledgements this work was funded by the faculty research grant – 2015, faculty of natural sciences, the open university of sri lanka. authors would like to acknowledge ms. d. s. w. wickramasinghe, undergraduate student at the open university of sri lanka for the support during the research project and dr. m. n kaumal, senior lecturer, department of chemistry, university of colombo, sri lanka for the assistance given in sample analysis. hettiarachchi and weerahewa heavy metal uptake by field grown basella alba ruhuna journal of science 30 vol 7: 21-31, december 2016 references abah j, mashebe p, ubwa st, denuga dd. 2014. some heavy metals content of cabbage and soil cultivated in the bezi bar farm area of katima mulilo, namibia. american journal of chemistry, 4, no.3, 101-108. bolan ns, szogi t, chusavathi b, seshadri m, rothrock jr j, panneerselvan p. 2010. uses and management of poultry litter world poultry science. journal, 66, 673-698. chen y, wu p, ying y. 2014. health risk assessment of heavy metals in vegetables grown around battery production area. scientia agricola, 71(2), 126-132. chiroma tm, ebewele ro, hymore fk. 2014. comparative assessment of heavy metal levels in soil, vegetables and urban grey waste water used for irrigation in yola and kano. international refereed journal of engineering and science, 3 (2), 01-09. ding tp, zhou jx, wan df, chen zy, wang cy. zhang f. 2008. silicon isotope fractionation in bamboo and its significance to the biogeochemical cycle of silicon, geochimica et cosmochimica acta 72, 1381-1395. epstein e. 1994. the anomaly of silicon in plant biology. proceedings of the national academy of science, usa, 91, 11-17. gal l. 2015. examining the role of environmental change on emerging infectious diseases and pandemic. a volume in the advances in human services and public health (ahsph) book series. gupta s, pandotra p, gupta ap, dhar jr, sharma g, ram g, husain mk, bedi ys. 2010. volatile (as and hg) and non-volatile (pb and cd) toxic heavy metals analysis in rhizome of zingiber officinale collected from different locations of north western himalayas by atomic absorption spectroscopy. food and chemical toxicology, 48, 2966-2971. jayasinghe bc , jayawardena mifp, pathiratne kas. 2005. comparison of chromium levels of water spinach samples with those of field water samples. proceedings of the 9 th international conference on environmental science and technology, rhodes island, greece. jayathilaka j, bandara jmrs. 1989. pesticide management by hill country vegetable farmers. tropical agricultural research, 1, 121-131. kananke t, wansapala j, gunaratne a. 2014. heavy metal contamination in green leafy vegetables collected from selected market sites of piliyandala area, colombo district, sri lanka. american journal of food science and technology, 2.5, 139-144. liu c, li f, luo c, liu x, wang s, liu t, li x. 2009. foliar application of two silica sols reduced cadmium accumulation in rice grains journal of hazardous materials, 161, 1466-1472 lux, a, luxova m, abe j, tanimoto e, hattori t, inanaga s. 2003. the dynamics of silicon deposition in the sorghum root endodermis, new phytologist. 158, 437-441. ma jf, takahashi e. 2002. soil, fertilizer and plant silicon research in japan, elsevier science, amsterdam. ma jf, yamaji n, mitani n, xu xx, su yh, mcgrath sp, zhao fj. 2008. transporters of arsenite in rice and their role in arsenic accumulation in rice grain. proceedings of the national academy sciences of the united states of america., 105, 9931-9935. mclaughlin mj, hamon re, mclaren rg, speir tw, rogers sl. 2000. review: a bioavailability-based rationale for controlling metal and metalloid contamination of agricultural land in australia and new zealand. australian journal of soil research 38(6), 1037-1086. http://www.ncbi.nlm.nih.gov/pubmed/18626020 http://www.ncbi.nlm.nih.gov/pubmed/18626020 hettiarachchi and weerahewa heavy metal uptake by field grown basella alba ruhuna journal of science 31 vol 7: 21-31, december 2016 neilson s, rajakaruna n. 2014. phytoremediation of agricultural soils: using plants to clean metal-contaminated arable lands. in: phytoremediation: management of environmental contaminants, pp. 159-168. neumann d, nieden u. 2001. silicon and heavy metal tolerance of higher plants. silicon and heavy metal tolerance of higher plants. phytochemistry 56, 685-692. ponzi r, pizzolongo p. 2003. morphology and distribution of epidermal phytoliths in triticum aestivum l. plant biosystems. 137, 3-10. premarathna hmpl, hettiarachchi gm, indraratne sp. 2005. accumulation of cadmium in intensive vegetable growing soils in the up country. tropical agricultural research, 17, 93-103. premarathne hmpl, hettiarachchi gm, indraratne sp. 2011. trace metal concentration in crops and soils collected from intensively cultivated areas of sri lanka. pedologist. 54(3), 230-240. rajapakshe rmcp, amarakoon id. 2011. response of lettuce and rhizosphere biota to successive additions of zinc and cadmium to a tropical entisol. communications in soil science and plant analysis, 42, 1336-1348. rathnayake rmps, premarathna hmpl, ariyaratne gmh. 2004. heavy metal accumulation by some selected leafy vegetables grown in wellampitiya area, proceedings of the peradeniya university research. rogalla h, romheld v. 2002. role of leaf apoplast in silicon-mediated manganese tolerance of cucumis sativus l. plant cell and environment, 25, 549-555. sangster ag, hodson mj, tubb hj. 2001. silicon deposition in higher plants. in: datnoff le, snyder gh, korndorfer gh (eds) silicon in agriculture, studies in plant science, elsevier, amsterdam, 8, 85-113. sharma rk, agrawal m, marshall fm. 2009. heavy metals in vegetables collected from production and market sites of a tropical urban area of india, food and chemical toxicology, 47, 583-591. wang lj, wang yh, chen q, cao wd, li m, zhang fs. 2000. silicon induced cadmium tolerance of rice seedlings. journal of plant nutrition. 23, 1397–1406. wijewardena jdh, gunarathna sp. 2004. annals of the sri lanka department of agriculture, 6, 245-253. http://www.springer.com/life+sciences/ecology/book/978-3-319-10394-5 http://www.springer.com/life+sciences/ecology/book/978-3-319-10394-5 sv-lncs ruhuna journal of science vol 8: 6775, june 2017 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v8i1.27 university of ruhuna  faculty of science, university of ruhuna sri lanka 67 short paper a simple algorithm for calculating the area of an arbitrary polygon k.r. wijeweera1, 2 and s.r. kodituwakku2, 3 1department of computer science, faculty of science, university of ruhuna, sri lanka 2postgraduate institute of science, university of peradeniya, sri lanka. 3department of statistics and computer science, faculty of science university of peradeniya correpondence: krw19870829@gmail.com received: october 28th 2016, revised: june 15th 2017, accepted: june 30th 2017 abstract. computing the area of an arbitrary polygon is a popular problem in pure mathematics. the two methods used are shoelace method (sm) and orthogonal trapezoids method (otm). in otm, the polygon is partitioned into trapezoids by drawing either horizontal or vertical lines through its vertices. the area of each trapezoid is computed and the resultant areas are added up. in sm, a formula which is a generalization of green’s theorem for the discrete case is used. the most of the available systems is based on sm. since an algorithm for otm is not available in literature, this paper proposes an algorithm for otm along with efficient implementation. conversion of a pure mathematical method into an efficient computer program is not straightforward. in order to reduce the run time, minimal computation needs to be achieved. handling of indeterminate forms and special cases separately can support this. on the other hand, precision error should also be avoided. salient feature of the proposed algorithm is that it successfully handles these situations achieving minimum run time. experimental results of the proposed method are compared against that of the existing algorithm. however, the proposed algorithm suggests a way to partition a polygon into orthogonal trapezoids which is not an easy task. additionally, the proposed algorithm uses only basic mathematical concepts while the green’s theorem uses complicated mathematical concepts. the proposed algorithm can be used when the simplicity is important than the speed. keywords. computational geometry, computer graphics programming, coordinate geometry, euclidian geometry, computer programming. 1 introduction a polygon is defined as the region of a plane bounded by a finite collection of line segments which forms a simple closed curve. let v0, v1, v2, …, vn-1 be n points in the plane. here and throughout the paper, all index arithmetic will be wijeweera and kodituwakku calculating the area of an arbitrary polygon ruhuna journal of science 68 vol 8: 67-75, june 2017 mod n, conveying a cyclic ordering of the points, with v0 following vn-1, since (n 1) + 1 ≡ 0 (mod n). let e0 = v0v1, e1 = v1v2, …, ei = vivi+1, …, en-1 = vn-1v0 be n segments connecting the points. then these segments bound a polygon if and only if 1) the intersection of each pair of segments adjacent in the cyclic ordering is the single point shared between them: ei ∩ ei+1 = vi+1, for all i = 0, 1, 2, ..., n – 1. 2) non adjacent segments do not intersect: ei ∩ ej = ø, for all j ≠ i + 1. 3) none of three consecutive points vi are collinear. these segments define a curve since they are connected end to end and the curve is closed since they form a cycle and the curve is simple since non adjacent segments do not intersect. the points vi are called the vertices of the polygon while the segments ei are called its edges. therefore a polygon of n vertices has n edges (o’rourke 1998). in the proposed algorithm, the polygon is separated into orthogonal trapezoids by drawing horizontal lines through each vertex of the polygon. the area of each trapezoid is computed and added up to find the area of the entire polygon. 2 methodology this section describes the proposed algorithm and its implementation. the c programming language has been used for the implementation. 2.1 representation of the polygon the polygon is represented using two arrays x and y. the points variable stores the number of vertices in the polygon. (x[i], y[i]) denotes the coordinates of the ith vertex where i = 0, 1, …, (points – 1). 2.2 drawing horizontal lines through each vertex of the polygon the polygon is separated into orthogonal trapezoids by drawing horizontal lines through each vertex of the polygon. this is done by using findyg function. first each y-coordinate is stored in yg array. then sortarray function is used to sort those y-coordinates in ascending order. wijeweera and kodituwakku calculating the area of an arbitrary polygon ruhuna journal of science vol 8: 67-75, june 2017 69 2.3 the refineyg function each horizontal line drawn in section 2.2 may go through more than one vertex in the polygon. in those cases yg array contains duplicate values. the refineyg function is used to remove those duplicate values from the yg array. using a for loop, initially the yg array is copied to yh array. after that only the distinct values are written to the yg array. the ygn variable will finally contain the number of distinct elements. since the yg array is already sorted in ascending order, duplicate values are always in consecutive cells. the first element of yg array is set into first element of yh array by yg[0] = yh[0] and now ygn is 1. using another for loop, values of yh are copied one by one to yg only if current value is not equal to the previous. the condition (yh[i 1] != yh[i]) is used for this purpose. 2.4 the pos function the pos function is used to decide the position of a given vertex (x[v], y[v]) with respect to a horizontal line y = yg[u]. if the vertex is on the horizontal line then the function returns zero. if the vertex is above the horizontal line then it returns 1 and if the vertex is below the horizontal line it returns -1. 2.5 the findgaps function the findgaps function takes u as a parameter which denotes the indices of yg array. a horizontal line drawn in section 2.2 may intersect edges of the polygon as shown in figure 1. two consecutive horizontal lines bound a set of orthogonal trapezoids. in order to compute the area of them, coordinates of the end points should be calculated. for a given two consecutive horizontal lines, the intersection points of upper and lower horizontal lines differ depending on the special fig. 1. horizontal lines through vertices wijeweera and kodituwakku calculating the area of an arbitrary polygon ruhuna journal of science 70 vol 8: 67-75, june 2017 cases as described in this section later. therefore x-coordinates of the intersection points corresponding to lower horizontal line are stored in gap1 array and that of upper horizontal line are stored in gap2 array each iteration. the n1 and n2 variables will contain the number of elements in gap1 and gap2 arrays respectively. the purpose of first for loop found in findgaps function is to deal with intersections of each edge with the given horizontal line y = yg[u]. the end points of each edge are (x[i], y[i]) and (x[j], y[j]) where i = 0, 1…, (n 1) and j = (i + 1) % points. the pi and pj variables decide the positions of the end points using pos function. depending on the way edges intersect with the horizontal line, there are five possible situations as shown in figure 2. from the left each possible situation is names as general case, down to down case, up to up case, down to up case, and up to down case respectively. in the figure, 1 and 2 numbers denote whether that intersection point considered for gap1 array or gap2 array respectively. following subsections describe how to deal with each case. general case if (pi * pj < 0) then end points are on opposite sides of the horizontal line. the intersection point is (xc, yg[u]). and the equation of the edge can be written as, (y – y[i]) / (x – x[i]) = (y[j] – y[i]) / (x[j] – x[i]); x = (x[j] – x[i]) * (y – y[i]) / (y[j] – y[i]) + x[i]; since the intersection point is on y = yg[u] line, xc = (x[j] – x[i]) * (yg[u] – y[i]) / (y[j] – y[i]) + x[i]; this intersection point should be included to both gap1 and gap2 arrays. fig. 2. possible intersections of edges wijeweera and kodituwakku calculating the area of an arbitrary polygon ruhuna journal of science vol 8: 67-75, june 2017 71 down to down case if (pi = pj = 0) then the entire edge goes through the horizontal line as in second to fifth situations in figure 2. the values of a and b are computed by a = i – 1 and b = j + 1. they denote the indices of neighboring vertices of the end points of the edge. when i = 0 then a = -1, but actually it should be (points 1). when j = (points 1) then b = points, but actually it should be 0. these two special cases should be handled. the pa and pb variables store the positions of a and b vertices respectively. if pa * pb > 0 then it should be either second or third case in figure 2. if pa < 0, it should be definitely the second situation. in this case both the end points should be included to gap2 array. up to up case if pa * pb > 0 and pa > 0, it is the third situation. in this case both the end points should be included to gap1 array. down to up case if pa * pb < 0 and pa < 0 then it is the fourth situation. in this case i end point should be included to gap2 array and j end point should be included to gap1 array. up to down case if pa * pb < 0 and pa > 0 then it is the fifth situation. in this case i end point should be included to gap1 array and j end point should be included to gap2 array. depending on the way vertices intersect with the horizontal line, there are three possible ways as shown in figure 3. the first two situations are ignored since they do not affect the area between given two horizontal lines. the second for loop is used to deal with vertex fig. 3. possible intersections of vertices wijeweera and kodituwakku calculating the area of an arbitrary polygon ruhuna journal of science 72 vol 8: 67-75, june 2017 intersections with the horizontal line. the vertices on the horizontal line are found by checking the condition y[i] = yg[u]. the pa and pb variables are found as in previous situation. if (pa * pb < 0), it is the third situation. in this situation the intersection vertex should be included to both gap1 and gap2 arrays. 2.6 the findsums function there are ygn number of horizontal lines which can be drawn through vertices of the polygon as shown in figure 1. each horizontal line has two arrays gap1 and gap2. inside the for loop, initially n1 and n2 are set into zero. then gap1 and gap2 array values are found using findgaps function for each horizontal line. the sortarray function is used to sort those two arrays in ascending order. 2.7 the computesum function figure 4 shows an example diagram of a horizontal line. the numbers indicate the indices of the gap array in computesum function. the thick line segments on the horizontal line shows the intersection regions with the polygon. the purpose of computesum function is to calculate the sum of the lengths of the thick line segments and store it in sum[index] element of sum array. the value of sum[index] is set into zero initially. then the elements of gap array are added or subtracted from the sum[index] depending on whether their indices are even or odd respectively. in each iteration of the for loop inside findsums function, the computesum function is invoked for gap1 and gap2 arrays. the sum1 and sum2 arrays correspond to gap1 and gap2 arrays respectively. 2.8 the getareas function figure 5 shows an example of polygon parts bounded by two consecutive horizontal lines drawn through vertices of polygon. the sum1[i] stores the sum of lengths of thick line segments corresponds to the lower horizontal line. and sum2[i + 1] stores the sum of lengths of thick line segments corresponds to the upper horizontal line. fig. 4. intersection regions example wijeweera and kodituwakku calculating the area of an arbitrary polygon ruhuna journal of science vol 8: 67-75, june 2017 73 the area variable in getareas function stores the area bounded by ith couple of consecutive horizontal lines. the gap between the two horizontal lines is (yg[i + 1] – yg[i]). therefore the area bounded by the two horizontal lines can be computed by 0.5 * (yg[i + 1] – yg[i]) * (sum1[i] + sum2[i + 1]). this formula is similar to the area formula for a trapezoid and it can be proved as follows. proof: let the gap between two horizontal lines is h and n be the number of trapezoids. then the height, lower base and upper base of the trapezoids are (h, a1, b1), (h, a2, b2)…, and (h, an, bn) respectively. sum of areas of trapezoids, = (h/2) * (a1 + b1) + (h/2) * (a2 + b2) + … + (h/2) * (an + bn) = (h/2) * [(a1 + a2 + … + an) + (b1 + b2 + … + bn)] 2.9 the getarea function the area variable inside getarea function will store the total area of the polygon. the for loop accesses each area bounded between consecutive couple of horizontal lines and sum them up. 3 results and discussion the algorithm was implemented using c programming language. following hardware and software were used. computer: intel(r) celeron(r) m; processor 1.50 ghz, 896 mb of ram; ide: turbo c++; version 3.0; copyright(c) 1990, 1992 by borland international, inc; system: microsoft windows xp professional; version 2002; service pack 2. to validate the proposed algorithm, it was compared with an implementation of shoelace method (wijeweera 2015, o’rourke 1998). a set of random polygons as shown in table 1 were used. fig. 5. polygon parts between two horizontal lines wijeweera and kodituwakku calculating the area of an arbitrary polygon ruhuna journal of science 74 vol 8: 67-75, june 2017 table 1. set of random polygons polygon coordinates of vertices 1 (10, 10), (110, 210), (90, 80), (250, 60) 2 (60, 240), (10, 190), (50, 20), (110, 150), (150, 190) 3 (140, 90), (200, 240), (40, 190), (70, 10), (280, 80), (210, 130) 4 (60, 60), (300, 130), (240, 220), (100, 250), (20, 200), (40, 20), (260, 70) 5 (100, 240), (130, 80), (270, 210), (190, 20), (150, 60), (90, 10), (10, 80), (80, 100) 6 (150, 40), (50, 10), (10, 190), (120, 250), (230, 220), (130, 90), (240, 160), (240, 30), (50, 120) 7 (140, 90), (240, 60), (140, 240), (250, 170), (290, 10), (30, 10), (10, 120), (130, 220), (80, 70), (170, 140) 8 (230, 40), (30, 10), (80, 70), (10, 90), (70, 90), (150, 140), (150, 60), (190, 140), (80, 250), (210, 200), (280, 10) 9 (90, 90), (150, 160), (130, 220), (240, 70), (250, 10), (160, 110), (110, 30), (40, 10), (20, 130), (90, 250), (130, 160), (90, 190) 10 (90, 70), (60, 150), (10, 20), (60, 190), (90, 150), (140, 240), (190, 130), (230, 170), (270, 60), (190, 10), (230, 100), (140, 30), (160, 150) the number of clock cycles to compute the area of a polygon is not measurable since the value is too small. therefore the number of clock cycles to compute the area of the same polygon 108 times was measured (kodituwakku et al. 2013). this was done for each polygon in table 1 using orthogonal trapezoid method (otm) and shoelace method (sm). the results are shown in table 2. table 2. set of random polygons polygon otm slm 1 4554 204 2 5345 241 3 9914 291 4 11641 327 5 14408 361 6 20791 406 7 28416 453 8 21939 466 9 33098 490 10 41047 532 according to the results the efficiency of the proposed method is lower than the existing method. 4 conclusion an algorithm to computerize orthogonal trapezoid method was proposed. and it was experimentally compared against shoelace method. the area of wijeweera and kodituwakku calculating the area of an arbitrary polygon ruhuna journal of science vol 8: 67-75, june 2017 75 the polygon was computed by decomposing the polygon into a set of trapezoids. the decomposition was not a trivial task. currently triangular polygonal meshes are used in computer graphics programming to model surfaces (hearn and baker 1998). the proposed decomposition technique can be to generate trapezoidal polygonal meshes. references hearn d, baker mp. 1998. computer graphics, c version, 2nd edition, prentice hall, inc., upper saddle river, pp. 305-309. kodituwakku sr, wijeweera kr, chamikara map. 2013. an efficient algorithm for line clipping in computer graphics programming, ceylon journal of science (physical sciences), volume 17: 1-7. o’rourke j. 1998. computational geometry in c: 2nd edition, cambridge university press 1-22. wijeweera kr. 2015. finding the area of an arbitrary polygon: shoelace formula and its implementation in c programming language. retrieved from http://www.academia.edu/9987996/finding_the_area_of_an_arbitrary_polyg on_shoelace_formula_and_its_implementation_in_c_programming_languag e. appendix program code is supplied as a supplementary file (appendix, pp i-iv) linked to wijeweera and kodituwakku (2017). ruhuna journal of science 8 (1): 67-75. doi: http://doi.org/10.4038/ rjs.v8i1.27 ruhuna journal of science vol 12 (2): 64-83, december 2021 eissn: 2536-8400 © faculty of science http://doi.org/10.4038/rjs.v12i2.103 university of ruhuna © faculty of science, university of ruhuna sri lanka 64 exploring marketing channels and market margins of tuna species: a case study of negombo fishery harbour in sri lanka in 2018 kamal edirisinghe1, jagath wansapala2, indira wickramasinghe2 and a.s.k. warahena3 1department of agriculture & food technology, university of vocational technology, ratmalana, sri lanka 2 department of food science and technology, faculty of applied sciences, university of sri jayewardenepura, gangodawila, nugegoda, sri lanka 3department of manufacturing technology, university of vocational technology, ratmalana, sri lanka *correspondence: sagarakamal@gmail.com; orcid: https://orcid.org/0000-0002-3312-534x received: 6th september 2020; revised: 8th july 2021; accepted: 15th december 2021 abstract this study was carried out to determine the domestic level marketing channels, marketing cost, and marketing margin for four commercially important marine fish species in negombo fishery harbour, western province, sri lanka, with a focus on marketing efficiency. data related to fish catch and prices were collected by direct observation during the period of january to december 2018. about 80 to 105 tons of fish catch per day were being landed. the annual average wholesale prices of fish (kg-1)g varied from lkr 450 to 670 for yellowfin tuna, lkr 233 to 414 for skipjack tuna, lkr 320 to 385 for frigate tuna, and lkr 140 to 190 for indian scad, which is typical market behaviour. though the net profit per 100 kg of fish received by the wholesaler was higher than the assembler, the retailer has made a significant amount of profit, nearly six times higher than the assembler. ten marketing channels were identified from fish producers to the ultimate consumer and some channels involved more than three intermediaries. as typical for long marketing channels, high prices were paid by the consumer for the low freshness quality of fish. though, the retail price of skipjack tuna at the negombo landing center is moderately correlated (r=0.634), the other three fish species did not give a clear indication. thus, the price factor of selected marine fish at the landing site was not having a significant impact on the market integration. consumers were concerned about both price and quality when purchasing the fish. to improve the marketing efficiency of the marine fish market, it is recommended to implement shortening of the fish market channel with less involvement of intermediaries and to develop the fish market facilities appropriately. keywords: consumer preference, fish price, market efficiency, negombo harbour. 1 introduction the fisheries sector of sri lanka plays an important role in the national economy generating considerable income with its contribution to gross domestic product https://rjs.ruh.ac.lk/index.php/rjs/index https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v12i2.103 https://creativecommons.org/licenses/by-nc/4.0/ mailto:sagarakamal@gmail.com https://orcid.org/0000-0002-3312-534x k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 65 (gdp) in 2018 of approximately 1.2% (mfard 2018). sri lanka has 15 fishing grounds and 22 fishing ports. most fishing ports have a set of facilities that can be used as an interface between fishing and consumption (mfard 2019). total fish production of sri lanka in 2018 was 527,060 metric tons, of which total marine fish production was 439,370 metric tons, whereas the remainder (87,690 metric tons) was attributed to inland fisheries and aquaculture (table 1). marine offshore fish production shows some growth over the years 2013-2018, whereas tuna fishing in this sector also shows growth in harvest, export and processing. tuna fishery is the mainly focused marine fish of the export market with high commercial value (yamao and de silva 2006). table 1: annual fish production in sri lanka by sub-sectors (mt). sector 2013 2014 2015 2016 2017 2018 marine coastal 267,980 278,850 269,020 274,160 259,720 249,020 marine offshore/deep sea 177,950 180,450 183,870 182,830 189,720 190,350 inland & aquaculture 66,910 180,450 183,870 73,930 81,870 87,690 total 512,840 535,050 520,190 530,920 531,310 527,060 (source: mfard 2019) according to ‘sri lanka fishery development strategy’, the average daily fish consumption per capita is sufficient to ensure adequate nutrition (mri 2010). fish products are an important source of animal protein accounting for about 70% of the country’s animal protein intake (food balance sheet 2017). currently, the average supply of seafood is 668,000 tons, and the average supply of seafood per capita is 32 kg per capita per year (food balance sheet 2017). livelihoods for most people in coastal areas are provided within the fisheries sector, which currently provides direct employment for approximately 650,000, including 150,000 in fisheries, 10,000 in related service activities and 400,000 in fish trade (weerasekara et al. 2015, mfar 2019). fish prices are highly dependent on fish quality, fish species and size, supply/ demand, market distance, weather/ climate change, and lack of long-term sales channels (hossain et al. 2015, perera et al. 2016). therefore, typical characteristics of fish supply chains include uncertainty, such as assembling fish from many fish landings ashore, the commodity and demand patterns of different types of fish, numerous marketing channels and intermediaries, and price fluctuations (alhassan et al. 2012, aswathy and abdussamad 2013, begum et al. 2014). currently, a business transaction can be done through fish markets, retail markets, fishing ports or online ordering. the scope of services at a fishing port has been expanded to provide for loading and unloading services, and logistics (ice, fuel, water, etc.) for fishing vessels. some studies on seafood sales include marketing channels (madugu and edward 2011, alhassan et al. 2012, rabby et al. 2015), sales patterns and trends (abila 1998), and sales systems (flowra et al. 2012, akhtar et al. 2013, jamali 2013), k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 66 economic analysis of fresh fish sales (ali et al. 2008), marketing management (sathiadhas and kanagam 2000), market intermediary and marketing profit (hussain et al. 2003), typical transportation systems used (rokeya et al. 1997) and marketing strategy (mutambuk 2014, hossain et al. 2015). distribution channels in consumer marketing including fish products play a heavy role in the flow of goods and services from producers to consumers (gorchels 2004, gorchels et al. 2004). even for fish products, the distribution channels consist of one or more organizations or individuals who participate in the process of goods, services, information and finance from the consumers or producers (coyle et al. 2003). sri lankan tuna species have the market popularity as a delicacy and a highly nutrient seafood in both local and foreign markets. tuna exports as fresh and processed foods have a sizeable market share for sri lanka in the world market. the export market is relatively intensive, and as a result, good quality tuna became expensive commodity in sri lanka. these facts reflect the socio-economic sensitivity and importance of tuna as a marine product. studying and exploring marketing channels and the market margin of tuna may provide directions to improve the socioeconomic advantage related to tuna fish industry and the marketplace. in the marine offshore sector, there is a consistent and slower growth in terms of tonnages of harvest (dissanayake and sigurdsson 2005). here also, the major commodity is the tuna species. the consistency in growth is most likely driven by tuna demand, while the slow growth is due to challenges in marketing channels and market margins. this study focuses to identify the marketing channel pattern of the marine fishery based on the negombo fish landing site (7.2040° n, 79.8277° e), marketing margin for selected tuna fishes, and how price impacts the market integration and consumer’s role of the marketing channel with reference to the domestic level of the marketing. 2 material and methods 2.1 study sites negombo fishery harbour is located in the gampaha district of the western province of sri lanka (7.2040° n, 79.8277° e). it was selected as the fish landing site on focus as it is one of the main fishery harbours that supply food fish in the province. the average annual fish supply from negombo is around 41,000 metric tons (ariyawansa et al. 2016). peliyagoda fish economic center (7.0147° n, 79.8997° e) at the suburbs of colombo city is chosen as the marketing center as it directly connects with the negombo fishery harbour landing site, and it is the largest for fish sales catering for colombo as well as for other inland areas, including mainly the central, sabaragamuwa and north-western provinces. at the start of the marketing channel, the negombo harbour and the subsequent peliyagoda central fish market were identified as the first two elements. beyond that, k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 67 till the end-user, there can be variants in marketing channel aspects. some channels may have more intermediatory elements. in the negombo fish landing, mostly large fish supplied by multi-day and one-day boats are landed at fishery harbour landing center. other fishing crafts are landed at the shore in negombo and reach the main landing area. table 2: sampling frame related to the marketers of negombo landing site and peliyagoda fish economic center, sri lanka (total 42). in marketing channels, though the first two elements are common, there are quality, quantitative, value, and timing variations along the channels. because some channels may have more quantities of fish flow (vs. less), take longer time (vs. shorter), carry more value (vs. less value), have better-assured freshness (vs variable freshness), more marketing chain elements were considered at the starting point as shown in table 2. the period of data collection was one year (january-december 2018). 2.2 selecting consumers for the study thirty consumers were randomly selected by representing each market during marketing time. information was collected from the consumers who purchased fish from suburbs of colombo representing the associated market areas of negombo fish market (n=5), central peliyagoda fish market (n=5), ceylon fisheries corporation (cfc) outlet (n=5), supermarket (n=5), retail stall (n=5) and fish vendors (n=5). 2.3 analyzing prices of selected marine fish species considering commercial value, fishing gear, availability throughout the year and marketing demand, this study selected yellowfin tuna (thunnus albacares), skipjack tuna (katsuwonus pelamis), frigate tuna (auxis thazard), and indian scad (decapterus russelli). fish prices from producer to consumer end were collected by secondary data of relevant authorities, by discussing with stakeholders and onsite visits. marketing chain element sampling number real-time total offshore fisherman 05 25 one day fishermen 05 around 50 assemblers/beach collectors 05 around 10 at a time wholesalers of negombo fishery harbour 05 around 10 wholesalers of peliyagoda central fish market 05 around 10 retail retailers 05 around 25 fish vendors 05 around 45 sales outlets of the supermarket chain 05 around 08 online traders 02 only 2 available k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 68 when analyzing marketing margin, the fish price of mean values of all selected fish species were clustered together and average values were calculated. average retail prices were calculated from mean values of all categories such as supermarket retailers, retail outlets, and fish vendors. 2.4 analysis of marketing channels due to a lack of previous studies and knowledge about activities within the logistic chain of sri lankan fisheries, primary data was gathered from stakeholders involved directly and indirectly in the fisheries logistic chain, i.e., fishermen, traders, processors, government officials and other informed groups. they were interviewed by on-site visits to negombo harbour, wholesale markets (negombo & peliyagoda) and retail markets. secondary data were obtained from organizations associated with fishery industry, community-level organizations, and the ministry of fisheries. structured interviews were used with groups with an adequate number of participants, such as assemblers, export assemblers, wholesalers/commission agents and retailers. table 3: key factors considered for designing framework to analyse fish marketing channels. key factor expected output element the structure of the value chain and the main actors and activity this is focused to identify the structure of the value chain and the participants and activities of the system a. the value addition in the value chain b. key actors c. main activity relationship within the value chain this may include a description of the product's capacity, quantity, and price, including the value-added in the value chain. a. collaboration between actors b. flow of information and knowledge c. power and trust d. governance the strategic position within the value chain this may include relationships within the value chain and how the collaboration between the actors a. driving forces within the industry b. competitive advantage source: gestsson et al. 2010 the questionnaire framework in this study was designed and used based on the value chain concept (porter 1985, segetlija et al. 2011, gestsson et al. 2010, perera et al. 2016) to investigate three key factors (table 3) affecting sri lankan marine fishery within the domestic (local) value chain. the number of fish trade retailers is spread all over the country and can be divided into several categories such as supermarkets, retail outlets and fish vendors (including motor-bike retailers, bicycle retailers and food peddlers). details of the quantity and prices of the fish caught were gathered from relevant organizations, such as the department of fisheries, national aquatic resources research and k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 69 development agency (nara), sri lanka custom and fishermen’s federation. for analyzing the value chain, prices were considered as an important indicator, more specifically how sellers and buyers meet and agree on quality and prices of fishes were indicated. all prices used for the study are in us dollars (usd) and data in sri lankan rupees (lkr) have been converted to usd, using the exchange rate in july 2018. 2.5 data collection purposive and simple convenient sampling techniques were used to collect the necessary data. structured interview schedules were used to collect information from five markets, namely, the major negombo fish landing center, wholesale, central fish market, supermarket, and retail fish markets through key informant surveys and focused group discussions with officials and members of fish trader associations/ cooperative societies. this was done after verifying the source of the fish from negombo fishery harbour by interviewing the marketer. 2.6 data analysis the average prices for selected fish species were computed for the period of study. the marketing margin at fishermen and consumers was calculated as per the formula described by (rahman et al. 2012), as marketing margin (%) = (selling pricepurchase price) x 100 selling price the cost-profit and marketing efficiencies of different marketing channels were also observed. descriptive statistics were analyzed using spss 23.00 and ms excel 2010. 3 results 3.1 status of the fish landing center and marketing destination table 4 shows the general information about different places in the fish market chain. the average operating duration of the fish auction in negombo fishing harbour is 5 hours, starting from 1.00 am to 6.00 am. at the auction market, all products were sold locally without weighing as fresh (without chilling). the traditional sales system in the negombo landing center is to sell whole fish in bulk without weighing. k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 70 table 4: general information about the fish market chain studied in negombo and colombo suburb areas. market characteristics negombo fishery habour central fish market (peliyagoda) super market (private) ceylon fisheries corporation (cfc) stall stall 01 stall 02 auction market open market market type wholesale retail wholesale/ retail retail retail retail retail market time 1.00 am-6.00am 6.00 am -10.00 am 4.00 am-10.00 am 8.00 am-10.00 pm 9.00 am-4.00 pm 9.00 am3.00 pm 10.00 am –4.00 pm platform cemented with tile cemented/ damaged cemented with top cover tile chilled display cabinet chilled display cabinet cemented damaged tile cemented damaged tile roof (shade) cover building building building building building rooftop cover asbestos sheet cover drainage facility present present present present present present but poor poor electricity supply present present present present present present present icing facility present but poor usage present but poor usage present but poor usage present chill storage condition to maintain present chill storage condition to maintain very poor not adequate ice very poor not adequate ice sanitation facility poor poor good good good very poor very poor water supply available available available available available available available hygienic practices very poor very poor very poor good good very poor very poor routine maintenance poor poor poor good good poor poor k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 71 about 100 ±12 tons of marine fish were being landed using the fishing boats/ vessels per day. two types of marketing systems have been identified in the negombo harbour area as auction market and open market. auctions usually sell all the caught fish in bulk, and fishermen place the caught fish on the floor alone instead of weighing it. each fish loading is usually composed of a mixture of species. the deepsea catch landings include much low-priced fish and some high-priced fish. when low-priced fish are mixed with high-priced fish, fishermen not only expect the price of low-priced fish to increase but also ensure timely disposal. due to the wide variety of precious fish such as seer fish (thora) and sail fish (thalapath) species, buyers are forced to buy the whole bulk of fish on auctions, but some studies have shown that ratings increase the overall fishing revenue. bidding is open and done in ascending order simply by verbally announcing the bids of all the approaching buyers for a certain fish lot. the highest bidder obtains the fish lot and that practice is allowed a price increase through trader competition. normally 1% of the total fish sales are reserved as auction charges for the registered people. payment is made at the point of sale and is guaranteed that the settlement is made at the bid price. though adequate infrastructure facilities are established in the negombo harbour, poor routine maintenance and inadequacy of hygienic practices seem to be problems also because of marketers’ improper use of the auction place. most of the fish were kept without adequate icing at the peer after unloading, before and after the auction affecting the freshness of fishes. the open market operation of the negombo fishery harbour is held for about four hours of average duration from 6.00 a.m. to 10.00 a.m. on regular landing days. the condition of the open market environment of the negombo fishing harbour is also the same as the auction site. considering the icing facility, only some of the marketers stored fish in the rigid foam boxes with ice. usage of lagoon water was observed for cleaning and washing purposes instead of pipeline (potable) water. the damages of the platform were observed in both auction and open market places with dull colours. five hours of average duration was noticed for market operation in the negombo fishery harbour, and central supermarket located closed to the colombo city. though all infrastructures are well-established at the central supermarket, usage of adequate icing facility, hygienic practices and routine maintenance have remained at a poor level. the sanitary facility of the central market was more satisfactory than the negombo. as a governmental trader, ceylon fisheries corporation (cfc) is actively engaged in the fish market chain distributing fish through their own sales outlets. around eight hours of marketing duration was noticed in cfc outlets, which is used with adequate icing. required infrastructure facilities were fulfilled in both supermarket outlets and cfc market outlets adhering to proper hygienic practices. both outlets were used to display fish in the chilled condition and the chilling temperature level was well maintained. except for these outlets, sanitation facilities, hygienic practices, routine maintenance (cleanliness), and usage of adequate ice were observed at a poor level in the other sales outlets. k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 72 3.2 marketing channels kohls and uhl (2005) described marketing channels/ value chains as the alternative routes of product flows from producers to consumers. in this study, ten marketing channels were identified in the flow of the selected fish market as given in figure 1. fig 1: identified fish marketing channels for marine fishes in negombo and colombo suburb areas. assembler is the first intermediator in the value chain and can have a variety of functions within the chain. retailers buy fish from auction markets, assemblers or wholesalers, and then resell them to consumers. the added value of the system is the purchase of large quantities of products from wholesalers. based on the surveyed data, the total marine fish catch from negombo fishery harbour is around 29,234 mt ±1,234 mt per annum. the catch was dispatched both to domestic and export markets, with the domestic fraction alluring around 59±6% of the marine fish production. in 2018, the export market accounted for 15±3% and the balanced percentage was used for dried fish production. the domestic fish market is k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 73 composed of a number of diverse end markets including urban wholesale fish markets, retailers, fish vendors, cfc and supermarket outlets. an online trading system is a newly added component to the value chain recently. fig 2: marketing channels (c1 to c10) of marine fish landings in the negombo fishery harbour this added value is mainly due to selection, transportation and sales. the general flow is from fishermen to auction markets, local assemblers, wholesalers and retailers. pricing is mainly based on the free marketing system of the auction market, with a few exceptions. the auction market has different market segments and different customer needs. the cost of these changes has been integrated into the price of fish, and is ultimately reflected in the price consumers are willing to pay. k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 74 the fisheries value chain is dominated by the private sector at all levels. the value chain map shown in figure 2 (the state-owned trader cfc) is a single channel that actually manages nearly 10% of the negombo-based marketing chain. obviously, with the dynamic changes in this sector, the structure of the fishery supply chain and the roles and responsibilities of its multiple stakeholders are changing rapidly. in emerging markets, some participants play multiple roles at different levels of the supply chain, and they are promoting system development. for instance, coastal operators engaged in deep-sea fisheries are shifting from local market suppliers to export market suppliers, which is particularly helpful in expanding the export processing industry. in some cases, the recent rise of fisheries cooperatives has created horizontal and vertical links, enhancing the bargaining power of small businesses, while keeping pace with industry trends. due to economic and practical barriers to entry into the fish trade, producers still cannot obtain fair prices for their products. however, cfc is stepping up its efforts to compete with the private fish trade, including buying fish at competitive prices through auctions with private merchants. scale is not important yet, but various types of cfc outlets are involved in the fish trade and various supermarket chains (such as cargills, keels, arpico) represent emerging markets. 3.3 prices of selected marine fish table 5: summary of average wholesale and retail price ranges of selected fish species in 2018 around negombo market chain. fish species wholesale price range lkr /kg retail price range lkr /kg price difference lkr /kg yellowfin tuna 450 to 670 (us$2.50 – 3.72) 722 to 1000 (us$4.01 – 5.55) 272-330 (1.51-1.83) 60% increased skipjack tuna 233 to 414 (us$1.29 2.30) 372 to 592 (us$2.01 – 3.29 133-178 (0.73-0.99) 57% increased frigate tuna 320 to 385 (us$2.13 2.13) 440 to 600 (us$2.44 – 3.33) 120-215 (0.31-1.2) 38% increased indian scad 140 to 190 (us$0.78 1.05) 200 to 250 (us$1.11 – 1.39) 60-60 (0.33-0.34) 42% increased (us$ 1= lkrs.180) prices of the fishes, namely yellowfin tuna skipjack tuna, frigate tuna, and indian scad varied as per the species, sizes, freshness, market demand, weather, and seasonal changes (table 5). the highest price difference (around 60% increased) was observed from yellowfin tuna. the least differences were recorded for frigate tuna and indian scad. figure 3 shows the monthly average prices of selected fish species in negombo wholesale market, central supermarket, retail outlets of supermarkets and retail k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 75 outlets in suburb areas in negombo without considering fish vendors (mobile) prices. the higher prices were indicated in may, june and july at the entire market chain, which may be the results of changing weather and rough sea conditions, thus the inadequate supply for the market, and then price factor goes up in the season. from september to november, the overall prices were considerably lower than in the other periods of the year. but in december, a recorded high price in the market was remained, which may be affected by seasonal demand. fig 3: monthly average prices of fish species in negombo and colombo suburb areas. a) yellowfin tuna, b) skipjack tuna, (c) frigate tuna (auxis thazard), and d) indian scad (decapterus russelli). average prices of the four fish species in the retail suburb areas (figure 4) were recorded higher than the other places over the year and the lowest value was observed in the negombo fish market. it was a fact that the fish price of retail suburb areas may add to the charges for handling, packing, icing, and transportation. k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 76 fig 4: average prices of selected marine fishes in negombo and colombo suburb areas. table 6 shows the summary of the average annual prices of the four fish species through different stages of the marketing channels. the retail prices of the supermarket outlets were higher than the prices of fish vendors for all fish varieties. that may include the employee wages and maintenance cost of the supermarket level. the fish prices at the consumer level were less than the vendors’ prices, which may happen due to the less involvement of intermediates of the marketing channel. some fish species like indian scad was not ordered online in this study period. table 6: average annual price (lkr per kg) of selected marine fish in different stages of marketing channel. mediator yellowfin tuna skipjack tuna frigate tuna indian scad producer 550.10 ± 50.50 275.15 ± 42.80 330.00 ±18.40 150.50 ±12.50 assembler 560.83 ± 67.20 297.67 ± 55.70 349.58 ± 20.50 167.50 ±15.70 wholesaler 798.42 ± 84.70 315.25 ± 48.4 354.17 ± 27.10 177.92 ±15.10 retailer super 883.83 ± 84.90 451.50 ± 60.30 587.33 ± 72.90 232.50 ±13.60 fish vendor 708.33 ± 77.00 415.83 ± 39.00 580.00 ± 60.60 215.83 ±11.40 online trader 710.00 ±83.10 409.17 ±36.00 523.33 ±72.60 not available 3.4 marketing cost and net marketing margin the cost of marketing a product refers to the expenses acquired by various sets of mediators in the process of achieving different marketing functions to reach the 0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 negombo wholesale central market retail super retail suburb a v e r a g e p r ic e ( l k r p e r k g ) market place yellow fin tuna skipjack tuna frigate tuna indian scad k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 77 product from the producer to the end-users. various components of fish marketing costs were recorded during this study period such as transportation cost, cost of fuels and lubricants, cost of storage and pack icing, cost of wastage, cost of fresh/ potable water, cost of other utilities, cost of wages and other miscellaneous expenditures. table 7 shows the average values of the marketing margin of lkr per 100 kilograms of all four species of fish within the study period. the marketing margin was computed (formula given by rahman et al. 2012) as the difference between the price obtained by the producers and the price paid by the consumers. net marketing margins consist of marketing costs and profits or losses earned by all intermediaries. the wholesalers (table 8) sold to the retailer gained a gross margin of lkr 6,800 (us$ 37.70) per 100 kgs and then, the net margin was received around lkr 4,800 (us$ 26.60) including marketing cost of lkr 2,000 (us$ 11.11) per 100 kgs. the retailer sold to the consumer to earn a gross margin of lkr 12,700 (us$ 70.50) and then the net margin reached up to lkr 7,200 (us$ 40.00) after deducting the marketing cost (lkr 5,500 us$ 30.50) per 100 kgs of fish. table 7: marketing margin of intermediaries (average price lkr per 100 kg) of fishes. intermediary purchase price (b) selling price (a) gross margin (c) =(ab) marketing cost (d) net margin (e) = (cd) assembler 32,600.00 us$ 181.10 34,300.00 us$ 195.50 1,700.00 us$ 9.44 500.00 us$ 2.70 1,200.00 us$ 6.60 wholesaler 34,300.00 us$ 195.50 41,100.00 us$ 6,800.00 us$ 37.70 2,000.00 us$ 11.10 4,800.00 us$ 26.60 retailer 41,100.00 us$ 228.30 53,800.00 us$ 298.80 12,700.00 us$ 70.50 5,500.00 us$ 30.50 7,200.00 us$ 40.00 3.5 analysis of retail prices the linear regression analysis was used to estimate the relationship between the prices of the negombo fish landing center and the retail price in the suburb area for the above-mentioned marine fishes. the strength of the relationship between the average weekly prices during 12 months period of fish landing center and retail outlet (figure 5) for the four selected high-value fishers (p<0.05 for rejecting ho of no relationship). weak relationship was indicated (figure 5a) for yellowfin tuna (r = 0.906, y=125.21+1.2004x, p=0.009), and it was significant. for skipjack tuna, the relationship has a significant strength (figure 5b), (r = 0.634, y=324.230+0.47x, p<0.001) and the relationship has a significant strength. both frigate tuna (figure 5c, r = 0.172, y= 207.836+0.137x, p = 0.121) and indian scad (figure 5d, r = 0.169, y= 824.312-0.871x, p = 0.126) were recorded very weak relationship between landing price and retail price. k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 78 fig 5: regression analysis of weekly average prices of fish between fish landing center and retail stall (january to december 2018). (a) yellowfin tuna, (b) skipjack tuna, (c) frigate tuna, (d) indian scad (y = retail price, x= landing center price, r= correlation coefficient) the retail price of skipjack tuna at the negombo landing center is moderately correlated (r = 0.634) which indicates considerable influences of market integration. but other three fish species did not give any clear indication. thus, the price factor of selected marine fish at the landing site was not a significant impact to the market integration except skipjack tuna fishery. 3.6 characteristic of respondents one part of this study concerned the status of the consumers related to the purchasing of fish from the market. 50% of the consumers were in the 30-39 years of age group and about 75% of them were employed with completing secondary education. based k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 79 on the descriptive statistics analysis, 73% of the consumers were concerned about both price and quality when purchasing the fish. about 60% of the consumers were selected to purchase from retailers considering the mode of purchasing fish (table 8 and 9). table 8: consumer’s priority factors when purchasing fish. criteria frequency percentage quality is ensured 2 6.7 price considered 5 16.7 quality and price considered 22 73.3 taste considered 1 3.3 total 30 100.0 table 9: consumer preferences when buying fish. place frequency percentage cumulative percentage fish port (landing) 2 6.7 6.7 wholesaler 4 13.3 20.0 supermarket 6 20.0 40.0 retailer 11 36.7 76.7 mobile 5 16.7 93.3 online 2 6.7 100.0 total 30 100.0 4 discussion though the required infrastructure facilities were fulfilled at the negombo fish landing site (including auction marketplace and open market) the maintenance work, cleanliness, and hygienic practices of the marketers were noticed as very poor. the overall characteristics of the marketplace related to infrastructure, maintenance, and hygienic practices were kept in good condition functioning under the ceylon fisheries corporation retail outlets. at the negombo auction market, the site visits confirmed that certain handling behaviour of personnel involved may affect the quality and freshness of the fish. most of the time, the cleanliness of the auctioning area was not up to a satisfactory level. when unloading the fish from fishing vessels and handling the fishes were dragged on the floor affecting the appearance and freshness, before and after auctioning unloaded fishes were kept for more than one hour without putting the ice affecting the freshness. it was noticed that poor handling practices when washing the fish, like using lagoon water instead of potable water. k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 80 in this study, ten marketing channels were identified between producer and consumer when purchasing or buying fish at the domestic market. some channels were involved with more than three intermediaries within a short time of period. those involvements were finally affected the purchasing price of the consumer. even freshness, quality of the fish is deteriorated due to less consideration of maintaining the minimum criteria such as inadequate ice, exposure to sunlight, improper packing and packaging, improper storing, and poor hygienic practices. most of the retail outlets were observed to have the routine maintenance of the working environment. channel 06 has been identified as having significant variations when compared with other channels because ceylon fisheries corporation was (cfc) involved as a wholesaler to their retailer shops and to the consumer. at this channel, prices of the fish were also considerably lower than the other channels, as well as freshness qualities of the fish were maintained by dedicated staff members. therefore, the demand of the consumer was also higher than the other retailers. marketing channel 9 also shows some peculiarity, due to the inclusion of a new market trend like online marketing, although that system is somewhat new to the consumers yet. communication with online customers showed that online mode is needed to be popularized in society. it may create a significant change in the efficiency of domestic fish marketing in sri lanka. up to now, very few online traders were involved from the city area of the western province. 5 conclusions the marine fish marketing channels associated with the national level are characterized by a large number of intermediates and redundant participants, ultimately increasing consumer costs. when distributing fresh fish from producers at the negombo fishery harbour to end consumers; ten short-run to long-run marketing channels are identified. in particular, longer channels (more steps in the channels) indicate higher retail prices due to lower fish freshness and quality, increased costs for handling, transporting, unloading, loading, packing, and storage activities. thus, the cost paid by consumers exceeds the actual value of low-quality fish. the profits of income are not absorbed by fishermen and the fishermen's community, and intermediates are making profits. the presence of collectors and assemblers as intermediates would be the cause for the longer length of the channels, also the presence of intermediates caused the delay in product delivery with additional transit points. unless there is an effective cold chain, the freshness and the quality of the fish at the customers’ end would not be achieved for such longer channels, but the prices would be still higher than the expectation. such situations were in practice for decades and because of this unsustainable business model and culture, many benefits were not provided to both the fisher community and the end consumer. especially the shorter marketing k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 81 channels involving cfc plays a great role compared to the other longer marketing channels. the higher prices reflected in the supermarket would be inevitable considering the overheads, cold chain, investments and expenses, and wastages due to the expiry of the product within their outlets. sustainability in consumption, food security, and production associated with marine fish marketing channels and consumers' consumption may be achieved through innovations in terms of technology interventions, the productivity of business and marketing channels, and national-level involvement. among the identified ten marketing channels the marketing channel involved in cfc could be identified as relatively the best, and hence the popularization of this marketing channel could be consequently nationally important. the less productive marketing channels with many intermediates should be rectified or discouraged as such models noted with banned business models and cultures challenge the sustainable consumption and production in the marine fishery sector. information flow from up and down the marketing channels could be used as an effective tool for improving the market and marketplaces. this serves the consumers for being able to understand the necessary information in selecting fish. in the upstream of the marketing, consumers need a proper upward flow of information along with the marketing channels. other aspects of information flow can be linked to the strategic minimization model, post-harvest losses, knowledge, technology management, and the swift transfer and delivery of products and thereby improving marketing channel cultures, practices, and values. acknowledgements the authors are grateful to the head of the post-harvest technology division of the national aquatic resources research and development agency (nara), the owners of fish carrier vessels, and other stakeholders, consumers for providing their valuable time during the research. the authors also wish to acknowledge the constructive criticism of two anonymous reviewers. references abila ro. 1998.the patterns and trends of marketing and consumption of the fish of lake victoria (kenya waters).african journal of tropical hydrobiology and fisheries. 8:45-50. ahmed s, hossain b. 2012. marketing strategies for frozen fish exporters in bangladesh.journal of arts, science and commerce. 3(3):1-8 aktar n, islam mr, hossain mb, rahman m. 2013.fish species availability and marketing system of fish in different markets of noakhali district in bangladesh.world applied science journal. 22(5):616-624. alam mj, yasmin r, rahman a, nahar n, pinky ni, hasan m. 2010.a study on fish marketing system in swarighat, dhaka, bangladesh.natural science. 8(12):96-103. alhassan eh, boateng vf, ndaigo c. 2012.smoked and frozen fish consumption and marketing channels in the tamale metropolis of ghana. ghana journal of development studies. 9(1):21-28. k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 82 ali ea, gaya him, jampada tn. 2008.economic analysis of fresh fish marketing in maiduguri gamboru market and kachallarialau dam landing site of northeastern nigeria.journal of agriculture and social sciences. 4:23-26 ariyawansa s, ginigaddarage p, jinadasa k, chandrika jm, ganegamaarachchi g,, ariyaratne s. 2016. assessment of microbiological and bio-chemical quality of fish in a supply chain in negombo, sri lanka, international conference of sabaragamuwa university of sri lanka 2015 (icsusl 2015) procedia food science 6: 246 – 252 aswathy n, abdussamad em. 2013. price behaviour and marketing efficiency of marine fish in tuticorin, tamil nadu. journal of fisheries economics and development. 13(2):29-35. begum r, akter t, barman pp, marine ss, hossain mm. 2014.potential for development of marine fish marketing systems in the chittagong district of bangladesh.journal of the sylhet agricultural university. 1(2):247-252 boorsma m. 2006. a strategic logic for arts marketing: integrating customer value and artistic objectives. international journal of cultural policy, 12(1), pp.73-92. coyle b, bardi ej, langley. 2003. the management of business logistics: a supply chain perspective‖. transcontinental louiseville. dimitrova b, rosenbloom b. 2010. standardization versus adaptation in global markets: is channel strategy different? journal of marketing channels, 17(2), pp.157-176. dissanayake dct, sigurdsson t. 2005. monitoring and assessment of the offshore fishery in sri lanka. reykjavik, iceland: the united nations university fisheries training programme. fishery industry outlook. 2019. nara, retrieved on 01 june 2020 http://www.nara.ac.lk/wp-content/uploads/2017/09/fisheries-industry-outlook-2018-convertedcopy.pdf. fisheries industry outlook. 2016. nara, http://www.nara.ac.lk/wp-content/uploads/2017/09/2016fisheries-outlook-2018.01.18-new.pdf fisheries statistics. 2007. ministry of fisheries and aquatic resources development & rural economy (mfar), https://www.fisheries.gov.lk/web/index.php?lang=en, 01 june 2020. fisheries statistics. 2019. ministry of fisheries and aquatic resources development & rural economy(mfar), https://www.fisheriesdept.gov.lk/web/images/pdf/fisheries_statistics_2018.pdf, 01 june 2020 flowra fa, bashar ahm, jahan ksn, samad ma, islam mm. 2012. fish marketing system and socioeconomic status of aratdars in natore and rajshahi, bangladesh. journal of our nature.34-43 food balance sheet (2013/17). department of census and statistics, sri lanka, http://www.statistics.gov.lk/agriculture/foodbalancesheet/fbs_report2013-2017.pdf garrow js, james wpt. 1994. human nutrition and dietetics. london, uk: churchill livingstone. gestsson, h., knútsson, ö. and thordarson, g., 2010. the value chain of yellow fin tuna in sri lanka gorchels l. 2004. connect with current, past customers. marketing news, 38(17), pp.43-44. gorchels l, west c, marien ej.2004. the manager's guide to distribution channels. mcgraw hill professional. haputhantri ssk, villanueva mcs, moreau j. 2008. trophic interactions in the coastal ecosystem of sri lanka: an ecopath preliminary approach. estuarine, coastal and shelf science, 76(2), pp.304318. hossain ma, abdulla-al-asif, zafar ma, hossain mt, alam ms, islam ma.2015. marketing of fish and fishery products in dinajpur and livelihoods of the fish retailers. international journal of fisheries and aquatic studies. 3(1):86-92 hussain sa, badar h, khokhar sb. 2003.market intermediaries and their marketing margins for inland fish – a case study of lahore district. international journal of agriculture and biology. 5(1):7376 jamali a. 2013. present status of fish marketing in gopalpurupazila of tangail district. journal of aquatic sciences. 1(2):24-30. kohls rl, uhl jn. 1998. marketing of agricultural products. englewood cliffs, nj: prentice hall. http://www.nara.ac.lk/wp-content/uploads/2017/09/fisheries-industry-outlook-2018-converted-copy.pdf http://www.nara.ac.lk/wp-content/uploads/2017/09/fisheries-industry-outlook-2018-converted-copy.pdf https://www.fisheries.gov.lk/web/index.php?lang=en https://www.fisheriesdept.gov.lk/web/images/pdf/fisheries_statistics_2018.pdf http://www.statistics.gov.lk/agriculture/foodbalancesheet/fbs_report2013-2017.pdf k. edirisinghe et al. marketing channels from negombo fishery harbour ruhuna journal of science vol 12 (2): 64-83, december 2021 83 kohls rl, uhl jn. 2005. marketing of agricultural products. macmillan publishing co., inc., new york. kotler p, rackham n, krishnaswamy s. 2006. ending the war between sales and marketing. harvard business review, 84(7/8), p.68. madugu aj, edward a. 2011. marketing and distribution channel of processed fish in adamawa state, nigeria. globaljournal of management and business research. 11(4):21-26. mfard. 2018. annual performance report. retrieved from https://www.parliament.lk/uploads/documents/paperspresented/performance-report-department-offisheries-aquatic-resources-2018.pdf mfard. 2019. fisheries statistics in 2018, ministry of fisheries and aquatic resources development, sri lanka. mutambuk mk. 2014. marketing strategies of commercial fish farming under economic stimulus programme (esp) in kenya: an empirical study of kitui county. international journal of humanities and social science. 4(8):111-121. ostrow r. 2009. the fairchild dictionary of retailing, second edition,new york, fairchild books, inc. perera kbp, herath hmlk, jayasinghe-mudalige uk, edirisinghe jc, udugama jmm, wijesuriya w. 2016.price analysis of selected marine fish available in colombo fish markets. porter’s vcm. 1985. what is value chain? retrieved on 20th feb, 2018. http://www.dspmuranchi.ac.in/pdf/blog/what%20is%20the%20first%20mover%20advantage.pdf rabby af, hossain ma, alam mt, uddin ms, dey t. 2015.status and economics of three upazilla fish markets in moulavibazar district in bangladesh.journal of sylhet agricultural university. 2(1):87-95 rahman m, akter m, odunukan k, haque se. 2020. examining economic and technology‐related barriers of small‐and medium‐sized enterprises internationalisation: an emerging economy context. business strategy & development. 3(1):16-27 rokeya ja, ahmed ss, bhuiyan as, alam ms. 1997. marketing system of native and exotic major carps of rajshahi district. bangladesh journal of fisheries. 20(1-2):99-103 sathiadhas r, kanagam a. 2000.distribution problems and marketing management of marine fisheries in india. in: marine fisheries research and management, pillai, v.n and menon, n.g. (eds.). mfri; kochi, 858-875. segetlija z, mesarić j, dujak d. 2011, january.importance of distribution channels-marketing channelsfor national economy. in 22nd cromar congress weerasekara kaws, jayampathi ommd, hettige nd, azmy sam, amarathunga aad, wickramarachchi wdn, maddumage mdsr, jayawardena jkpc, narangoda srcnk, rajapaksha rmng, liyanage npp. 2015. assessment of water pollution status of selected fishery harbours located in the southern province of sri lank, journal of environmental professionals sri lanka: 4(2): 36-46 wickramasinghe wr, bavinck m. 2015. institutional landscapes affecting small-scale fishing in southern sri lanka-legal pluralism and its socio-economic effects. maritime studies, 14(1), pp.119. yamao m. de silva dam. 2006. export oriented tuna industry in sri lanka: an analysis of the sources of export success. zaragosa ec, pagdilao cr, moreno e.p. 2004. fisheries for tuna and other large pelagic fishes. in turbulent seas: the status of philippine marine fisheries. cebu city (philippines): coastal resource management project, pp.38-41. https://www.parliament.lk/uploads/documents/paperspresented/performance-report-department-of-fisheries-aquatic-resources-2018.pdf https://www.parliament.lk/uploads/documents/paperspresented/performance-report-department-of-fisheries-aquatic-resources-2018.pdf sv-lncs ruhuna journal of science vol 9: 70-77, june 2018 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v9i1.37 university of ruhuna  faculty of science, university of ruhuna 70 sri lanka short paper comparative study of the nutritional, phytochemical and mineral compositions of the nuts of tropical almond (terminalia catappa) and sweet almond (prunus amygdalus) r.a. salawu*, a.f. onyegbula, i.o. lawal, s.a. akande and a.k. oladipo nigerian stored products research institute, headquarters. km, 3, asa-dam road, p. m. b. 1489, ilorin. nigeria. correspondence: * salawuadenike05@yahoo.com; orcid 0000-0002-0882-822x received: 23rd march 2018, revised: 21st june 2018, accepted: 30th june 2018 abstract. the study was conducted to compare the nutritional, phytochemical and mineral compositions of tropical almond with sweet almond. sample of terminalia catappa nuts were collected within the premises of nigerian stored products research institute (nspri), ilorin, nigeria while prunus amygdalus was purchased from shoprite palms mall ilorin, nigeria. proximate, phytochemical and mineral analyses were carried out using standard procedures. results showed that t. catappa was significantly (p<0.05) high in ash (4.84%), crude fibre (15.54%), carbohydrates (2.91%) and some mineral elements such as potassium, zinc, iron, magnesium and copper. prunus amygdalus was significantly (p<0.05) high in ether extract (50.96%) while no significant difference (p>0.05) was recorded in their protein contents (33.00 and 32.89% respectively). p. amygdalus was significantly (p<0.05) high in phytochemicals such as tannin (748.49µg/g), phenols (1,781.50 µg/g), flavonoids (456.38 µg/g), saponin (158.70 µg/g) and alkaloids (240.11µg/g) while t. catappa was significantly (p<0.05) high in glycosides (220.27µg/g).the differences in phytochemicals might be due to the differences in drying and other processing methods. t. catappa can well compete with p. amygdalus if the value chain is improved upon by proper packaging and storage for commercial purposes. keywords. almond, nuts, nutrition, ilorin, nigeria 1 introduction plants continue to revolutionize the face of the earth through distinctive benefits they provide across the world (agronigeria 2017). plant seeds form an important part of human diet and their significance especially in the diet of the population of developing countries is increasing for several reasons (alozie and udofia 2015). almond seeds are a good source of proteins, edible oils and fats in the diets as well as potential raw materials for local industries. mailto:salawuadenike05@yahoo.com r. a. salawu et al. comparative study of tropical almond and sweet almond ruhuna journal of science vol 9: 70-77, june 2018 71 this seed is also used by many rural dwellers in southern nigeria to fortify the local complimentary foods, which are usually low in protein (christian 2007). almond is a large tree that grows mainly in the tropical regions of asia, africa and australia (pankaj and robert 2008). this is also identified by common english names such as; country-almond, indian-almond, malabaralmond, sea-almond, tropical almond (usda 2016). terminalia catappa l. (tropical almond) belonging to the family combretaceae is one of the underutilized tree species found in tropics including nigeria. it is planted extensively in the tropics for shade and ornamental purposes, esp. in parks, along avenues, as well as home gardens (mbah et al. 2013). the fruit of the tree is a drupe 5-7 cm long and 3-5.5 cm broad, green at first, then yellow and/red when ripe. the fruit contains a single seed, with sweet edible fibrous pulp which is eaten by children as forage snacks and there has been no report of associated toxicity with its consumption (mbah et al. 2013, arumugam et al. 2015). sweet almond (prunus amygdalus var. dulcis) is another variety which belongs to the family roseceae. it is also a drupe with a thick leathery grey-green exocarp called the hull. it has been shown to be a nutritious food providing more than 20% of daily value of riboflavin, niacin, vitamin e, calcium, iron, magnesium, manganese, phosphorus and zinc in each 100 g (berryman et al. 2011). literature has revealed the nutritional composition of tropical almond nuts which is predominant in nigeria. mbah et al. (2013) published some data on nutrient potential of almond seeds (t. catappa l.) sourced from three eastern states of nigeria. they reported t. catappa seeds to be rich in protein and could reduce the level of malnutrition in most impoverished countries of africa and hence encouraged its use in food supplements either in raw or roasted form to improve the food nutrient content. ezeokonkwo and dodson (2007) reported the potential of t. catappa seed as a source of dietary protein. olatidoye et al. (2011) examined the chemical composition and physicochemical characteristics of tropical almond nuts (t. catappa l.) cultivated in south western nigeria and reported that the physicochemical properties of the almond seed oil indicated that it is edible, drying and suggested its suitability for industrial purposes as well as the nutritional potentials of the nut, which could serve as an alternative food ingredient for unsaturated vegetable oil. christian and mark (2006) also evaluated the nutritional potential of the nuts of tropical almond (t. catappa l.) and also reported that tropical almond nut can contribute useful amounts of essential nutrients to the diet of man. the variety of almond nuts sold in most of nigerian supermarkets is sweet almond (p. amygdalus dulcis) which is native to the middle east and south asia (agronigeria 2016). the objective of the present study is to compare the nutritional, phytochemical and minerals composition of two varieties of almond with a view of revealing the potential of the local variety (t. catappa) of almond, r. a. salawu et al. comparative study of tropical almond and sweet almond ruhuna journal of science 72 vol 9: 70-77, june 2018 and to provide information regarding the processing of tropical almond into a value added product that could be marketed worldwide, as a dietary supplement. 2 materials and methods 2.1 sample collection and preparation mature almond nuts (terminalia catappa) were collected within the premises of nigerian stored products research institute (nspri), ilorin, nigeria. the fruits were sun dried and manually de-hulled to get the nuts in their whole form. the nuts were then stored in an air tight bag in a cool (20ºc) and dried place till usage. the foreign almond was purchased from shoprite, palms mall, ilorin, nigeria under the trade name: blue diamond almonds (blue diamond growers: sacramento, ca 95812 usa). 2.2 determination of the proximate composition the proximate analysis was carried out following the method of aoac (2016) while carbohydrate content was determined by difference (pearson, 1976). the moisture was determined by hot air oven (dhg-9055a drying oven, searchtech instruments) method at 105ºc for 5 h. the micro kjeldahl method was used for the determination of protein content. the fat content or ether extract was determined by extracting 5g of sample with petroleum ether (boiling point of 40ºc to 60ºc) using soxhlet (borosilicate glass, 45/50 kimax usa 24/40) solvent extraction method. ash was determined by weighing 5 g of charred sample into a tarred porcelain crucible, which was incinerated at 6000c for 6 h in an ash muffle furnace (sx-5-112 boxresistance furnace, searchtech instruments) until ash was obtained. the crude fibre was determined by exhaustive extraction of soluble substances in the sample using 1.25% h2so4 acid and 1.25% naoh solution after the residue was ashed and the loss in weight was recorded as crude fibre. 2.3 mineral analysis the minerals; sodium (na), potassium (k), calcium (ca), magnesium (mg), iron (fe), copper (cu) and zinc (zn) were determined using the dry ash procedure in accordance with the method of oshodi and fagbemi (1991). accordingly, 1 g of sample was weighed into a crucible and placed in a muffle furnace at 6000c for 5 h to ash and then transferred into desiccators to cool to room temperature (23ºc). the ash was dissolved in 10% hydrochloric acid (10 ml), filtered and diluted to 100 ml volume with distilled water. from the digest, the required elements were determined; na and k were measured r. a. salawu et al. comparative study of tropical almond and sweet almond ruhuna journal of science vol 9: 70-77, june 2018 73 by the use of jenway digital flame photometer, pfp7/c analytical flame photometer as described by bonire et al. (1990). ca, mg, fe, cu, and zn were measured using atomic absorption spectrophotometer (aas 969 buck scientific vgp 210, buck scientific inc. 58 fort point st, east norwalk, ct. 06855) in accordance with aoac (2000) and compared with absorption of standards of the elements. 2.4 phytochemical screening the qualitative phytochemical analysis was done to determine the presence of alkaloids, flavonoids, saponin, tannin, phenols, glycosides, steroids and phlobatannin in the nuts using the methods described by dey et al. (2003), mehta et al. (2013) and ifemeje et al. (2014).the quantitative analyses to determine the amount of alkaloids, glycosides and saponin in the nuts were carried out using the methods described by obdoni and ochuko (2001) and ifemeje et al. (2014) while the concentration of tannin, phenols and flavonoids were determined by the method described by vijay and rajendra (2014). 2.5 statistical analysis data on proximate, minerals and phytochemical analyses were reported as mean ± sd of triplicate (n=3) determinations. all data were pooled together and analysed by using student’s t-test. means were considered significant at probability level of 5 % (p<0.05). 3 results and discussion 3.1 proximate composition the results presented in fig. 1 are the proximate analysis of terminalia catappa and prunus amygdalus. it showed that t. catappa was significantly (p<0.05) high in moisture (30.47%), crude fibre (15.54%), ash (4.84%) and in nitrogen free extract (nfe) or carbohydrates content (2.91%) compared to p. amygdalus; low in moisture (8.55%), crude fibre (11.08%), ash (4.13%) and nfe (0.96%) respectively. in contrast, p. amygdalus was significantly (p<0.05) high in crude fats (50.96%) compared to t. catappa (43.71%) while there was no significant difference (p>0.05) between the protein contents of t. catappa (33.00%) and p. amygdalus (32.87%).the protein content of t. catappa (33.00%) in this study was higher than some data available in literature. the protein contents of t. catappa (33.00%) and p. amygdalus (32.89%) compared well and were even higher than the protein contents of most conventional oil seeds; for instance, groundnut (25.00%), black-eyed r. a. salawu et al. comparative study of tropical almond and sweet almond ruhuna journal of science 74 vol 9: 70-77, june 2018 beans (27.13%), brown beans (28.00%), cowpea (27.80%), bambara nuts (23.41%) and pigeon pea (21.88%), (ezeokonkwo and dodson 2007). according to ezeoknkwo and dodson (2007) the protein content of t. catappa was 25.81%. similarly, mbah et al. (2013) reported the nutritional composition of t. catappa as; protein (23.40%), fats (22.0%), ash (4.1%), fibre (6.4%) and carbohydrates (34.6%). also the proximate composition results in the present study are higher than those report by agunbiade and olanlokun (2006); protein (11.52%), ash (6.76%), fibre (5.09%) and carbohydrates (54.87%) and those reported by olatidoye et al. (2011); protein (32.6%), fats (3.3%), ash (4.8%), fibre (0.4%) and carbohydrates (49.9%). the differences in protein contents may be due to varietal and other environmental differences such as soil, water, climate etc. fig. 1. nutritional composition of terminalia catappa and prunus amygdalus (dry matter basis). bar represents mean of triplicate readings (n=3) while error bar represents standard deviation. bars with unshared alphabets are significantly different (p<0.05). 3.2 phytochemical composition phytochemical screening of the two varieties of almond showed that all the parameters tested (in exception of steroids and phlobatannin) were present; these include, tannin, phenols, flavonoids, saponin, alkaloids and glycosides. the quantitative analysis of the phytochemicals (fig. 2) showed that p. amygdalus was significantly (p<0.05) higher in tannin (748.49 µg/g), phenols (1,781.50 µg/g), flavonoids (456.38 µg/g), saponin (158.70 µg/g) and alkaloids (240.11 µg/g) as against t. catappa which are; tannin (388.95 µg/g), phenols (410.83 µg/g), flavonoids (73.28 µg/g), saponin (86.32 µg/g) and r. a. salawu et al. comparative study of tropical almond and sweet almond ruhuna journal of science vol 9: 70-77, june 2018 75 alkaloids (210.65 µg/g) respectively. conversely, t. catappa was significantly (p<0.05) higher in glycosides (220.27 µg/g) compared to p. amygdalus (181.65 µg/g). the reason for higher concentrations of tannin, phenols, flavonoids, saponin and alkaloids in p. amydgladus may be due to differences in varietal differences, soil and other environmental factors as it was also reported by mbah et al. (2013). the tannin contents of t. catappa and p. amygdalus in the present study are higher than some literature reports. agunbiade and olanlokun (2006) reported the tannin content of indian almond to be 18.2 microgram per gram while mbah et al. (2013) reported the tannin content of t. catappa to be 0.03 microgram per gram. this may be linked with vegetative loss during processing. fig. 2 quantitative phytochemical composition of terminalia catappa and prunus amygdalus. bar represents mean of triplicate readings (n=3) while error bar represents standard deviation. bars with unshared alphabets are significantly different (p<0.05). 3.3 mineral content the mineral elements tested for in t. catappa and p. amygdalus were sodium (na), potassium (k), calcium (ca), zinc (zn), iron (fe), magnesium (mg) and copper (cu). sodium and calcium were not detected in both samples. the results of the mineral analysis (fig. 3) revealed that t. catappa was significantly (p<0.05) high in k (4.80 mg/ 100g), zn (0.0075 mg/ 100g), fe (0.0035 mg/ 100g), mg (0.064 mg/ 100g) and cu (0.0020 mg/ 100g) compared to p. amygdalus k (3.80 mg/ 100g), zn (0.0025 mg/ 100g), fe (0.0030 mg/ 100g), mg (0.0615 mg/ 100g) and cu (0.0015 mg/ 100g) respectively. the presence of macro and micro elements in t. catappa from the present study showed that this seed has some beneficial nutritional potential (mbah et al., 2013). r. a. salawu et al. comparative study of tropical almond and sweet almond ruhuna journal of science 76 vol 9: 70-77, june 2018 fig. 3 mineral composition of terminalia catappa and prunus amygdalus. bar represents mean of triplicate readings (n=3) while error bar represents standard deviation. 4 conclusions the study has shown that t. catappa compares well with p. amygdalus in terms of nutritional, phytochemical and mineral compositions. this research work would serve as a source of information for processors or stakeholders regarding the value addition of the indigenous variety of almond. acknowledgements authors wish to acknowledge the management of nigerian stored products research institute, ilorin for the use of her laboratory and equipment during the course of this research work. two anonymous reviewers are acknowledged for critical comments on the initial draft. references agronigeria news, 2016. tropical almond (terminalia catappa). [internet document] retrieved from https://www.agronigeria.com.ng on 03122018@11:15 am. agronigeria news, 2017. tropical almond (terminalia catappa). [internet document] retrieved from https://www.agronigeria.com.ng on 03122018@10215 am. agunbiade so, olanlokun, jo. 2006. evaluation of some nutritional characteristics of indian almond (prunus amygdalus) nut. pakistan journal of nutrition 5(4): 316–318. alozie ye, udofia us. 2015. nutritional and sensory properties of almond (prunus amygdalu var. dulcis) seed milk. world journal of dairy & food sciences 10(2):117-121. aoac, 2000. official methods of analysis of association of official analytical chemists international 20thed. aoac international suite 500 481 north frederick avenue. gaithersburg, maryland 20877–2417 usa. arumugam va, natarajan d, and pannerselvam pk. 2015. an updated review of terminalia catappa. pharmacognosy review 9(18): 93–98. https://www.agronigeria.com.ng/ https://www.agronigeria.com.ng/ https://www.ncbi.nlm.nih.gov/pubmed/?term=anand%20av%5bauthor%5d&cauthor=true&cauthor_uid=26392705 https://www.ncbi.nlm.nih.gov/pubmed/?term=divya%20n%5bauthor%5d&cauthor=true&cauthor_uid=26392705 https://www.ncbi.nlm.nih.gov/pubmed/?term=kotti%20pp%5bauthor%5d&cauthor=true&cauthor_uid=26392705 r. a. salawu et al. comparative study of tropical almond and sweet almond ruhuna journal of science vol 9: 70-77, june 2018 77 berryman ce, preston ag, karmally w, deckelbaum rj, kris-either ton pm. 2011. effects of almond consumption on the reduction of ldl cholesterol: a discussion of potential mechanisms and future research directions. journal of nutritional revolution 69:171-85. bonire jj, jalil nsn, lori ja. 1990. sodium and potassium content of two cultivars of white yam (dioscorea rotundata) and their source soils. journal of science and food agriculture 53: 271-274. christian a. 2007. fluted pumpkin (telfairia occidentalis hook f.) seed: a nutritional assessment. electronic journal of environmental agricultural & food chemistry 6(2): 1787-1793. christain a, mark eu. 2006. nutritional potential of the nut of tropical almond (terminalia catappia l.) pakistan journal of nutrition 5(4): 334-336. dey p, chaudhuri tk, dutta s. 2003. quantification and correlation of bioactive phytochemicals of croton bonplandianum leaves of sub-himalayan region of west bengal. asian journal of pharmaceutical clinical research 6(3): 124-147 ezeokonkwo ca, dodson wl. 2007. the potential of terminalia catappa (tropical almond) seed as a source of dietary protein. journal of food quality 27(3): 208-219. ifemeje jc, egbuna c. eziokwudiazo jo, ezebuo, fc. 2014. determination of the anti-nutrient composition of ocimum gratissimum, corchorus olitorius, murraya koenigii spreng and cucurbita maxima. international journal of innovation and scientific research 3(2): 127133. mbah bo, eme pe, eze cn. 2013. nutrient potential of almond seed (terminalia catappa) sourced from three states of eastern nigeria. african journal of agricultural research 8(7): 629-633. mehta k, patel bn, jain bk. 2013. phytochemical analysis of leaf extracts of phyllnthus fraternus. research journal of recent sciences 2:12-15 obdoni b, ochuko p. 2001. phytochemical studies and comparative efficacy of the crude extracts of some homostatic plants in edo and delta states of nigeria. global journal of pure and applied sciences. 8: 203-208. olatidoye o p, sobowale ss, akinlotan jv, olorode oo. 2011. chemical composition and physicochemical characteristics of tropical almond nuts (terminalia catappa l.) cultivated in south western nigeria. journal of medical and applied biosciences 2: 1–10. oshodi aa, fagbemi tn. 1991. chemical composition and functional properties of full fat fluted pumpkin seed flour. nigerian food journal 9: 26–32. pankaj o, robert ep. 2008. west indian almond terminalia catappa l. combretaceae. encyclopedia of fruit and nuts. j. janick and r.e. paull, (eds). cabi, wallingford, united kingdom. pp. 273-276. pearson d. 1976. the chemical analysis of foods, 7th edition, churchill livingstone, london, p. 147. usda 2016. terminalia catappa l. germplasm resources information network (grin), agricultural research service (ars), united states department of agriculture; retrieved 3 july 2016. vijay dt, rajendra, sb. 2014. estimation of total phenol, tannin, alkaloid and flavonoid in hibscus tiliaceus linn. wood extracts. journal of pharmaognosy and phytochemistry 2 (4): 41-47. ruhuna journal of science vol 10(1): 88-95, june 2019 eissn: 2536-8400 faculty of science doi: http://doi.org/10.4038/rjs.v10i1.53 university of ruhuna  faculty of science, university of ruhuna sri lanka 88 short paper potential of cow dung as insect herbivore repellent on cowpea (vigna unguiculata) a.l. ogunyebi 1* , a. olubiyo 1 , k. o. omoyajowo 1, 2 , t.s. fingesi 1 1 department of cell biology and genetics, university of lagos, akoka, lagos, nigeria 2 department of science policy and innovation studies, national centre for technology management, victoria island, lagos, nigeria *correspondence: logunyebi@unilag.edu.ng; https://orcid.org/0000-0003-2315-470x received: 27 th may 2018, revised: 13 th november 2018, accepted: 13 th may 2019 abstract. this study examined the potential of using cow dung as a repellent of herbivore pest on cowpea (vigna unguiculata) and the work was carried out in the botanical and zoological garden of the university of lagos, akoka, lagos, nigeria. four accessions of cowpea were grown in the soil samples collected from the botanical and zoological garden of the university of lagos. the experiment was arranged in a randomized block design with three replicates to monitor the potential of using cow dung as a repellent of herbivore pest on cowpea for a period of six weeks. the data collected was analyzed using anova. the results showed that cow dung treatments did not have significant effect (p>0.05) in repelling the insects of cowpeas. however, the plants treated with high concentrations of cow dung attracted significantly (p<0.05) higher number of insect pests and as the days of the application of the treatments increases, the number of leaves damaged by the insects increases. this study therefore recommends that a further study should be carried out using some other different plants species at different locations and different environmental conditions. keywords: cow dung, cowpea, insect repellent 1 introduction cowpea vigna unguiculata is an economically important food crop which is a source of plant protein, minerals, lipids and vitamins for humans and livestock (dolvo et al. 2014). it is valued as a cheap source of protein for many people that could not afford for animal protein such as meat, fish, and eggs in nigeria. it is processed in various form but often consumed as “moi-moi” and “akara”− a special delicacy often traded by roadside food vendors in nigeria https://orcid.org/0000-0003-2315-470x a. l. ogunyebi et al. potential of cow dung as insect repellent on cowpea ruhuna journal of science vol 10(1): 88-95, june 2019 89 for public consumption. it is mainly grown in northern nigeria but its cultivation has recently extended to southern part of nigeria where it is cultivated in the western and eastern part respectively (emosairue et al. 2014). cowpea has different pests that pose great threat to its proper growth and yield (nampala et al. 1999, ajeigbe and singh 2006). major insect pests that cause severe damage to cowpea during growth stages include; cowpea aphid (aphis craccivora koch), foliage beetles (ootheca spp.)and medythia spp.), the flower bud thrips (megalurothripssj ostedti tryb) the legume pod borer (marucavitrata fabricius) and the sucking bug complex of species such as clavigralla spp., anoplocnemis spp., riptortus spp., mirperus spp., nezara viridula and aspavia armigera are the most important and prevalent (olatunde et al. 2013). pod bugs for example suck sap from green pods, causing abnormal pod and seed formation and yield losses of 30-70% (sexena 2010). the prevalence and distribution of cowpea insect pests may however be influenced by some factors. recent study suggested that the use of effluent water to irrigate crops may increase incidence, abundance and damage caused by cowpea insect pests possibly reducing cowpea plant productivity and vigor (tiroesele et al. 2017). sawadogo et al. (2009) also posited that there may be variation in prevalence and distribution of cowpea insect pest among agroclimatic zones. two common aphid parasites, lysiphlebus spp. and diaeretiella spp. have also been identified as major parasites that determine the abundance of aphids (ucanr 2017). hence, their presence on fields may reduce the abundance of aphids. the use of biological control of insect pest of cowpea is receiving attention lately. this is because conventional methods such as synthetic insecticides cannot be used to control these pests for safe and sustainable cowpea production. taken together, cowpea is an important crop but insect pest is a cultivation-constrain. moreover, insecticide applications are costly and have many environmental consequences including direct human health hazards. hence, it requires alternate non-chemical approach toward insect pest control. cow dung could be a good option since a number of studies have documented its insect-repellent nature (okogbue and ojo 2003, mandavgane et al. 2005). hence, in this study insect-repellent nature of cow dung is tested on cowpea. 2 material and methods 2.1 study area the experiment was carried out in the botanical and zoological garden of university of lagos, akoka, lagos which is situated in the northeast of a. l. ogunyebi et al. potential of cow dung as insect repellent on cowpea ruhuna journal of science vol 10(1): 88-95, june 2019 90 lagos, in the yaba local government area of lagos state nigeria. it lies in the latitude 6 º 31’ 0’n and 3 º 23’10’e and longitude 6.51667 º n and 3.38611 º e. the university is located within the tropical lowland region with two distinct seasons; the wet season is between march–october, while the dry season is between november–february. the mean annual rainfall is between 1250–2500 mm, while the mean monthly temperature ranges between 25.7 º c usually in july and 30.2 º c in february. 2.2 experimental procedures germination experiment was carried out for six weeks on the four accessions of cowpea collected from national centre for genetic resources and biotechnology (nacgrab).three seeds of each of the accessions were sown in plastic containers filled with soil collected from the botanical and zoological garden of the university of lagos, akoka, lagos. the experiment was 4 x 4 factorial experiment laid out in completely randomized design (crd). the treatment combinations applied were three different concentrations of cow dung (0%, 25%, 50% and 75% v/w) replicated three times. the treatments were applied six times after germination of the cowpea at 5 days interval (5, 10, 15, 20, 25, and 30 days respectively). the treatment combinations are as follows: nolo = t1 – nasa/sa/07/028 + 0% cd nol1 = t2– na/sa/07/063 + 0% cd nol2 = t3 – na/10/11/08/047 + 0% cd nol3 = t4 – na/sa/07/009 + 0% cd n1lo = t5 – nasa/sa/07/028+ 25% cd n1l1 = t6 – na/sa/07/063 + 25% cd n1l2 = t7 – na/10/11/08/047 + 25% cd n1l3 = t8 – na/sa/07/009 + 25% cd n2lo = t9 – nasa/sa/07/028+ 50% cd n2l1 = t10 – na/sa/07/063 + 50% cd n2l2 = t11 – na/10/11/08/047 + 50% cd n2l3 = t12 – na/sa/07/009 + 50% cd n3lo = t13 – nasa/sa/07/028+ 75% cd n3l1 = t14 – na/sa/07/063 + 75% cd n3l2 = t15 – na/10/11/08/047 + 75% cd n3l3 = t16 – na/sa/07/009 + 75% cd a. l. ogunyebi et al. potential of cow dung as insect repellent on cowpea ruhuna journal of science vol 10(1): 88-95, june 2019 91 2.3 data analysis data on cow dung treatment performance were subjected to analysis of variance (anova) and the treatment mean were separated using least significant difference (lsd). pearson correlation was used to show the relationship between the number of leaves damaged and number of days of the treatment. the statistical tool used was graphpad prism 7.04. 3 results table 1 showed the mean values of leaves damaged after the treatment. the result showed that 50% and 75% had the highest mean value of leaves damaged (t15 in accession na/10/11/08/047), while the least number of leaves damaged was observed in control (t1, t4 and t12 in accessions na/sa/07/028 and na/sa/07/009). table 1: mean values of the damaged leaves in each accession after treatment. treatment control 25% 50% 75% t1 0 0.33 0.67 0.67 t2 0.33 0.33 0.67 0.33 t3 0.67 0.33 0.67 0.67 t4 0 1 0.67 0.67 t5 0.67 0.33 1.67 1 t6 1.67 1.33 2 1.67 t7 2 2 2 1.67 t8 0.33 1.67 1.67 2 t9 2.33 1 2.33 3.33 t10 2.33 3.33 5 3.33 t11 4.33 4 3.67 3.67 t12 0 2.67 2 1.67 t13 3 3.67 4 4.33 t14 2.67 4.33 5.33 4 t15 4.67 3.33 4.33 4.67 t16 0.33 0.67 3 3 a. l. ogunyebi et al. potential of cow dung as insect repellent on cowpea ruhuna journal of science vol 10(1): 88-95, june 2019 92 table 2. analysis of variance for the mean values of damaged leaves in vigna unquiculata analysis of variance ss df ms f (dfn, dfd) p value treatment (between columns) 7.761 3 2.587 f (2.396, 35.94) = 6.168 0.0032* individual (between rows) 117.1 15 7.804 f (15, 45) = 18.61 <0.0001* residual (random) 18.87 45 0.4194 total 143.7 63 * significant at p<0.01 probability level. the results presented in table 2 showed that there were significant differences among the treatment at 5% probability level for the three conditions [f (15, 45) = 18.61, p = 0.0001]. treatments 1-9, 11, 13 and 15 were not significant at (p>0.05). fig 1. representation of leaves damaged for all the treatments on vigna unquiculata a. l. ogunyebi et al. potential of cow dung as insect repellent on cowpea ruhuna journal of science vol 10(1): 88-95, june 2019 93 however, treatment 12 at 25% concentration was significant at (p<0.05) when compared to control. similarly, treatments 10, 14, and 16 at 50% concentration were significant at (p<0.05) while treatments 16 at 75% concentration was strongly significant at (p<0.05) as shown in figure 1. 3.1 treatment performance leaves damaged in all the accessions of vigna unquiculata with increased in the number of days are shown in figure 2. leaves damaged increases with increasing number of days of exposure to cow dung in all the accessions of vigna unquiculata. major insect pest identified in all accessions were cowpea aphid (aphis caraccivora koch), legume bud thrips (megalurothrips sjostedti tryb) and armyworm (spodoptera exigua hübner). fig 2. treatment performance on all the accessions of vigna unquiculata 3.2 correlation this study shows a positive correlation between the number of leaves damaged and number of days of the treatment. as days of treatments increases, the number of leaves damaged by insects increased (table 3). a. l. ogunyebi et al. potential of cow dung as insect repellent on cowpea ruhuna journal of science vol 10(1): 88-95, june 2019 94 table 3. correlation matrix between the number of leaves damaged and number of days of the treatment. control 25% 50% 75% control 0.8890 0.8825 0.8976 25% 0.8890 0.8747 0.8651 50% 0.8825 0.8747 0.9472 75% 0.8976 0.8651 0.9472 all correlations are significant at p<0.01 level (2-tailed) 4. discussion there is an increase usage of cow dung for different purposes in different countries around the world (okogbue and ojo 2003). it has been used as mosquito repellent (mandavgane et al. 2005). in this study, the potential of cow dung as insect repellent in vigna unquiculata showed negative results. it was observed that more leaves were damaged at high concentration of cow dung treatment with respect to increasing number of days. the cowpea used for control and those that were treated, all had the leaves damaged. it was observed in this study that cowpeas treated with higher concentrations (50% and 75%) had more leaves damaged than 25% concentration and the control. consequently, the treatments did not repel the insect pests of cowpea (vigna unguiculata). this was contrary to the potential of cow dung being used as mosquito repellent, which is now a common practice in india (atulesh 2011). however, the inability of the treatment to repel insect pest maybe due to other factors that serve as limitations or barriers to this study. one of the limitations was the raining period during which the research was conducted. another limitation could be using only one location for the study. other different locations with different environmental conditions may produce different results. sawadogo et al. (2009) posited that there may be variation in prevalence and distribution of cowpea insect pest among agro-climatic zones. similarly, the study was carried out using one species of plant though with different accessions, but using different species of crops for this study may however, produce a different result. 5 conclusions this present study revealed that cow dung at different concentration treatments does not have the potential to repel insect pests of cowpeas, rather, a. l. ogunyebi et al. potential of cow dung as insect repellent on cowpea ruhuna journal of science vol 10(1): 88-95, june 2019 95 it attracts more insect pests to these crops. hence, it cannot be used as insect pest repellent of cowpea. additionally, the properties of cow dung that is responsible for attracting insect pests of cowpea should be further examined. nonetheless, government should encourage more research on alternate nonchemical approach to pest control on cowpea and other economic food crops through adequate funding. acknowledgements authors sincerely thank kehinde sowunmi for his assistance on statistical analysis. many thanks go to the editor and our anonymous reviewers for their critical feedback and suggestions which has led to great improvement in the final preparation of this manuscript. we also thank the environmental biology research team of the department of cell biology and genetics, university of lagos, akoka for their collegiality, work ethic and willingness to illuminate our minds with great ideas. references ajeigbe ha, singh b. 2006. integrated pest management in cowpea: effect of time and frequency of insecticide application on productivity. crop protection 25(9):920-925. atulesh 2011. cow dung: a composted fertilizer. retrieved from http://www.wealthywaste.com/ cow-dung-a-composted-fertilizer on september 9 th , 20i8 dolvo f, williams c, soaka l. 2014. cowpea: home preparation and use in west africa. international research centre. ottawa. 308pp. emosairue s, eze d, okore k. 2014. timing of insecticidal application in vignaunguiculata (l.) walp and its potential as late season crop in calabar area, nigeria. journal of applied chemistry and agricultural resources 8 (2): 6-12. mandavgane sa, pattalwar vv, kalambe ar. 2005. development of cow dung based herbal mosquito repellent. natural product radiance 4(4): 270-273. nampala p, ogenga-latigo mw, kyamanywa s, adipala e, karungi j, oyobo n, obuo je, jackai le. 1999. integrated management of major field pests of cowpea in eastern uganda. african crop science journal. 7(4): 479-486. okogbue ec, ojo bo. 2003. local production of renewable energy (biogas) from animal waste for domestic and laboratory uses. nigerian journal of solar energy 14: 121-125. olatunde ja, odebiyi j, chaing hs, jackai ln. 2013. identification of sources of resistance in cowpea (vigna unguiculata) l. walp. to clavigralla tomentosicollis stai (hemipera, coreidae). journal on insecticide application 2(4): 455-461. sawadogo a, thio bb, kiemde s, drabo i, dabire c, ouedraogo j, mullens tr, ehlers jd, roberts pa. 2009. distribution and prevalence of parasitic nematodes of cowpea (vigna unguiculata) in bukina faso. journal of nematology 41 (2): 120-127. sexena rc. 2010. natural resistance of plant to pests: roles of allelochemicals. american chemical society. washington d. c. 235pp. ucanr (university of california, agriculture and natural resource). 2017. uc pest management guidelines. retrieved from ipm.ucanr.edu on 27 th november 2018. tiroesele b, sitwane m, obopile m, ullah mi, ali s. 2017. effects of effluent water on the abundance of cowpea insect pests. environmental monitoring and assessment 189 (11): 110. ruhuna journal of science vol 8: 44-54, june 2017 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v8i1.25 university of ruhuna  faculty of science, university of ruhuna sri lanka 44 short paper impact of climate variability on vegetable crops in ilorin, kwara state, nigeria adeniyi adedapo department of geography, kwara state polytechnic, ilorin, nigeria correpondence: dapomoses@gmail.com received: 3rd february 2017, revised: 7th june 2017, accepted: 15th june 2017 abstract climate is one of the key factors in agricultural productivity. climate variability plays an important role in food production. vegetable crops are sensitive to climate variability. this paper examines the impacts of climate variability on vegetable crops in ilorin, nigeria. climatic data on temperature, relative humidity, rainfall and sunshine hours were collected for a period of ten years. agricultural data on the yield of sweat potato, okra, pepper, tomato and amaranthus were also collected for the same period. multiple regression, trend analysis, correlation statistics and standardized anomaly index (sai) were employed in the data analysis. the results obtained indicate that the selected climatic elements have weak impact on the yield of okra, sweat potato, pepper, and amaranthus. however, the selected climatic elements have strong impact on the yield of tomato. these results imply that variation in the yield of okra, sweat potato, pepper, and amaranthus could be attributed to nonclimatic factors like soil fertility and crop management practices while variation in the yield of tomato could be attributed to climatic factors. the paper thus suggests measures such as breeding new varieties of tomato that are tolerant to climatic stress or variability, application of fertilizer to improve the fertility of the soil and the use of modern agricultural techniques to improve the productivity of vegetable crops in ilorin. keywords: climate, crops, ilorin, variability, vegetable. 1 introduction according to ward (2016) the term ‘vegetable’ in its broadest sense refers to any kind of plant life or plant product. in the narrower sense, it refers to the fresh, edible portion of an herbaceous plant consumed in either raw or cooked form. vegetable crops can be classified as fruit vegetables such as tomato, cucumber, watermelon, peas; root and tuber/root vegetables such as carrot, adeniyi adedapo impact of climate variability on vegetable crops in ilorin ruhuna journal of science vol 8: 44-54, june 2017 45 potato, sweet potato, radish, elephant foot yam; green leafy vegetables such as amaranthus, celery, cabbage, curry leaf and bulb vegetables such as small onion, ballery onion, and garlic. vegetables are a rich source of vitamins, carbohydrate, salts and protein. they are the best resources for overcoming micronutrient deficiencies and provide small holder farmers with much higher income and more jobs per hectare than staple crops. the worldwide production of vegetables has doubled over the past quarter century and the value of global trade in vegetables now exceeds that of cereals (bhardwaj 2012). among vegetable crops, tomato is the most important vegetable crop worldwide and is grown over a four million hectares of land area (bhardwaj 2012). climate is one of the fundamental factors that determine the agricultural productivity of an area. it is so fundamental that it affects virtually all aspects of crop production (adeniyi 2013). the production and yield of agricultural crops are climate determined. according to anju et al. (2014) climate variability has significant impacts on agricultural sector especially during the last 40 year period. climatic variability has become an issue today with significant effects on agricultural productivity. climate determines the quality and quantity of crop productivity. olanrewaju (2012) declares that many of the problems facing agricultural products are climate related. climatic variability poses a threat on agriculture especially in an environment where agriculture is rain fed. ernest (2002) stated that agricultural production depends on climatic variables such as temperature, precipitation and light. adebayo (2010) observed that high temperature can slow down or completely stop the process of photosynthesis. yusuf (2012) also reported that excessive heat has desiccating effects on plants and rapid rates of transpiration can cause wilting. vegetables crops, like other agricultural crops, are sensitive to climate variability. according to bhardwaj (2012) vegetables are generally sensitive to environmental extremes, and thus high temperature are the major causes of low yields and will be further magnified by climate change. he also noted that increasing temperature, reduced irrigation water availability, flooding and salinity will be major limiting factors in sustaining and increasing vegetable productivity. global climatic change, especially erratic rainfall pattern and unpredictable high temperature spells, will reduce the productivity of vegetable crops. high temperature stress disrupts the biochemical reactions fundamental for normal cell function in plants. it primarily affects the photosynthetic function of higher plants (bhardwaj 2012). high temperature can cause significant loss in tomato productivity due to reduced fruit set, and smaller and lower quality fruits (bhardwaj 2012). yusuf (2012) also reported that environmental factors have negative effects on tomato production. thakur and jahn (2012) stated that continuous declining weather conditions and changes in climate due to the escalating temperature, erratic rainfall, more adeniyi adedapo impact of climate variability on vegetable crops in ilorin ruhuna journal of science vol 8: 44-54, june 2017 46 demand for water and enhanced incidence of diseases are all set to affect the production of various vegetable crops. adeniyi (2013) reported that rainfall is one of the most important factors affecting crop production. bhardwaj (2012) also observed that water greatly influence the yield and quality of vegetables; drought conditions drastically reduced vegetable productivity and tomatoes in particular are considered to be one of the vegetable crops most sensitive to excess water. from the above studies, climatic variables, especially temperature, precipitation and sunshine hours affect the productivity of vegetable crops. this, therefore, suggests that to improve the production and the yield of vegetable crops climate-crop relationship must be taken into consideration. it is against this background that this paper is put forward to ascertain the impact of climatic variability on vegetable crops in ilorin. 2 material and methods 2.1 study area ilorin, the capital of kwara state is located on latitude 8o24’n and 8o36’n and longitude 4o10’e and 4o 36’e. it is situated at a strategic point between the densely populated south-west and the sparsely populated middle belt of nigeria. ilorin is located in the transition zone between the deciduous woodland of the south and dry savannah of north nigeria (jimoh 2003). figure 1 shows the map of kwara state and the study area. the climate of ilorin is characterized by both wet and dry seasons. the rainy season begins towards the end of april and last till october while the dry season begins in november and ends in april. the temperature of ilorin ranges from 33oc to 35oc from november to january while from february to april; the value ranges between 34oc to 37oc. days are very hot during the dry season. the diurnal range of temperature and the mean monthly temperatures are characteristically high in the area. the total annual rainfall in the area ranges from 990.3mm to 1318mm. rainfall in ilorin city exhibits the double maximal pattern and greater variability both temporarily and spatially. the relative humidity at ilorin city ranges from 75% to 88% from may to october, while in the dry season it ranges from 35% to 80%. the geology of ilorin consists of precambrian basement complex rock. the soils of ilorin are made up of loamy soil with medium to low fertility. because of the high seasonal rainfall coupled with the high temperature, there is tendency for lateritic soil to constitute the major soil types in ilorin due to the leaching of minerals nutrients of the soil (ajibade and ojelola 2004). the elevation of the area varies from 273m to 333m in the western side with isolated hill (sobi adeniyi adedapo impact of climate variability on vegetable crops in ilorin ruhuna journal of science vol 8: 44-54, june 2017 47 hill) of about 394m above the sea level while on the eastern side it varies from 273m to 364m (ajibade and ojelola, 2004). the lowest level is along the river valley of asa and oyun while the highest point is sobi hill. fig.1: map of nigeria showing kwara state and the study area (source: ministry of land and urban development) ilorin is mainly drained by asa river which flows in a south-north direction (ajibade and ojelola 2004). the pattern of the drainage system of ilorin is dendritic. asa river occupies a fairly wide valley and goes a long way to divide ilorin into two parts; namely the eastern and the western part. the eastern part covers those areas where the gra is located while the core indigenous area of ilorin falls under the western part. other rivers in ilorin that drain into asa river are river agba, river alalubosa, river okun, river osere, river aluko, river yalu, river odota and river loma. the soils of ilorin are easy to farm. they are loamy soil with low sodium and low fertility. because of the high seasonal rainfall coupled with the high temperature, there is tendency for lateritic soil to constitute the major soil types due to the leaching of mineral nutrients of the soil (ajibade and ojeola 2004). the socio-economic activities in ilorin have increased tremendously from agricultural practices of growing food crops to local craft of cloth weaving, leather works, pottery, embroidery, tie and dye, mat making, etc., to modern adeniyi adedapo impact of climate variability on vegetable crops in ilorin ruhuna journal of science vol 8: 44-54, june 2017 48 commerce with viable trading industry and administrative activities (olorunfemi 2001). agricultural activities in ilorin are limited to small garden plots of maize, beans, and vegetables which are cultivated mainly for domestic consumption. cultivation of tuber crops like yam and cassava are mainly done at the outskirts of the city. both climate and soil of ilorin support the growth of vegetable crops. 2.2 climatic data there are many ways of estimating the impact of climatic variability on agricultural productivity. these include the study of radiation and moisture balance of various crops under different climatic environment and the study of the relationship between plant and climate in a controlled environment (olaniran 1981). however, this study employs the method of analyzing agricultural and climatic data in ilorin over a period of ten years (2002-2011) to determine if there are strong correlations between climate factors and crop production. climatic data on temperature, rainfall, relative humidity and sunshine hours were collected from ilorin international airport, ilorin for the period of ten years. these climatic variables were selected based on the important role they play in the production and yield of vegetable crops. agricultural data on sweet potato, tomato, pepper and amaranthus were also collected from kwara state agricultural development project office, ilorin. the selected vegetable crops are the common and major vegetable crops grown in the area. descriptive and inferential statistics were employed in data analysis. simple correlation and multiple regression were used in showing the relationship between climatic parameters and vegetable yield, mannkendull statistics was used in showing the trend and variations in vegetable yield while standardize anomaly index was used in the analysis of fluctuations in climatic parameters over ten years in the study area. statistical package for social science (spss) was used in the analysis. 3 results and discussion 3.1 patterns of agricultural data in ilorin (2002-2011) table 1 shows the descriptive analysis of the agricultural data on vegetable crops in ilorin. out of the five selected vegetable crops, sweat potato has the highest mean value (8.43 tons). this was followed by amaranthus (4.43) while pepper has the lowest mean value (2.62). this implies that within the years under review, sweat potato has the highest yield value. the highest deviation was obtained in tomato production (1.70). the dispersion adeniyi adedapo impact of climate variability on vegetable crops in ilorin ruhuna journal of science vol 8: 44-54, june 2017 49 characteristics of sweat potato, okra, and amaranthus are low. the coefficient of variation which shows the relative deviation between crop yields indicated that tomato and pepper are heterogeneous with values greater than 33% while that of sweat potato, okra and amaranthus are homogeneous with values less than 33%. this suggests that the values of tomato and pepper produced in the area differ significantly. the relative deviation in the yield of the crops could be a result of climate, soil fertility or crop management techniques. table 1: descriptive analysis of agricultural data (2002-2011). crop mean (yield) (tons) standard deviation co-efficient of variation sweet potato 8.43 0.72 8.49 okra 4.06 0.49 12.09 tomato 3.58 1.70 47.52 pepper 2.62 1.10 41.91 amaranthus 4.43 0.30 6.68 source: own data computation 3.2 patterns of climatic data in ilorin (2003-2013) figures 2-5 show the fluctuations of the selected climatic variables in ilorin. from figure 2, the annual temperature fall below the long term average in 2004 to 2006 and 2011. however, it falls above the long term average in 2003 to 2008 to 2010. this suggests a decline in temperature in 2004 to 2006 and 2011. the result also shows that temperature fluctuate around the long term average. from figure 3, the annual rainfall fall above the long term average in 2003 to 2009 while the annual relative humidity falls above the long term average in 2004, 2006 and 2009 to 2011 (figure 4). similarly, from figure 5, the annual sunshine hours fall above the long term average in 2006, 2008, and 2010 to 2011. these suggest that there is an increase in rainfall, relative humidity and sunshine hours in those years. from figure 3, rainfall has the longest consecutive years of annual rainfall above the long term average. the above results indicate that the selected climatic variables fluctuate around the long term mean. adeniyi adedapo impact of climate variability on vegetable crops in ilorin ruhuna journal of science vol 8: 44-54, june 2017 50 fig 2. annual temperature fluctuation for ilorin (20022011). fig 3. annual rainfall fluctuation for ilorin (20022011). -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 s ta n d e r d iz e d a n o m a ly in d e x years -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 s ta n d a r d iz e d a n o m a ly in d e x year adeniyi adedapo impact of climate variability on vegetable crops in ilorin ruhuna journal of science vol 8: 44-54, june 2017 51 fig 4. annual relative humidity fluctuation for ilorin (20022011). fig 5. annual sunshine hour fluctuation for ilorin (20022011). 3.3 relationship between climatic variables and crop yield the result of the regression analysis shows that 46.5%, 32.9%, 44.7%, 26.7% and 79.7% of the variance in sweat potato, okra, pepper, amaranthus and tomato yields can be respectively explained by the climatic parameters (table 2). this implies that the impact of climate variability on variation of vegetable -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 s ta n d a r d iz e d a n o m a ly in d e x year -2 -1.5 -1 -0.5 0 0.5 1 1.5 s ta n d a r d iz e d a n o m a ly in d e x year adeniyi adedapo impact of climate variability on vegetable crops in ilorin ruhuna journal of science vol 8: 44-54, june 2017 52 crops in ilorin over the years under review is low except that of tomato. this also implies that variation in the yield of sweat potato, okra, pepper and amaranthus could be attributed to non-climatic factors like soil fertility while variation in tomato yield could be attributed to climate variability. table 2. statistical relationship between climate and crop yield. crop r r2 regression coefficient standard error f significant sweat potato 0.682 0.465 -10.525 31.901 1.08 0.453 okra 0.574 0.329 9.550 24.513 0.613 0.672 pepper 0.669 0.447 -65.781 49.794 1.01 0.481 amaranthus 0.516 0.267 -2470.779 8103.555 0.454 0.768 tomato 0.893 0.797 -117.355 45.147 4.899 0.056 source: own data computation 3.4 crop climate correlation the simple correlation coefficient (r2) between the selected climatic variables and the selected vegetable crops were computed (table 3). the result shows that the correlation values of temperature and sweet potato, okra, pepper and tomato. this implies that temperature has weak correlation with the selected crops. rainfall also has a weak correlation with all the crops. relative humidity and sunshine hour have a weak correlation with pepper and amaranthus. however, relative humidity and sunshine hours have a strong correlation with tomato. the result also shows that rainfall has a negative correlation with all the crops except amaranthus. the implication is that as rainfall increases, the values of the crop yield reduce except that of amaranthus. table 3. climate – crop correlation crop temperature rainfall (mm) relative humidity sunshine hours sweet potato 0.181 -0.296 0.365 0.618 okra -0.185 -0.118 0.512 0.085 pepper 0.885 -0.384 0.388 0.314 amaranthus 0.115 0.125 -0.129 -0.467 tomato 0.449 -0.328 0.561 0.684 source: own data computation adeniyi adedapo impact of climate variability on vegetable crops in ilorin ruhuna journal of science vol 8: 44-54, june 2017 53 3.5 trend in crop yield table 4 shows the result of the trend analysis using mann-kendull method. the result shows that there is no significant difference or decline in the values of the crop yield at 95% or 99% probability levels. this implies that no differential pattern of variation exists in yield of selected vegetable crops in each successive year under study. table 4: trend analysis of crop yield crop (r)t sweet potato 0.73 okra 0.42 tomato 0.78 pepper 0.47 amaranthus 0.24 source: own data computation 4 conclusion and recommendations the study examines the impact of climatic variability on vegetable crops in ilorin. the results of the regression analysis show that climate has little impact on the selected vegetable crops except tomato. this implies that variation in the climatic variables have little effect on sweet potato, pepper, okra, and amaranthus. the results, therefore, suggest that variation in vegetable yield could be as a result of other non-climatic factors. such factors could be soil fertility, disease and insect pests or farm management techniques. based on the above findings, we recommend additional research to determine if the application of fertilizer can improve soil fertility resulting in higher crop production, if the use of insecticides to control crop pest and diseases will impact crop yield, and if practicing modern agricultural techniques can improve the yield of vegetable crops. references adebayo aa. 2010. climate: resource and resistance to agriculture: eighth inaugural lecture delivered at federal university of technology yola, 19th may, 2010 adeniyi a. 2013. impact of climate on productivity of selected crops i ilorin, kwara state, nigeria. ilorin journal of business and social sciences 15 (1): 59-66 ajibade lt, ojelola ao. 2004. effects of automobile mechanics activities on soil in ilorin, nigeria geo-studies forum an international journal of environmental and policy issues 2 (1): 18-27 adeniyi adedapo impact of climate variability on vegetable crops in ilorin ruhuna journal of science vol 8: 44-54, june 2017 54 anju l, ambily pg, gopikrishna vg, amairaj m. 2012. a study on the scope and importance of tuber crops with special reference to cassava as resilient crop towards climate change. earth science and climate change 5 (6): 1-6 bhardwaj ml. 2012. effect of climate change on vegetable production in india in vegetable production under changing climate scenario edts by bhardwaj, m.l, sharma, h.d., kumar, m., kumar, r., kansal, s., thakur, k. singh, s.p., kumari, d., kumari, s., gupta, m. and sharma, v. ernest lm. 2002. climate variability, vulnerability and effectiveness of farm-level adaptation options: the challenges and implications for food security in southwestern cameroon. environmental and development economics. 7: 529-545 jimoh hi. 2003. erosion tolerance range of landuse surface: implication on land resource use and management techniques in ilorin, nigeria, international journal on environmental studies, 60 (5): 445-452 olaniran oj. 1981. research in agroclimatology in nigeria, journal of agric. research, 19 (1): 15-29 olanrewaju rm. 2012. effect of climate on yam production in kwara state, nigeria environmental issues 3 (1) yusuf ro. 2012. coping with environmentally induced change in tomato production in rural settlement of zuru local government area of kebbi state. environmental issues 5 (1): 47-54 ward aw. 2016. encyclopedia britannica. sv-lncs ruhuna journal of science vol 9(2): 160-168, december 2018 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v9i2.44 university of ruhuna  faculty of science, university of ruhuna 160 short paper effects of partially purified enterocins from enterococcus faecalis strains on the growth of some phytopathogenic fungi o.m. david1* and o.e. onifade2 1department of microbiology, ekiti state university, ado-ekiti, nigeria 2department of science laboratory technology, ekiti state university, ado-ekiti, nigeria *corresponding author: david.oluwole@eksu.edu.ng; orcid: 0000-0002-1396-3450 received: 29th may 2018, revised: 12th november 2018, accepted: 22nd november 2018 abstract. plant protection is an important area which needs attention since most of the hazardous inputs added into agricultural systems are in form of synthetic chemicals. the inhibitory activity of partially purified enterocins (ppes) produced by enterococcus faecalis strains on plant pathogenic fungi was investigated in this study. the ppes were preliminarily screened against bacteria using agar-well diffusion method. the active extracts were partially purified using ion exchange chromatography. the in vitro anti-fungal properties of the ppes were determined using agar dilution and broth dilution techniques. the ppes tested in this study inhibited the growth of botryodiplodia theobromae, aspergillus niger, pythium ultimum, penicillium expansum and fusarium oxysporum. at different concentrations ppes had varying inhibitory effects on the dry mycelial weight of pythium ultimum and f. oxysporum. at the 96th hour of the experiment, enterocin unad 012 had higher percentage inhibition ranging between 37.63 and 84.11% than enterocin unad 046 with percentage inhibition ranging between 28.77% and 67.27% on the test fungi. this inhibitory activity of enterocins produced by e. faecalis on fungi makes them as potential biocontrol agents due to their ability in suppressing their growth. keywords. bacteriocin, enterocins, enterococcus faecalis, fungi, phytopathogens. 1 introduction phytopathogenic fungi are capable of causing infectious diseases in plants. they damage plants and plant product on which human beings depend for http://doi.org/10.4038/rjs.v9i2.44 o.m. david and o.e. onifade effects of enterocins on phytopathogenic fungi ruhuna journal of science vol 9(2): 160-168, december 2018 161 food, clothing, shelter, furniture and the environment. most of them belong to the family ascomycetes and basidiomycetes. common species include pythium ultimum, penicillium expansum, fusarium oxysporum. aspergillus fumigatus, botryodiplodia theobromae and phytophthora spp. (aderiye et al. 1996, fagbohun et al. 2008). enterococcus is a lactic acid bacteria (lab) found in gastrointestinal flora, oral cavity and human vagina. they are widespread in nature and have been detected in the fecal samples from humans, lower vertebrates and insects (david et al. 2012). enterococci has been reported to produce bacterocins; an extracellular macro-molecular protein/peptides which exert a lethal effect on bacteria or the related groups (papagiani et al. 2004). bacteriocins as antimicrobial peptides could be a better replacement to chemical fungicides. all species of enterococci are capable of producing bioactive bacteriocins named as enterocin (gilmore et al., 2002). bacteriocins have been reported to act against both related species and distantly related genera (vidaver et al. 1972, okkers et al. 1999). they act on food-borne pathogenic and spoilage micro-organisms and in the recent time, their activity against plant pathogens was reported (schillinger et al. 1996). the potential of bacteriocin from b. subtilis has antagonistic and bactericidal effects on agrobacterium spp., the causative agent of crown gall. the proprieties of bacteriocins indicated that they have a strong potential to be used in biological control of crown gall disease (hammami et al. 2009). fungi have been reported to cause numerous diseases in plants. some of the chemicals used to control these diseases bio-accumulate in the plants and eventually enter food chains. current campaign for the fungicide-free fruits and vegetables products, and rise in fungal resistance to common chemocontrol agents necessitate the search for alternative control methods for mycophyto-pathogens. in this study we evaluated the anti-fungal ability of entrocins produced by two strains of enterococcus faecalis. the antimicrobial spectrum and some properties of the bacteriocins are described and their antifungal properties against some phytopathogenic fungi were also studied. 2 materials and methods 2.1 preparation of cell free supernatant (cfs) two enterococcus faecalis strains were collected from the stock cultures maintained in the department of microbiology, ekiti state university, adoekiti, nigeria. the organisms were separately revived in de man rogosa and sharpe (mrs) broth. the broth was incubated at 37oc for 24 h after which it was centrifuged for 10 min at 10,000 g at 4ºc. the supernatant was decanted gently and later filtered through a membrane filter with a pore size of 0.22 o.m. david and o.e. onifade effects of enterocins on phytopathogenic fungi ruhuna journal of science 162 vol 9(2): 160-168, december 2018 μm. the interfering effects of peroxides and organic acids in the cfs were eliminated by addition of 1 n naoh and 130 u/ml of catalase (sigma chemical co., st. louis, mo, usa) respectively. 2.2 determination of antibacterial activity of cfs the bacteria used in the primary screening of the enterocin include bacillus subtilis, escherichia coli, klebsiella pneumoniae and staphylococcus aureus. they were collected from the department of microbiology, ekiti state university, ado-ekiti. the organisms were grown for 18 h at 37oc and the turbidity adjusted to 0.5 mcfarland standard. antibacterial potential of the cfs was determined using agar-well diffusion assay. the reciprocal of least serial dilution of cfs with antibacterial activity was taken to be activity unit (au) as described by david et al. (2017). 2.3 partial purification of enterocin by ammonium sulphate precipitation to saturation level, ammonium suphate (ranging between 60 and 90 %) was added to 50 ml of csf with constant stirring and kept overnight at 4oc. the solution was centrifuged at 10,000 g for 20 min at 4oc and later dissolved in 500 ml of 20 mm sodium phosphate buffer (ph 5.0). the supernatant was stored at 4oc until used. 2.4 determination of protein content of the enterocin produced the protein content of the cfs was determined according to bradford (1976). the optical density of each of the samples was calculated from the bestfit equation line obtained from the graph of the bovine serum albumin (bsa) standard curve. 2.5 source of phytopathogenic fungi fungi isolates primarily isolated from infected plants were collected from the stock cultures maintained at the department of microbiology, ekiti state university, ado-ekiti, nigeria. the fungi include aspergillus niger, botrydiplodia theobromae, fusarium oxysporum, penicillium expansum and pythium ultimum. the test fungi were maintained on slants of potato dextrose agar at 4oc until use. o.m. david and o.e. onifade effects of enterocins on phytopathogenic fungi ruhuna journal of science vol 9(2): 160-168, december 2018 163 2.6 determination of antifungal property of enterocin the partially purified enterocin was sterilized by filtering it through filters with 0.22 μm pore size and the 2 ml of the filtrate was added into 10 ml of sterile potato dextrose broth. the broth was incubated at 37°c for 24 h and the sterile enterocin did not produce any turbidity. the anti-fungal activity of the different extracts of the partially purified enterocins was determined according to poisoned food assay method described by nene and thapilyal (2002). at the right concentrations, the sterile extract was mixed with sterilized potato dextrose agar (pda) just before the setting of the agar. agar plug (10 mm) from the advancing edge of five-day culture of each of the test fungi was inverted on the center of each plate and incubated at 25°c for 96 h. the pda plate without enterocin was also maintained at the same condition to serve as the control and the experiment was performed in triplicate. the diameter of fungal colony was measured to the nearest centimeter. 2.7 analysis of data results of this study were presented as the mean values of the replicates. oneway analysis of variance (anova) was carried out using spss 16.0. significance was accepted at p ≤ 0.05. 3 results and discussion out of four isolates screened for bacteriocinogenic potential, only two of the bacteriocin-producing strains (unad 012 and unad 046) showed a prominent activity against the test organisms (table 1). table 1. antibacterial activity (inhibition zones in mm) of crude enterocin produced by strains of enterococcus faecalis. test organisms enterocins from e. faecalis strains unad 012 unad 046 unad 019 unad 033 gram positive b. subtilis 10 36 10 s. aureus 10 15 10 10 gram negative e. coli 15 16 10 k. pneumoniae 19 10 10 o.m. david and o.e. onifade effects of enterocins on phytopathogenic fungi ruhuna journal of science 164 vol 9(2): 160-168, december 2018 table 2. activity, protein concentration, yield and fold of selected enterocin. parameter purification steps crude (nh4)2so4 ion exchange unad 012 unad 046 unad 012 unad 046 unad 012 unad 046 volume (ml) 10 10 5 6 5 2 activity (au/ml) 122 158 145 207 95 125 protein conc. (mg/ml) 15.8 17.2 6.2 6.0 1.4 1.9 total activity (au) 1220 1570 725 1102 285 263 total protein (mg) 158 148 31 48 4.2 5.2 specific activity (au/mg) 7.7 9.9 23.3 20.3 67.8 40.4 yield % 100 100 25.4 59.0 3.4 18.4 purification fold 1 1 3 2.04 8.7 2.9 fig. 1. elution profile of bacteriocin unad 012 deduced from the determination of bacteriocin activity. the zone of inhibition ranged between 19 and 36 mm against the test organisms. enterocins unad 012 and unad 046 have better activity against gram negative and gram positive bacteria, respectively. enterocin produced by strain unad 033 had the least effect on the isolates. enterocins have been reported to inhibit bacteria (laukova et al. 1993, casula and cutting 2002, foulquie et al. 2003). in this study, four test bacteria were used at the primary screening stage of enterocin production. the bacteria were used to determine o.m. david and o.e. onifade effects of enterocins on phytopathogenic fungi ruhuna journal of science vol 9(2): 160-168, december 2018 165 the potency of the bacteriocins produced by the strains of e. faecalis. table 2 shows the effects of purification on specific activity of two promising enterocinogenic producing e. faecalis. the specific activity increased with purification processes while a decrease was noticed in the yield, protein concentration and total protein of the enterocins. this observation was comparable to the findings of whitford et al. (2001). the elution profile of bacteriocins deduced from the determination of bacteriocin activity was represented in figures 1 and 2. fig. 2. elution profile of bacteriocin unad 046 deduced from the determination of bacteriocin activity. compared with the control, enterocin unad 012 had a significant effect on p. ultimum at p<0.05. at p<0.05 significant level, the growth of the fungi at 24h differs significantly from those of 72h and at 96h. at the same significant level, the difference of the growth of the fungi at 48 h differs from the growth at 96h. there was a significant difference (p<0.05) on the percentage inhibition of the effects on the enterocin on the test fungi except unad 012 on p. ultimum (table 3). as shown in table 4, the percentage inhibition of the fungi increased with time of exposure to the enterocins. o.m. david and o.e. onifade effects of enterocins on phytopathogenic fungi ruhuna journal of science 166 vol 9(2): 160-168, december 2018 table 3. antifungal activities of enterocins, at their arbitrary units (au), on selected fungi isolates (radial mycelial in cm). test organisms enterocins time (h) 24 48 72 96 control 3.30±1.79 5.85±1.78 9.80±3.78 12.65±3.31 p. expansum unad 012 2.85±1.02 3.25±1.03 7.80±2.56 7.89±3.97 unad 046 1.85±0.45 4.25±1.99 8.80±2.78 9.01±3.78 b. theobromae unad 012 1.59±0.89 2.07±1.02 2.90±1.45 3.02±1.34 unad 046 2.73±1.06 3.00±1.74 8.56±3.97 9.00±3.45 a. niger unad 012 1.10 ±0.98 2.50±1.52 3.50±0.46 3.52±1.49 unad 046 2.10±1.16 3.75±1.48 4.65±1.66 5.97±3.34 f. oxysporum unad 012 1.50±0.48. 2.20±1.93 4.50±2.09 5.81±3.39 unad 046 1.10 ±0.41 2.05±1.68 2.65±1.88 4.14±1.59 py. ultimum unad 012 1.50±0.56 1.50±1.46 1.52±1.34 2.01±1.78 unad 046 1.50±0.91 4.00±1.33 4.80±2.94 5.17±3.97 table 4. percentage inhibition of the enterocins on test fungi. test organisms enterocins time (h) 24 48 72 96 control 0 0 0 0 p. expansum unad 012 13.64 44.44 20.41 37.63 unad 046 43.94 27.35 10.20 28.77 b. theobromae unad 012 51.82 64.62 70.41 76.13 unad 046 17.27 48.72 12.65 28.85 a. niger unad 012 66.67 57.26 64.29 72.17 unad 046 36.36 35.90 52.55 52.81 f. oxysporum unad 012 54.55 62.39 54.08 54.07 unad 046 66.67 64.96 72.96 67.27 p. ultimum unad 012 54.55 74.36 84.49 84.11 unad 046 54.55 31.62 51.02 59.13 compared with the control, unad 046 had a better inhibitory effect on b. theobromae, a. niger, penicillium expansum and p. ultimum than unad 042. on the other hand, f. oxysporum was more susceptible to unad 046 than unad 012. these results were similar to those of aruna and madhuri (2016), schillinger et al. (1996) reporting the susceptibility of different fungi (spoilage and pathogenic) to enterocins. smaoui et al. (2010) reported that lactobacillus spp. produce bacteriocins that are active against gram-negative bacteria and also particularly inhibit fungi. bacteriocin has been proposed to be a promising treatment of plant infections, and its application has been reported to be safe to animals and humans (cleveland et al. 2001, ogunbanwo o.m. david and o.e. onifade effects of enterocins on phytopathogenic fungi ruhuna journal of science vol 9(2): 160-168, december 2018 167 et al. 2004, cole et al. 2006). very few bacteriocins with antifungal properties have been reported and most enterocins studied have bacteriostatic and bacteriocidal activities on food-borne bacteria pathogens and not mould (suzuki et al. 1991). 4 conclusion from this study, we observed that partially purified enterocins produced by e. faecalis had inhibitory spectrum on selected phytopathogenic fungi. enterocin from enterococcus faecalis could be a good candidate for biocontrol of phytopathogenic fungi. nevertheless, more studies need to be done to further validate the results of this report. acknowledgements the authors appreciate dr. bamidele femi for his technical assistance in this study and also the technologists in the department of microbiology, ekiti state university for their assistance. comments on the initial manuscript from two anonymous reviewers are acknowledged. references aderiye bi, ogundana sk, adesanya sa, robert mf. 1996. antifungal properties of yam (discorea alata) peel extract. folia microbiology 4(5):407-412 aruna b, madhuri gsj. 2016. antifungal activity of bacteriocin produced by lactic acid bacteria from fermented green gram. international journal of science and research 5(8): 1824-1831. bradford mm. 1976. a rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. analytical biochemistry 72: 248254. casula pt, cutting tr. 2002. enterocin b, a new bacteriocin from enterococcus faecium t136 which can act synergistically with enterocin. african journal of microbiology 143: 22872294. cleveland j, montville tj, nes if, chikindas ml. 2001. bacteriocins: safe, natural antimicrobials for food preservation. international journal of food microbiology 71:1-20 cole k, farnll mb, donigbue am, stern nj, svetoch ea, eruslanov en, volodina li, kovalev yn, perelygin vv, mitsevich ev, mitsevich ip, levchuk vp, pokhilenko vd, borzenkov vn, svetoch oe, kudryavtseva ty, reyes-herrera i, blore pj, solis de los santos f, donogbue dj. 2006. bacteriocins reduce campylobacter colonization and alter gut morphology in turkey. poultry science 85:1570-1575. david om, alese mo, komolafe dm, adejare ij, alese oo, omonisi ae. 2017. in vitro and in vivo antimicrobial activity of partially purified enterocin produced by enterococcus faecalis and its application in wound healing. african journal of clinical and experimental microbiology 18(1): 1-10. o.m. david and o.e. onifade effects of enterocins on phytopathogenic fungi ruhuna journal of science 168 vol 9(2): 160-168, december 2018 david om, oluduro ao, famurewa o. 2012. property and antibacterial spectrum of partially purified enterocin produced by entrocinogenic enterococcus faecalis isolated from the gut of cockroach. au journal of technology 16(2): 74-80. fagbohun de, falegan cr, arowolo ja. 2008. phytophthora black pod disease in ekiti state: species differentiation and pathogenic variation. bioscience, biotechnology research asia 5(1): 157-160. foulquie mnr, callewart db, van beeumen j, de vuyst l. 2003. isolation and biochemical characterization of enterocins produced by enterococcus from different sources. journal of applied microbiology 94:214-229. gilmore ao, riley ma, wertz je. 2002. bacteriocins: evolution, ecology and application. annual review of microbiology 56: 117-137 hammami i, rhouma a, jaouadi b, rebai a, nesme x. 2009. optimization and biochemical characterization of a bacteriocin from a newly isolated bacillus subtilis strain 14b for biocontrol of agrobacterium spp. strains. letters in applied microbiology 48: 253–260. laukova a, marekova m, javorsky p. 1993. detection and antimicrobial spectrum of a bacteriocin-like substance produced by enterococcus faecium ccm4231 journal of applied microbiology 16: 257-260. nene y, thapilyal l. 2002. poisoned food technique of fungicides in plant disease control. 3rd ed, oxford and ibh publishing company, new delhi. ogunbanwo st, sanni ai, onilude aa. 2004. influence of bacteriocin in the control of escherichia coli infection of broiler chickens in nigeria. world journal of microbiology 20: 51-56. okkers dj, dicks lmt, silvester m, joubert jj, odendaal hj. 1999. characterization of pentocin tv35b, a bacteriocin-like peptide lactobacillus pentosus isolated from with a fungistatic effect on candida albicans. journal of applied microbiology 87: 726–734 papagiani m. 2004. ribosomally synthesized peptides with antimicrobial properties: biosynthesis, structure, function and applications. journal of advanced biotechnology 21: 465-499. schillinger ur, geisen r, holzapfel wh. 1996. potential of antagonistic microorganisms and bacteriocins for the biological preservation of foods. trends food science and technology 7: 158-164. smaoui s, elleuch l, bejar w, karray-rebai i, ayadi i, jaouadi b, mathieu f, chouayekh h, bejar s. mellouli l. 2010. inhibition of fungi and gram-negative bacteria by bacteriocin bactn635 produced by lactobacillus plantarum sp. tn635. applied chemistry and biotechnology 1132-1146. suzuki i, nomura m, morichi t. 1991. isolation of lactic acid bacteria which suppress mould growth and show antifungal action. milchwissenschaft 46: 635–639. vidaver ak, mathys ml, thomas me, schuster ml. 1972. bacteriocins of the phytopathogens pseudomonas syringae, p. glycinea, p. gtycinca and p. phaseolicola. canadian journal of microbiology 18(6): 705-713. whitford wf, mcpherson ma, foster rj, teather rm. 2001. identification of bacteriocin-like inhibitors from rumen streptococcus spp. and characterization of bovicine 255. journal of applied microbiology 67(2): 569-574. sv-lncs ruhuna journal of science vol. 7: 3242, december 2016 eissn: 2536-8400  faculty of science university of ruhuna  faculty of science, university of ruhuna sri lanka 32 study on the physico-chemical properties, sensory attributes and shelf life of pineapple powder incorporated biscuits m. thivani, t. mahendran 1 and m. kanimoly 2 department of agricultural chemistry, faculty of agriculture, eastern university, sri lanka correspondence: 1 thevamahen@yahoo.com 2 kanimolym@yahoo.com received: 5 th september 2016, revised: 26 th november 2016, accepted: 21 st december 2016 abstract. the bakery industry is growing very fast and the products are becoming increasingly popular among consumers world-wide. among ready-to-eat snacks, biscuits possess several attractive features including wider consumption base, relatively long shelf-life and preferred eating quality. pineapples constitute important natural and valuable material in producing functional foods due to the presence of several anti-oxidants and bioactive compounds. therefore, the aim of the present study was to investigate the nutritional composition and sensory quality of wheat flour biscuits incorporated with pineapple powder at the rates of 3, 5, 10, and 15% (w/w basis). the protein, fat, total sugar, moisture and mineral contents were significantly (p<0.05) changed during storage. nutritional analysis indicated that biscuits incorporated with 5% pineapple powder had a slow rate of increase in moisture content (4.19 to 5.0%) and decreasing trend in protein (22.15 to 22.06%), fat (18.83 to 18.79%), total sugar (58.52 to 58.34%) and mineral contents (2.212 to 1.591%) compared to other treatments during storage. among the treatments, the biscuits prepared with 5% pineapple powder had the highest nutritional and sensory quality, having the overall acceptability score of 7.7 in a 9-point hedonic scale. the shelf life evaluation showed that these biscuits could be stored for 6 weeks at the ambient conditions of average temperature at 301ºc and rh at 7580% with acceptable quality. keywords: biscuits, physico-chemical properties, pineapple powder, sensory attributes, shelf life 1 introduction biscuits, which are leavened baked products, are one of the most delicious foods preferred by most people. biscuits represent the largest category of snack item among bakery products (pratima and yadav, 2000). they are stable foods and have advantages such as ready-to-eat form, wide m. thivani, t. mahendran and m. kanimoly quality and shelf life of pineapple powder biscuits ruhuna journal of science vol. 7: 32-42, december 2016 33 consumption, long shelf life and eating quality (siddiqui and nasreen, 2014). biscuits are widely accepted and consumed by almost all profiles of consumers in many countries and therefore offer a valuable supplementation vehicle for nutritional improvement (arshad et al., 2007). currently, cookies are prepared from composite flour or fortified with some other good sources of nutrients (wani et al., 2015). modifications of basic recipes and incorporation of new ingredients such as fibers, fat replacers, cereals other than wheat, etc. have led to novel biscuit formulations with improved functionality and nutritive value (handa et al., 2012; yadav et al., 2012). fruits are produced in considerable quantities and consumed locally, but are seldom processed in order to add value. fruits exhibit relatively high metabolic activity compared with other plant derived foods such as seeds and tubers. these metabolic activities continue after harvesting, thus making most fruits highly perishable commodities (offia-olua and ekwunife, 2015). therefore, there is need for diversity in commercial utilization of fruits into different forms. there are numerous ways of utilizing and processing fruits such as processing into juice, jams, concentrates, jellies and dehydrated products. the introduction of fruit based composite flour is novel, as recently, fruits and vegetables have received much attention as a source of biologically active substances because of their anti-oxidant, anti-carcinogenic and antimutagenic properties (tortoe et al., 2014). pineapple (ananas comosus) is cultivated in the tropical and subtropical regions of the world and is well known for its attractive sensorial and nutritional characteristics. further, it is a good source of carotene and ascorbic acid and rich in vitamin b1; it is also containing carbohydrate, protein, fiber, calcium and iron (ade et al., 2015). fresh pineapple is often expensive because of its delicate nature. pineapples continue to ripen at appropriate temperatures after harvesting. the shelf life of ripe pineapple is short and limited to 4-6 days (kadam et al., 2012) and the average yield of processing ranges from 45-55%. therefore, an alternate solution is the development of a pineapple powder from pineapple pulp as a value added ingredient for the bakery and confectionary industries (deliza et al., 2005). pineapple is commonly found in sri lankan market and is being consumed raw and also processed into various food products, including jam, jellies, juices, puree, powder and nectar. pineapple fruit is used for the preparation of pineapple powder because of difficulty in ready consumption as fresh due to their big size, hardy nature and also to prevent postharvest losses. production of biscuits incorporated with pineapple powder lead to innovations in new product development. therefore, the objective of the present study was to develop pineapple powder incorporated biscuits and to assess the physico-chemical properties and sensory attributes in the formulated biscuits during storage at ambient conditions. m. thivani, t. mahendran and m. kanimoly quality and shelf life of pineapple powder biscuits ruhuna journal of science 34 vol. 7: 32-42, december 2016 2 materials and methods 2.1 preparation of pineapple powder fully ripe and fresh pineapples (variety: mauritius) were selected, washed, peeled and their thorny eyes were removed. they were sliced lengthwise; each slice was cut into 2 cm thickness. the slices were dipped in to the potassium metabisulphite solution (0.02%) for 5 minutes to prevent discoloration. these slices were placed in a single layer on stainless steel trays. the vacuum oven (model dzf 6020-w, optima, india) was pre-heated to 70°c, and the trays were loaded. the oven temperature was maintained at 60°c. fruit slices were dried for 12-14 h until the slices became leathery and not sticky. the dried slices were milled using hammer mill (model: dy-19, spain) and sieved into particle size of 100 µm using a stainless steel double sieve (pro-100, china). the pineapple powder was packed and stored in a glass bottle at ambient conditions for subsequent use in the study. 2.2 formulation of pineapple powder incorporated biscuit biscuits were prepared using creamery method for making biscuit dough (manley, 2011). hydrogenated fat (margarine 50g) and powdered sugar (175g) were creamed together by electric beater (ghm-05, zhejiang, china). all-purpose flour (225 g) and baking powder (3 g) were sieved twice together. sieved flour was added to creamed paste. different amount of pineapple powder was incorporated to produce the following treatments: t1: 3% of pineapple powder (3g pineapple powder/100g mixture) t2: 5% of pineapple powder (5g pineapple powder/100g mixture) t3: 10% of pineapple powder (10g pineapple powder/100g mixture) t4: 15% of pineapple powder (15g pineapple powder/100g mixture) firm dough was prepared with each mixture. the dough were rolled out to 2.5mm thickness and cut into round pieces having 5cm diameter which were placed on oiled stainless steel trays and baked in a pre-heated oven (model: per 1700, china) at 150°c for 4 min to produce biscuits. 2.2 determination of physico-chemical properties the nutritional properties such as protein, fat, total sugar, moisture and mineral contents of pineapple powder incorporated biscuits were analyzed according to the standard method as described by aoac (2002). all analyses were carried out in triplicates. m. thivani, t. mahendran and m. kanimoly quality and shelf life of pineapple powder biscuits ruhuna journal of science vol. 7: 32-42, december 2016 35 2.3 evaluation of sensory attributes sensory evaluation was carried out using a trained panel of thirty members consisting equal men and women. samples of biscuit were presented to each of the panelist and were asked to assess the taste, colour, texture, crispness, pineapple flavour, absence of off flavour and overall acceptability using ninepoint hedonic scale with 1 representing the least score (dislike extremely) and 9 the highest score (like extremely). the sensory evaluation was carried out either at 10 am for the morning session and at 3 pm for the afternoon session to get accurate results. the analysis was performed for the freshly made biscuits and at the end of the storage period. 2.4 determination of shelf life the pineapple biscuits were packed in laminated plastics materials (metabolic, ma 18240, usa) and stored at the ambient conditions of average temperature at 30º1ºc and rh at 75-80% for 6 weeks. the biscuits were assessed for the quality characteristics during the storage period at two week intervals. visual observations were made daily to evaluate the microbial spoilage and to determine the shelf life of the biscuits. 2.5 statistical analyses the experiment was conducted using completely randomized design, consisting of four treatments replicated three times. data were subjected to analysis of variance (anova) and means were separated using duncan’s multiple range test at p<0.05 by using statistical analysis system (sas 9.1 version, north carolina, usa) software statistical package. the mean scores of the sensory evaluation were tested using analysis of variance (anova) method and differences were separated by friedman test. 3 results and discussion 3.1 nutritional composition of pineapple powder the nutritional composition of the pineapple powder was moisture 6.07%, titrable acidity 0.64% (as citric acid), ascorbic acid 57.2 mg/100g, total sugar 12.8%, fiber 0.47% and total soluble solids 13.2%. the values are in accordance with rodgers et al. (2007). according to kadam et al. (2012) nutritional composition of pineapple powder however depends on the variety, maturity, size of the fruit, harvesting interval and climatic conditions of the growing area. m. thivani, t. mahendran and m. kanimoly quality and shelf life of pineapple powder biscuits ruhuna journal of science 36 vol. 7: 32-42, december 2016 3.2 physico-chemical characteristics of pineapple powder incorporated biscuits the nutritional analysis indicated that all the biscuits contained favourable proportion of protein, fat, total sugars, moisture and mineral contents. substitution of pineapple powder for wheat flour improves the nutritive value of the product. there was a decrease in protein, fat, total sugars and mineral contents while increases in moisture level were observed in pineapple powder incorporated biscuits during storage. 3.2.1 protein content the data for changes in protein content of biscuits, during storage under ambient condition are shown in fig. 1. typically, all-purpose wheat flour contains 9.5 to 11.5% protein. according to the results, the protein content decreased significantly (p<0.05) through the storage period. this may occur due to interaction between reducing sugars and amino acids (maillard reaction) and it is a major cause of quality change and degradation of nutritional level of many foods. fig. 1: changes in protein content of pineapple powder incorporated biscuits during storage. (values are means of triplicates; vertical bars indicate the standard errors; t1:3 g pineapple powder/100 g mixture; t2:5 g pineapple powder/100 g mixture; t3:10 g pineapple powder/100 g mixture; t4:15 g pineapple powder/100 g mixture) maillard reaction impairs with the nutritional value of protein and also this reaction results in the loss of protein solubility and stability (fennema, 1996). treatment 2 (5% pineapple powder incorporated biscuits) had a lower rate of decreasing trends of protein from 22.15 to 22.06% compared to the other m. thivani, t. mahendran and m. kanimoly quality and shelf life of pineapple powder biscuits ruhuna journal of science vol. 7: 32-42, december 2016 37 treatments. various proteins present in wheat flour can undergo changes such as protein cross-linking, protein-carbohydrate interactions and denaturing of protein during processing (singh, 2000). 3.2.2 fat content the fat content decreased significantly (p<0.05) throughout the storage period (table 1). treatment 2 (5% pineapple powder added to 100g mixture) had a lowest rate of decreasing trend in fat than other treatments. this may be due to low initial moisture content in the biscuits. reduction in fat content was due to the oxidation of unsaturated fatty acids with the atmospheric oxygen and moisture uptake during storage (brooker, 1998). table 1: changes in fat content (%) of pineapple powder incorporated biscuits during storage treatments storage duration (weeks) 0 2 4 6 t1 18.770.07 b 18.720.08 c 18.640.08 c 18.570.30 c t2 18.830.03 a 18.820.03 a 18.810.03 a 18.790.03 a t3 18.820.06 a 18.800.03 b 18.780.03 b 18.740.02 ab t4 18.830.03 a 18.80.03 b 18.760.03 b 18.720.08 b values are means of triplicates ± standard error. means values with the same superscript letters in the same column are not significantly different from each other at p=5% by dmrt. fat and oil contribute to the tenderization of baked products through inhibition of gluten development and starch gelatinization (patrignani et al., 2014). this is through a water proofing effect, possibly due to complexion with carbohydrate and/or protein. according to o'brien et al., (2003), the reduction of fat content in the biscuits during storage is due to the autooxidation of unsaturated fatty acids in the lipids. 3.2.3 total sugar content the total sugars content of the biscuits decreased significantly (p<0.05) throughout the storage period (table: 2), due to the thermal degradation of sugars during baking and sugar polymerization during storage. treatment 2 (5% pineapple powder added biscuits) showed reduced rate of a decreasing trend compared to the other treatments. it had higher sugar content at the end m. thivani, t. mahendran and m. kanimoly quality and shelf life of pineapple powder biscuits ruhuna journal of science 38 vol. 7: 32-42, december 2016 of the storage period possibly due to the low moisture gain and reduced complex reaction with sugars. table 2: changes in total sugar content (%) of pineapple powder incorporated biscuits during storage. treatments storage duration (weeks) 0 2 4 6 t1 58.340.08 b 58.340.28 d 58.250.34 d 58.160.13 d t2 58.520.10 a 58.520.23 b 58.410.27 a 58.340.16 a t3 58.530.06 a 58.530.13 c 58.360.13 c 58.210.29 c t4 58.610.07 a 58.610.09 a 58.220.33 b 58.120.18 b values are means of triplicates ± standard error. means values with the same superscript letters in the same column are not significantly different from each other at p=5% by dmrt. 3.2.4 moisture content the study showed that the moisture content of the biscuits significantly (p<0.05) increased during storage (fig. 2). biscuits are very hygroscopic and have an equilibrium relative humidity around 30%. therefore, in most cases the biscuits must be protected from the atmosphere to prevent or at least reduce, moisture pick up from the atmosphere. treatment 2 (5% pineapple powder incorporated biscuits) had the lowest rate of increasing in moisture content from 4.19 to 5.0% compared to other treatments. the moisture gains of biscuits occurred due to water vapour transmission from the storage environment as the biscuits are very hygroscopic and the moisture absorption from the water vapor present in the air inside the package. another possibility of the moisture increase could be through air intake from the package seal (kumar et al., 2016). biscuits with low moisture content will have longer shelf life if they are stored under control conditions such as appropriate packaging which is impervious to moisture and gasses (bertagnolli et al., 2014). our results are in contrast with the findings of sindurani et al. (2000) who reported that the moisture content of value-added cookies decreased during storage at ambient conditions. the different directions of values may be due to the different packaging materials, methodologies and conditions used for storage of cookies. m. thivani, t. mahendran and m. kanimoly quality and shelf life of pineapple powder biscuits ruhuna journal of science vol. 7: 32-42, december 2016 39 fig. 2: changes in moisture content of pineapple powder incorporated biscuits during storage (values are means of triplicates; vertical bars indicate the standard errors; (t1:3g pineapple powder/100g mixture; t2:5g pineapple powder/100g mixture; t3:10g pineapple powder/100g mixture; t4:15g pineapple powder/100g mixture) 3.2.5 mineral content mineral content decreased significantly (p<0.05) during the storage of biscuits (table: 3). minerals in all treatments have undergone very little changes and the treatment 2 (5% pineapple powder added biscuits) had a very slower rate of decreasing trend compared to other treatments. table 3. changes in mineral content (%) of pineapple powder incorporated biscuits during storage treatments storage duration (weeks) 0 2 4 6 t1 2.1160.027 d 2.2090.04 c 2.0600.025 b 1.5000.010 c t2 2.2120.032 c 2.2010.06 b 2.1860.02 a 1.5910.005 a t3 2.2230.011 b 2.2050.03 b 2.1740.02 a 1.5690.005 b t4 2.2340.017 a 2.2300.01 a 2.1760.034 a 1.5760.012 b values are means of triplicates ± standard error. means with the same superscript letters in the same column are not significantly different from each other at p=5% by dmrt m. thivani, t. mahendran and m. kanimoly quality and shelf life of pineapple powder biscuits ruhuna journal of science 40 vol. 7: 32-42, december 2016 mineral losses can occur by heat-induced chemical reaction between reducing sugars and amino acids and/or proteins to form compounds that bind minerals (passos et al., 2013). these reaction products are more resistant to digestion and hence capable of having their mineral-binding properties to remain intact. considerable amounts of some soluble minerals are dissolved in water and leads to mineral loss during processing and storage due to hygroscopic nature of the product. 3.3 sensory analysis of pineapple powder incorporated biscuits during storage the sensory scores of the pineapple biscuits during storage is shown in fig. 3. the results revealed that there were significant (p<0.05) differences among the treatments on the sensory characteristics such as taste, colour, texture, crispiness, pineapple flavour, absence of off flavor and overall acceptability at 5% significance level. organoleptic characteristics of biscuits have slightly changed during the storage period. this may be due to non-enzymatic browning reaction (maillard reaction) and auto-oxidation of fats. fig. 3 sensory scores of pineapple powder incorporated biscuits during storage (t1:3g pineapple powder/100g mixture; t2:5g pineapple powder/100g mixture; t3:10g pineapple powder/100g mixture; t4:15g pineapple powder/100g mixture) the biscuits incorporated with 5% pineapple powder (t2) had the highest overall acceptability characteristics based on the physic-chemical and organoleptic point of view when compared to other treatments. from the overall acceptability rating, the 5% pineapple powder added biscuits had the highest mean value of 7.7 in a 9-point hedonic scale and no significant(p>0.05) changes in organoleptic characters were observed up to 6 m. thivani, t. mahendran and m. kanimoly quality and shelf life of pineapple powder biscuits ruhuna journal of science vol. 7: 32-42, december 2016 41 weeks of storage in ambient condition of average temperature at 30º1ºc and relative humidity of 75-80%. similar results were reported by ade et al. (2015) in which addition of 5% dried pineapple pulp is feasible in producing nutritionally and organoleptically acceptable biscuits. 4.0 conclusions incorporation of pineapple powder into wheat flour for the production of biscuits is possible based on the physico-chemical properties and sensory quality of the biscuits. the results revealed that the 5% pineapple powder incorporated biscuits had the highest overall quality during the storage compared to other treatment combinations. therefore, the 5% pineapple powder added biscuits could be stored for 6 weeks at the ambient condition of 301ºc and 75-80% rh without any significant changes in quality characteristics. references ade kd, lal a, mishra aa. 2015. development and quality evaluation of pineapple pomace and wheat bran fortified biscuits. the allahabad farmer journal 70(2): 92-97. aoac. 2002. official methods of analysis. (17 th edition). association of official analytical chemists. washington, usa. arshad mu, anjum fm, zahoor t. 2007. nutritional assessment of cookies supplemented with defatted wheat germ. food chemistry 102: 123–128. bertagnolli smm, silveira mlr, fogaça ado, umann l, penna ng. 2014. bioactive compounds and acceptance of cookies made with guava peel flour. journal of food science and technology 34: 303-308. brooker b. 1998. the role of fat in biscuits: strategy for fat reducing products. woodhead publishers, cambridge, uk 127-168 pp. deliza r, rosenthal a, abadio fbd, silva cho, castillo c. 2005. application of high pressure technology in the fruit juice processing: benefits perceived by consumers. journal of food engineering 67: 241-246. fennema or. 1996. food chemistry. crc press, wisconsin, usa, 157-412 pp. handa c, goomer s, siddhu a. 2012. physico-chemical properties and sensory evaluation of fructoligosaccharide enriched cookies. journal of food science and technology 49(2): 192–199. kadam dm, wilson ra, kaur v, chadha s, kaushik p, kaur, rai dr. 2012. physicochemical and microbial quality evaluation of foam‐mat‐dried pineapple powder. international journal of food science and technology 47(8): 1654-1659. kumar r, samsher, chandra s. 2016. studies on the proximate analysis of biscuits using multigrain flours during ambient condition. international journal of food science and nutrition 1(2):39-41. manley d (ed.). 2011. manley’s technology of bakery products: biscuits, crackers and cookies. elsevier, usa. 234-249 pp. m. thivani, t. mahendran and m. kanimoly quality and shelf life of pineapple powder biscuits ruhuna journal of science 42 vol. 7: 32-42, december 2016 o'brien cm, chapman d, neville dp, keogh mk, arendt ek. 2003. effect of varying the micro-encapsulation process on the functionality of hydrogenated vegetable fat in short dough biscuits. food research international 36(3): 215-221. offia-olua bi, ekwunife oa. 2015. production and evaluation of the physico-chemical and sensory qualities of mixed fruit leather produced from apple (musa pumila), banana (musa sapientum), pineapple (ananas comosus). nigerian food journal 33(1): 22-28. passos mead, moreira cff, pacheco mtb, takase i, lopes mlm, valente-mesquita vl. 2013. proximate and mineral composition of industrialized biscuits. food science and technology 33(2): 323-331. patrignani m, conforti pa, lupano ce. 2014. the role of lipid oxidation on biscuit texture during storage. international journal of food science and technology 49(8):1925-1931. pratima a, yadave mc. 2000. effect of incorporation of liquid dairy by-product chemical characteristic of soy fortified biscuits during storage. journal food science and technology 37:158–161. rodgers gl, bhangale nj, robert s. 2007. quality evaluation of pineapple powder fortified crackers during storage. indian food packer 48: 47‐55. siddiqui aa, nasreen l. 2014. effects of the baking temperature and time on the quality of biscuits. journal of state university of bangladesh 5(1): 1-12. singh rp 2000. scientific principles of shelf life evaluation protein biscuits. journal of food science and technology 31: 117-121. sindurani ja, smith jl, rajmohan t. 2000. effects of different levels of fruit powders on the storage stability of value-added cookies. indian journal of food science and technology 44(1): 97-100. tortoe c, johnson pt, slaghek t, oduro-yeboah c, addo p, nyarko a, tamakloe i. 2014. physico-chemical, proximate and sensory properties of organic side-stream pineapple (ananas sp.) flour. international journal of food science, nutrition and engineering 4(1): 1-8. wani sh, gull a, allaie f, safapuri ta. 2015. effects of incorporation of whey protein concentrate on physic-chemical, texture, and microbial evaluation of developed cookies cogent food and agriculture. 1(1): 109-116. yadav rb, dhull n, yadav bs. 2012. effect of incorporation of plantain and chickpea flours on the quality characteristics of biscuits. journal food science and technology 49(2): 207–213. sv-lncs ruhuna journal of science vol 8: 1223, june 2017 eissn: 2536-8400  faculty of science doi: http://doi.org/10.4038/rjs.v8i1.23 university of ruhuna  faculty of science, university of ruhuna sri lanka 12 physical quality characters of cookies produced from composite blends of wheat and sweet potato flour m.b.f. jemziya1 and t. mahendran2 1 department of biosystems technology, faculty of technology, south eastern university of sri lanka, university park, oluvil, sri lanka 2 department of agricultural chemistry, faculty of agriculture, eastern university, sri lanka correpondence:1fathimajemziya@yahoo.com, 2 thevamahen@yahoo.com received: 27th march 2017, revised: 17th april 2017, accepted: 23rd may 2017 abstract. a research study was carried out to develop cookies with good nutritional and physical quality from sweet potato flour and wheat flour and to evaluate the quality characteristics. the mature sweet potato (cv. wariapola red) tubers were washed, peeled, cut into thin slices of 1mm thickness, dried under the sun until the pieces were quite brittle, milled and sieved. different composite blends of wheat flour and sweet potato flour were mixed in the ratios (w/w) of 100:00, 80:20, 60:40, 40:60, 20:80 and 00:100. cookies were then prepared from separate mixtures following manufacturing procedures. sweet potato flour was nutritionally analyzed and it contained 2.3% protein, 9.4% dietary fiber and 85.5% soluble carbohydrate, and therefore, sweet potato flour appeared to be suitable for a successful combination with wheat flour for the production of cookies. physical characteristics such as thickness, volume, diameter and spreading factor of the cookies decreased from 0.969 to 0.910 cm, 41.66 to 30.41 cm3, 7.4 cm to 6.52 cm and 6.43 to 5.61 respectively while density of cookies increased from 0.474 g/cm3 to 0.652 g/cm3 with increasing of sweet potato flour up to 100%. the sensory qualities showed that the cookies supplemented with 40% sweet potato flour were well acceptable in terms of colour, texture, taste and overall acceptability. the mixture of 40% sweet potato flour and 60% wheat flour had been successful for the formulation of composite cookies with better physical, nutritional and organoleptic qualities within the universally accepted standards. keywords. cookies, physical characters, organoleptic qualities, sweet potato flour, wheat flour. 1 introduction sweet potato (ipomoea batatas) is an extremely versatile and delicious vegetable that possesses high nutritional value (mohanraj and sivasankar m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science vol 8: 12-23, june 2017 13 2014). among the world's major food crops, sweet potato produces the highest amount of edible energy per hectare per day (sukhcharn et al. 2008). sweet potato consists of about 70% carbohydrates (dry basis) of which a major portion is starch, which can be utilized as a functional ingredient in certain food preparations (avula 2005). it is an excellent source of vitamin a (in the form of beta carotene) and also a very good source of vitamin c and manganese. in addition, sweet potatoes are a good source of dietary fiber, natural sugars, protein, niacin, vitamin b5, vitamin b6, vitamin e, potassium, biotin, iron, calcium and copper (“the world healthiest foods”, 2017). sweet potato is commonly referred to as a subsistence, food security or famine relief crop. its uses have diversified considerably in the developing countries. sri lanka has a long history of cultivation of sweet potatoes. it is considered as a crop of exotic origin, but people regard it as indigenous because it has been in cultivation in sri lanka as an important traditional food crop since ancient times (karunathilake 2005). options for sweet potato products are numerous, and based on recent diagnostic assessments carried out in developing countries, dried chips, starch and flour have been identified as among the most promising products (collins 1989). the raw material of foremost importance in bakery product is the wheat flour. wheat is the main raw material for bakery, biscuit and paste (diana et al. 2007). bakery products are commonly made from wheat flour containing gluten, whereas gluten contributes to the typical texture, flavour and form of the usual bread and cake products. without gluten, baked goods will not hold their shape. that is why wheat flour is used in baking. there is however circumstances in which wheat flour are not readily available, or if people in these areas nevertheless wish to change to the consumption of the bread so that they will need to import wheat and pay for it. a mixture of wheat flour and sweet potato flour could make a good baking product, which should increase its economic value (zuraida 2003). most of the technical research on sweet potato flour focused on the development of new products using sweet potato flour rather than on efficient methods to produce and store the flour (lizado and guzman, 1982; sukhcharn et al. 2008). addition of various proportion of sweet potato flour in wheat flour can increase the nutritive values in terms of fiber and carotenoids and also helps in lowering the gluten level and prevent humans from coeliac disease (tilman et al. 2003). sri lanka has a long history of sweet potato cultivation. extent and production of sweet potato are 3,270 hectares and 25,780 metric tons in 2010 (department of census and statistics of sri lanka, 2010). sweet potato can be grown successfully throughout the year in all agro-ecological zones. in sri lanka, they are cultivated in all the districts but mainly in ratnapura, hambantota, kurunagale, gampaha, kagalle, badulla and kalutura districts. in dry zone the cultivation area of sweet potato is low compared to wet zone m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science 14 vol 8: 12-23, june 2017 and intermediate zone (department of census and statistics of sri lanka, 2010). it is considered as a crop of exotic origin, but people regard it as indigenous because it has been in cultivation in sri lanka as an important traditional food crop from very ancient time (karunathilake 2005). sweet potato based products are of high quality and could compete with the existing products in the market (sneha et al. 2012). the use of sweet potato flour for supplementing with wheat flour in the bakery industry could substantially reduce the need for wheat being imported, reduce the usage of sugar on the products and increase the value of sweet potato. the objective of present study is to replace the wheat flour in cookies with sweet potato flour (gluten – free flours) in order to increase the fiber and other nutrients and develop cookies resembles as closely as possible to the wheat flour based product. 2 materials and methods 2.1 procurement of materials fresh sweet potatoes (wariapola red) without any bruises were procured locally from the field of commercial grower. roots were washed, trimmed and cured to make them free from soil and other foreign materials, rotting or insect damage. trimming was carried out manually and curing was done at 35°c for 2–3 days, stored at 12-15°c at 80% relative humidity till further use. other major ingredients i.e. wheat flour, sugar, baking powder, salt and margarine were purchased from the super market of batticaloa. 2.2 preparation of sweet potato flour purchased sweet potatoes (cv. wariapola red) were washed, peeled and cut into thin slices at around 1mm thickness. drying of sweet potato slices was done on perforated trays in the sun until the pieces were quite brittle and then stored in air tight container till further use. the dried chips were milled into flour using electric mill grinder (aikeleyisi, flat wheel grinder) and passed through sieves (250 µm) to obtain flour of uniform size. the flour was then packed in air tight container and stored under ambient conditions until further use. 2.3 experimental plan following types of cookies were made, and designated as t1 to t6. t1 100% wheat flour m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science vol 8: 12-23, june 2017 15 t2 20g sweet potato flour /80g wheat flour (20+80 g) t3 40g sweet potato flour /60g wheat flour (40+60 g) t4 60g sweet potato flour /40g wheat flour (60+40 g) t5 80g sweet potato flour /20g wheat flour (80+20 g) t6 100% sweet potato flour 2.4 development of wheat and sweet potato blend cookies cookie dough was prepared according to the following formula: 100 g of flour (contain different proportion of sweet potato flour and wheat flour), 50 g of sugar, 20 g of margarine, 2 g of baking powder, 0.5 g of sodium chloride and various proportion of water to make required consistency of dough. the firm dough was rolled out to 5 mm thickness in a baking tray and cut into round having 7.4 cm diameter with a cookie cutter. the cookies were placed on a greased aluminum tray and baked in a pre-heated oven at 200 0c for 10 minutes to produce cookies. these cookies were assessed for physico chemical and organoleptic qualities. 2.5 determination of physical characteristics of wheatsweet potato flour composite cookies physical parameters of wheat – sweet potato cookies were measured such as diameter (mm by vernier caliper, mitutova, japan), and thickness (mm) and other parameters such as volume, density and spread ratio were determined as follows. volume volume of cookies is defined as the area of the cookies multiplied by its thickness. volume (cm3) = 3.14 hd2 4 where h = thickness of cookie (cm), and d = diameter of cookie (cm) density after calculating the volume of cookies, density of them was obtained by ratio of mass to volume. density (gcm-3) = mass of cookie (g) volume of cookie (cm3) spread ratio the spread ratio was determined by using this formula. m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science 16 vol 8: 12-23, june 2017 spread ratio = diameter of cookie thickness of cookie 2.6 sensory analysis of wheat – sweet potato flour composite cookies the sensory attributes, including texture, colour, taste, mouth-feel and overall acceptability were evaluated by a trained twenty-member panel. ranking test was used to evaluate the perceptible differences in intensity of an attribute among samples. samples were presented in identical disposable plastic dishes, coded with 3 digit random numbers. each sample was given a different code number. all the samples were simultaneously presented to each panelist in a balanced or random order. trained panelists are asked to rank the coded samples for the intensity of a specific characteristic, by ordering the samples from the most intense to the least intense. the panelists were allowed to re evaluate the samples necessary to make the required comparisons among them. 2.7 microbial examination potato dextro agar (pda) preparation was carried out without any external contamination. total plate count was taken as described below. peeled potato was cut into small pieces and added in 250 ml of distilled water and boiled. weighed agar was boiled with 250 ml of distilled water until agar dissolve and placed in a 1000 ml of flask. then required amount of sucrose and potato extraction were added into the flask and stirred. then conical flask containing media was plugged with a cotton wool and wrapped aluminum foil. then it was put into the autoclave at 121oc, 15psi for 20 minutes and the media was allowed to cool. petri dishes, forceps and needles were kept in the oven at 180oc for one hour and allowed to cool. all the equipment were sterilized by 70% of alcohol. then it was poured into petri dishes and they were kept in lamina flow until solidify. different treatment samples were placed in agar plate. then petri dishes were covered and labeled. the plates were observed after 4 days for plate count. 2.8 shelf life evaluation the shelf life of cookies was assessed based on the nutritional and sensory qualities. the cookies were organoleptically examined once in two weeks. m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science vol 8: 12-23, june 2017 17 3 results and discussion 3.1 nutritional composition of the freshly made sweet potato flour nutritional composition of the freshly prepared sweet potato flour is presented in table 1. the results of nutritional composition of sweet potato flour are closely related with the results obtained by sukhcharn et al. (2008). the moisture content of sweet potato flour was 8.1%. sweet potato flour had fiber content 9.4%. this high fiber increases the utility of sweet potato flour in various food products and also had lesser extent of protein content (2.3%) compared to the wheat flour (12.6%). most non wheat flours have less protein but higher carbohydrate content than wheat flour (tindall 1968, okorie and onyeneke 2012). table 1. nutritional composition of the freshly made sweet potato flour (the values are means of four replicates ± standard error). constituents content (%) moisture 8.10 ± 0.15 ash 3.60 ± 0.15 fat 0.52 ± 0.01 fiber 9.40 ± 0.16 protein 2.30 ± 0.20 total sugar 11.20 ± 0.17 reducing sugar 6.30 ± 0.22 total soluble carbohydrate 85.48 ± 0.41 the high level of carbohydrates is desirable in baked products, because on heating, starch granules in the presence of water swells and forms a gel which is important for the characteristic textures and structures of baked goods (amendole 1972; okorie and onyeneke 2012). sweet potato flour was produced from white flesh sweet potatoes and was dull white in colour, had a somewhat sweet flavour. greater nutritive value is achieved due to the fact that as greater part of water content is removed, while the carbohydrates, pectin, proteins, oils and mineral salts are concentrated in the tissue of dried food products. 3.2 physical parameters of developed wheat – sweet potato cookies the physical parametric analysis of the cookies revealed that, there was a significant differences between the treatments as the level of sweet potato flour was increased (0-100%) in respect of diameter, thickness, volume, m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science 18 vol 8: 12-23, june 2017 density and spread ratio of cookies at the 5% level of significance according to anova. fig. 1. diameter and spread ratio of different combination of wheat – sweet potato flour cookies (the values are means of four replicates; the vertical bars indicate the standard errors). diameter there was a significant decrease in the diameter of control treatment (t1) and other different treatments (t2, t3, t4, t5, t6) after incorporating cookies with sweet potato flour. according to dmrt, control treatment which containing 100% wheat flour has the highest mean value (7.4 cm) followed by 20% sweet potato flour added cookie and 100% sweet potato cookie has the least mean value. even though, there are no significant difference between control treatment and sweet potato flour added cookies up to 40%. diameter and spread ratio of different combination of wheat – sweet potato flour cookies are shown in figure 1. thickness according to dmrt, control treatment which containing 100% wheat flour has the highest mean value (0.969 cm) followed by 20% sweet potato flour added cookie and 100% sweet potato cookie has the least mean value. the result showed that increase in level of sweet potato flour resulted in linear decrease of thickness and diameter of cookie. this might be due to the higher water holding capacity of sweet potato flour. mean values of thickness, m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science vol 8: 12-23, june 2017 19 volume and density of different combination of wheat – sweet potato flour cookies according to dmrt are shown in table 2. volume there was a significant decrease in volume of control treatment (t1) and other different treatments (t2, t3, t4, t5, t6) after incorporating cookies with sweet potato flour. according to dmrt, control treatment containing 100% wheat flour has the highest mean value and 100% sweet potato cookie has the least mean value. table 2. thickness, volume and density of different combination of wheat – sweet potato flour cookies (means of four replicates ± standard error; the means with shared letters are not significantly different from each other at 5% level based on dmrt) treatments thickness (cm) volume (cm3) density (gcm-3) t1 0.969 ± 0.003 a 41.66 ± 0.84a 0.474 ± 0.009c t2 0.949 ± 0.006 b 40.52 ± 0.91ab 0.487 ± 0.011c t3 0.931 ± 0.003 c 38.87 ± 0.82b 0.508 ± 0.011c t4 0.930 ± 0.002 c 34.33 ± 0.41c 0.575 ± 0.007b t5 0.920 ± 0.005 cd 32.99 ± 0.42c 0.598 ± 0.007b t6 0.910 ± 0.002 d 30.41 ± 1.21d 0.652 ± 0.027a density the densities of the cookies gradually increase (0.474 0.652 gcm-3) due to marginal increase of sweet potato flour. according to dmrt, there was no significant difference between control treatment and sweet potato flour added cookies up to 40% (t2 and t3). the volume of sweet potato cookies decreased linearly whereas, density increased in the similar manner. this may be due to higher fiber content in the sweet potato flour. spread ratio there was a significant decrease in the spreading factor of control treatment (t1) and different treatments (t2, t3, t4, t5, t6) after incorporating cookies with sweet potato flour. according to dmrt, control treatment which containing 100% wheat flour has the highest mean value followed by 20% sweet potato flour added cookie and 100% sweet potato cookie has the least mean value. even though, there was no significant difference between control treatment and sweet potato flour added cookies up to 40%. the differences in m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science 20 vol 8: 12-23, june 2017 spread factors of cookies containing different proportion of sweet potato flour may be attributed to the differences in swelling patterns and rheological properties. the results may be due to the higher fiber content, solid matter content and also high water holding capacity of the sweet potato flour. 3.3 organoleptic evaluation of freshly made wheat – sweet potato flour composite cookies the sensory evaluation of the cookies revealed that there were significant differences between the treatments as the level of sweet potato flour increased from 0 to100% for texture, mouth-feel, taste, colour and overall acceptability at the 5% level of significance according to anova. mean values of treatments according to tukey's studentized range (hsd) test are shown in table 3. the results of the sensory attributes of wheat and sweet potato flour blend cookies such as texture, colour, taste and overall acceptability have deviated pattern of scores and also score of mouth feel attribute have similar way of pattern compared with the results obtained by sukhcharn et al. (2008). these different directions of score patterns may be due to the different rates of preference and acceptable values of panels and quality of finished cookies that were studied. texture texture is one of the most important parameters connected to product quality. it is defined as the sensory manifestation of the structure of food and the manner in which that structure reacts to the applied force (jean-xavier and rossella, 1996). texture analysis involves measuring the properties related to how a food feels in our mouth (initial bite). according to dmrt, there was no significant difference between the control treatment and series of 20 60% sweet potato flour added cookies (t2, t3, and t4). the control treatment and t3 have the highest mean value and t6 has the least mean value. mouth-feel mouth-feel analysis involves measuring the properties of cookies such as crunchy, granular, flaky and teeth clogging. mouth-feel attributes of cookies decreased from 4.50 to 3.30 with increasing in the substitution of sweet potato flour. according to dmrt, there were no significant differences between control treatment (t1) and sweet potato flour added cookies up to 60%. the control treatment (t1) has the highest mean value and t6 has the least mean value. m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science vol 8: 12-23, june 2017 21 table 3. mean values of organoleptic attributes of cookies incorporated with sweet potato flour at different levels (the values are means of 20 replicates ± standard error) treatments texture mouth-feel taste colour overall acceptability t1 4.85 ± 0.05 a 4.50 ± 0.11a 4.25 ± 0.27abc 4.60 ± 0.11ab 4.75 ± 0.08a t2 4.55 ± 0.11 a 4.45 ± 0.11a 4.30 ± 0.15abc 4.65 ± 0.10a 4.45 ± 0.11a t3 4.85 ± 0.05 a 4.30 ± 0.12a 4.50 ± 0.11ab 4.80 ± 0.07a 4.70 ± 0.09a t4 4.45 ± 0.11 a 4.20 ± 0.13ab 4.80 ± 0.07a 4.75 ± 0.08a 4.60 ± 0.11a t5 3.95 ± 0.15 b 3.60 ± 0.16bc 3.95 ± 0.14bc 4.15 ± 0.12bc 3.75 ± 0.17b t6 3.55 ± 0.15 b 3.30 ± 0.22c 3.60 ± 0.15c 3.70 ± 0.20c 3.40 ± 0.12b *the means with the shared letters are not significantly different from each other at 5% level based on tukey's studentized range (hsd) test. *the sensory attributes were analyzed in ranking test with the values from 1 to 4.9. taste cookie containing 60% sweet potato flour had sweet taste and 100% sweet potato flour contained cookie had either caramel or burnt taste. this might be due to the caramelization of free sugar in sweet potato during baking (sukhcharn, et al. 2008). according to dmrt, t4 as highest mean value and 100% sweet potato flour added cookie has the least mean value. colour the quality of food is generally based on colour, flavour, texture and nutritive value. an attractive colour leads to the food to make good demand. the colour of cookie changed from light brown to dark brown. the darker colour may be due to maillard reaction between reducing sugar and protein (raidi and klevin 1983; dhingra and jood 2000). according to dmrt, t3 has highest mean value and 100% sweet potato flour added cookie t6 has the least mean value. overall acceptability overall acceptability includes many implications, which is the important parameter in sensory estimation. there were no significant differences between control treatment and 20 – 60% sweet potato flour added cookies (t2, t3 and t4). the 40% sweet potato flour added cookie (t3) has highest mean value and 100% sweet potato flour added cookie (t6) has the lowest mean value. m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science 22 vol 8: 12-23, june 2017 3.4 sensory analysis of wheat – sweet potato cookies following storage organoleptic characteristics of the cookies were changed slightly following the storage period. this may be due to the non – enzymatic browning reaction (maillard reaction) and fat oxidation. table 4. mean values of sensory attributes of stored wheat – sweet potato cookies (the values are means of 20 replicates ± standard error). treatments texture mouth-feel taste colour overall acceptability t2 4.52± 0.11 b 4.40±0.12a 4.30±0.15b 4.65±0.10a 4.35±0.11a t3 4.82±0.05 a 4.25±0.15a 4.50±0.11ab 4.80±0.07a 4.60±0.11a t4 4.42±0.11 b 4.16±0.15a 4.80±0.07a 4.75±0.08a 4.50±0.11a the means with the shared letters are not significantly different from each other at 5% level based on tukey's studentized range (hsd) test. the sensory attributes were analyzed in ranking test with the values from 1 to 4.9. the 40% sweet potato flour added cookie (t3) has the best shelf life in nutritional and organoleptical point of view compared to other combinations of wheat and sweet potato flour. mean values of sensory attributes of stored wheat – sweet potato cookies are shown in table 04. from the overall acceptance rating, the 40% sweet potato flour added cookie has the highest mean value and no remarkable changes in organoleptic characters were observed up to three months of storage in ambient condition of average temperature 300c and relative humidity of 75 – 80%, indicate that the 40% sweet potato flour added cookies could be preserved up to three months. 4 conclusions the finding of the present research revealed that sweet potato contains considerable amount of protein, although rich in dietary fiber content and carbohydrate, so that a successful combination with wheat flour for cookie production would be nutritionally advantageous and also sweet potato flour with wheat flour has significant effect on the physical properties of the flour blends. sweet potato flour (40%) incorporated cookies have highly acceptable functional and organoleptic quality characters compared to other combinations. the outcome of the present research can be used as valuable information for the development of high fiber low gluten sweet crunchy cookies. the results obtained could be very valuable in decision making for industries that want to take nutritional advantage of sweet potato flour as alternative or supplement to cereal flours. m.b.f. jemziya and t. mahendran cookies from sweet potato and wheat flour ruhuna journal of science vol 8: 12-23, june 2017 23 references amendole j. 1972. the baker’s manual, 3rd edn. anhriers series. new jersey: hyden book co. inc. l 3 pp. avula ry. 2005. rheological and functional properties of potato and sweet potato flour and evaluation of its application in some selected food products. phd thesis. department of fruit and vegetable technology central food technological research institute mysore – 570020 india. collins m. 1989. economic analysis of wholesale demand for sweet potatoes in lima, peru. m.sc. thesis. department of agricultural and resource economics, university of florida, gainesville, usa. 1-164 pp. department of census and statistics, sri lanka. 2010. http://www. statistics.gov.lk/agriculture/seasonalcrops/seasonalcropsnationaltotals.html (accessed on 24th of may, 2013). dhingra s, jood s. 2000. organoleptic and nutrition evaluation of wheat breads. journal of food chemistry 77:479 – 488. diana nr, mirela vg, jianu i. 2007. studies regarding the chemical composition of several wheat species, flour types and pastes assortments. bulletin usamv – cn 64/2007. jean-xavier g, rossella m. 1996. the sensory perception of texture and mouth feel. trends in food science and technology 7(7): 213 – 219. karunathilake ek. 2005. vegetable science, horticulture (vegetables, tubers, and spice crops). 120 – 122 pp. lizado mlc, guzman mp. 1982. development of new products using sweet potato flour. economic journal 10(1): 62. mohanraj r, sivasankar s. 2014. sweet potato (ipomoea batatas [l.] lam)--a valuable medicinal food: a review. journal of medicinal food 17 (7): 1–9. okorie su, onyeneke en. 2012. production and quality evaluation of baked cake from blend of sweet potatoes and wheat flour. part-i: natural and applied sciences 3(2):171-177. raidi ma, klevin bp. 1983. effect of soy or field pea flour substitution on physical and sensory characteristics of chemistry leavened quick bread. journal of cereal chemistry 60: 367 – 370. sneha s, genitha tr, vrijesh, y. 2012. preparation and quality evaluation of flour and biscuit from sweet potato. journal of food processing and technology, omics international 3(12):192. sukhcharn s, riar cs, saxena dc. 2008. effect of incorporating sweet potato flour to wheat flour on the quality characteristics of cookies. african journal of food science 2: 65-72. the world’s healthiest foods. “sweet potatoes”, accessed on april 17-23, 2017. http://whfoods.org/genpage.php?tname=foodspice&dbid=64. tilman jc, colm mob, denise mc, anja d, elke ka. 2003. influence of gluten free flour mixes and fat powder on the quality of gluten free biscuits. journal of european food research and technology 216: 369-376. tindall hd. 1968. commercial vegetable growing. oxford: oxford university press. 234 pp. zuraida n. 2003. sweet potato as an alternative food supplement during rice shortage. journal lit bang pertanian 22 (4):150 -155. ruhuna journal of science vol 10(1): 77-87, june 2019 eissn: 2536-8400 faculty of science doi: http://doi.org/10.4038/rjs.v10i1.xx university of ruhuna  faculty of science, university of ruhuna sri lanka 77 short paper types and frequency of fingerprint minutiae in individuals of igbo and yoruba ethnic groups of nigeria u. u. akpan1*, t. awe1, d. o. idowu1, 2 and k. o. adekoya1 1department of cell biology and genetics, university of lagos, lagos nigeria. 2 professor abideen oluwasola special diagnostic centre, department of pathology, university college hospital, ibadan, nigeria. *correspondence: akpanutomobong@gmail.com; orcid: 0000-0002-4062-1034 received: 27th may 2018, revised: 8th march 2019, accepted: 29th april 2019 abstract the population distribution of fingerprint minutiae is necessary to improve efficiency of fingerprints in identifying individuals in a population-specific manner. the objective of the study was to determine the distribution of different types of minutiae fingerprint feature in two nigerian ethnic populations. fingerprints from forty-four (44) igbo individuals and forty-four (44) yoruba individuals, both of nigeria were collected using a manual impression method that uses ink pad and paper. of all the minutiae types considered, bifurcations and convergences accounted for 54.85% of the study’s total minutiae counts (tmc). this study shows that the igbo ethnic group consistently have higher count of all minutiae types and higher total minutiae counts for both hands statistically significant at p<0.001. we found association between gender and minutiae distribution in some minutiae types including fragment/point or dot (fp), overlap (ol) and break (br). this work revealed important variations among individuals from the two ethnic groups on the distribution and variability of minutiae in nigeria populations. keywords: dermatoglyphics, fingerprints, forensic, igbo, minutiae, nigeria, yoruba. 1 introduction the study of dermatoglyphics in humans involves analyzing the epidermal ridges on the surface of the palms, soles, fingers, and toes of humans (cummins and midlo 1943). dermatoglyphics have been widely employed in areas as anthropology, genetics and evolutionary studies in characterizing populations, analyzing the nature and origin of human variability, population doi: http://doi.org/10.4038/rjs.v10i1.52 https://orcid.org/0000-0002-4062-1034 https://creativecommons.org/licenses/by-nc/4.0// u.u. akpan et al fingerprint minutiae of igbo and yoruba ethnic groups ruhuna journal of science vol 10(1): 77-87, june 2019 78 structure assessment, and the micro-differentiation among populations (meier 1980, durham and plato 1990, segura-wang and barrantes 2009). previous studies of dermatoglyphics in human populations have hinted on the usefulness of pieces of information obtained from these characteristics in the understanding of the evolution and genetic structure in human populations (blangero 1990, crawford and duggirala 1992), in characterizing syndromes and diseases (schaumann and alter 1976), and in personal identification in the field of forensic sciences (dankmeijer et al. 1980, champod et al. 2004, faigman et al. 2008). worldwide, only features such as main pattern type, pattern intensity index or ridge count have been widely studied (aase and lyons 1971, karthick et al. 2015), whereas other features, such as the minutiae or epidermal ridge breadth have received comparatively less attention, despite being of considerable interest due to their direct relevance in personal identification (loesch 1983, dankmeijer et al. 1980, champod et al. 2004, gutierrezredomero et al. 2007, gutierrez-redomero et al. 2011). the minutiae representation of the fingerprints has now been adopted by most of the commercially available automatic fingerprint matching systems because it has been relatively demonstrated to be stable (pankanti et al. 2002) and can be extracted from low quality images whereas it is difficult to extract ridge features from low quality images (iloanusi 2011). nigeria is a country of more than 150 million people, and her ethnic landscape is dominated by three major tribes: hausa, igbo and yorubas according to national population commission (npc) report of 2006. although there has been considerable amount of dermatoglyphic studies carried out on nigerian populations, only a few characteristic features have been focused on, while some others have received less or no attention. the earliest works on fingerprint characteristics for individuals of nigerian origin were performed by lestrange (1953) and ojikutu (1964). jantz and brehme (1978) performed the first extensive in yoruba finger and palmar dermatoglyphics, though lestrange (1953) had reported a small number of females and ojikutu (1964) had included yoruba as part of his nigeria sample. these studies established the frequencies of the arch, loop and whorl pattern types in different nigerian ethnic groups. several other studies in different ethnic groups reporting frequencies of pattern types include jaja et al. (2011) who studied the dematoglyphics of the ogoni people focusing on pattern frequencies, ridge count and a-b ridge count and otobo and jarimbootobo (2016) who assessed the digital and palmar characteristics in the ijaw ethnic group with a surprising high radial loop frequency. eboh studied the digital dermatoglyphics of anioma and urhobo people of nigeria and found no association between ethnicity and pattern while mohammed (2014) studied digital dermatoglyphics in the kanuri. several other studies assessed the association between different pattern types and health condition/phenotypes and often generated pattern frequencies u.u. akpan et al fingerprint minutiae of igbo and yoruba ethnic groups ruhuna journal of science vol 10(1): 77-87, june 2019 79 as part of it report. dike et al. (2012) performed a comparative dermatoglyphics study of digital pattern, tri-radii and palmar distances in diabetic and hypertensive individuals in rivers state. an association was observe between ulnar loop and hypertensive subject as well as between whorls and diabetic subject. ethnicity was not considered. oladipo et al. (2009) studied the association of digital and palmar dermatoglyphics pattern in nigeria women with malignant mammary neoplasm and obtain significantly high mean dat and a reduced total right ridge count as being indicative of the condition. oladipo et al. (2010) examined the digital and tfrc as well as dat and atd in obese ibibio individuals and concluded that arch on the first digit was significantly associated with obesity with a greater atd and dat in unobese individual. an association study of dermatoglyphics in cancer patients indicated a 74.9% loop pattern on the right hand and 68.5% on the left hand, both significantly higher than in non-cancer individuals (umana et al. 2013). adekoya et al. (2013) assessed the relationship between dermatoglyphics and multiple intelligence among selected secondary schools students in nigeria and found significant associations. the yoruba in jos were examined by akingbade et al. (2014) while nigerians of undefined ethnicity were evaluated for association of dermatoglyphics and abo blood group with no significant associations (eboh 2013).the association of dermatoglyphics pattern in congenital deaf and mute as well as handedness has been reported (osunwoke et al. 2010, ogunaike et al. 2014). the type and distribution of minutiae in the fingerprints of nigerian have received little or no attention. however, orike et al (2016) have reported an effort to explore the forensic applications of fingerprint minutiae for inferring the gender and ethnicity of an individual from fingerprints of an unknown nigerian origin. several studies have recently been undertaken on fingerprints from the hausas of nigeria. recent studies by adamu et al. (2016a, 2016b, 2017a, 2017b), adamu (2017), and adamu and taura (2017) have investigated the minutiae in the hausa population in kano and estimated its association with sexual dimorphism in epidermal ridges, sex variation in thumb minutiae, estimation of body mass index, and sex estimation.. there is therefore a need to assess the type and frequencies of different fingerprint minutiae in nigerians of other ethnic origins, including igbo and yoruba. 2 material and methods the population site for this study was university of lagos. samples were collected from staff and students of the university of lagos, akoka, lagos state, nigeria, who volunteered to participate in the study. the selected participants were verified to be of igbo and yoruba ancestry which extends to at least 2-3 previous generations on both parent’s genealogies. consent of the u.u. akpan et al fingerprint minutiae of igbo and yoruba ethnic groups ruhuna journal of science vol 10(1): 77-87, june 2019 80 participant was obtained adequately educating them on what the study aims to achieve and the risk involved in the data collections. every participant was made to sign an informed consent after reading it to them. the participant’s demography including sex and tribal classification were obtained through a designed questionnaire are shown in table 1. participants were between the ages of 18-60; 44 yorubas and 44 igbos. table 1: demography of the participants ethnicity gender total male female frequency frequency yoruba 30 14 44 igbo 36 8 44 total 66 22 88 fig 1. the types of minutiae considered in this study (gutiérrez-redomero et al. 2011) plain fingerprint of participants was collected manually using a blue-inked pad and a sheet of paper. individual participant’s fingers were placed on the ink pad and then pressed on the paper which contained spaces for each of the ten fingers and also had provision for obtaining the participants personal and demographic data. the definitions of the minutiae used are according to gutierrez-redomero et al. (2011) (figure 1). the fingers/ digits were u.u. akpan et al fingerprint minutiae of igbo and yoruba ethnic groups ruhuna journal of science vol 10(1): 77-87, june 2019 81 designated the following letters: thumb (t), index (i), middle (m), ring(r) and little (l). the hands were designated left (l) and right (r). the fingerprints were carefully examined with the aid of a hand lens. the occurrence of each type of minutiae were counted and recorded against each fingerprint. the difference in proportions of minutiae between the ethnic groups was tested for statistically significance using fisher’s exact test and univariate analysis on spss 21. the significance level was set at 0.053. 3 results the comparison made between the frequencies of the different classes of minutiae is also presented in table 2. among the study population the most frequent type is the bifurcations and convergence, which accounted for 54.85% of the study’s total minutiae counts (tmc). fragment/point or dot and the bridge type, were the next most frequent representing 12.40% and 10.16% of the study’s tmc respectively. association of the distribution of the minutiae types (mts) on ethnic group was found for return and bifurcations and convergence at p< 0.01, while those found in overlap and crossbar were at a higher significant level of p<0.001 (table 2). table 2: distribution of different minutiae types in the two populations (total n= 21450). * significantly higher at p<0.01, ** significantly higher at p<0.001 minutiae type count yoruba (n=10,206) igbo (n=11,244) n (%) total (%) total (%) bifurcations and convergence (bc) 11,764 (54.85) 5,916 (57.97)* 5,848 (52.01) fragment/point or dot (fp) 2,660 (12.40) 1,256 (12.3) 1,404 (12.49) bridge (bg) 2,178 (10.15) 1,066 (10.45) 1,112 (9.88) ridge ending (er) 1,328 (6.19) 556 (5.45) 772 (6.87) overlap (ol) 854 (3.98) 338 (3.3) 516 (4.59) ** break (br) 808 (3.77) 292 (2.86) 516 (4.59) enclosures (en) 758 (3.53) 338 (3.31) 420 (3.74) crossbar (cp) 628 (2.93) 288 (2.82) 340 (3.02) ** return (rt) 406 (1.89) 90 (0.88) 316 (2.81) * others (ot) 66 (0.31) 66 (0.65) 0 (0) u.u. akpan et al fingerprint minutiae of igbo and yoruba ethnic groups ruhuna journal of science vol 10(1): 77-87, june 2019 82 table 3: the total count of minutiae (mean ± sd) over whole fingerprint for both hands. hands total (n=88) ethnic group (n=88) yoruba (n=44) igbo (n=44) both hands tmc 243.75 ± 61.15 232.51 ± 76.16 256.20±38.07* right hand tmc 124.02 ± 38.65 117.38 ± 40.69 131.30± 37.15 left hand tmc 119.73 ± 38.03 115.20 ± 47.02 124.46± 25.27 * significantly higher at p<0.001 in the study population, the counts (tmc) for the right and left hands were unequal between the tribes for both hands (table 3). however, there was no significant association between tribal classifications and the varying tmcs for the right and left hands. on the other hand, this significant association was discovered for the total tmc. table 4: presence or absence of different types of minutiae on all or any of the ten fingers and their descriptive statistics (mean ± sd). minutiae type total (n=88) yoruba igbo present (%) mean ± sd present (%) mean ± sd present (%) mean ± sd bc 88 (100) 133.68 ± 32.37 44 (100) 134.45 ± 36.48 44 (100) 132.91 ± 27.19 er 82 (93.18) 15.09 ± 14.99 38 (86.36) 12.64 ± 14.55 44 (100) 17.55 ± 15.17 fp 88 (100) 30.23 ± 14.12 44 (100) 28.55 ± 13.91 44 (100) 31.91 ± 14.28 ol 80 (90.91) 9.70 ± 7.56 36 (81.82) 7.68 ± 8.28 44 (100) 11.73* ± 6.22 cp 78 (88.64) 7.14 ± 6.68 38 (86.36) 6.55 ± 7.24 40 (90.91) 7.73 ± 6.10 bg 88 (100) 24.75 ± 12.33 44 (100) 24.23 ± 13.57 44 (100) 25.27 ± 11.08 br 80 (90.91) 9.18 ± 8.46 40 (9 0.91) 6.64 ± 7.604 40 (90.91) 11.73 ± 8.59 en 86 (97.73) 8.61 ± 8.11 42 (95.46) 7.68 ± 9.72 44 (100) 9.55 ± 6.08 rt 68 (77.73) 4.61 ± 6.08 28 (63.64) 2.05 ± 2.56 40 (90.91) 7.18* ± 7.39 ot 12 (13.64) 0.75 ± 3.295 12 (27.27) 1.50¶ ± 4.56 0 (0) 0 ¶ significantly higher at p<0.001; * significantly higher at p<0.01 bifurcations and convergence (bc); fragment/point or dot (fp); bridge (bg); ridge ending (er); overlap (ol); break (br); enclosures (en); crossbar (cp); return (rt); others (ot) u.u. akpan et al fingerprint minutiae of igbo and yoruba ethnic groups ruhuna journal of science vol 10(1): 77-87, june 2019 83 from our data, we discovered that, not all minutiae types (mts) were not present in all individuals. the only constant minutiae types (mts) in the population were the bifurcations and convergence (bc), fragment/point or dot and the bridge types (table 4), which showed varying levels of presence ranging from 77 93%, while the mt with the lowest presence level was the trifurcations; dock and others types, found in only 13% of the study population. furthermore, the frequencies of presence or absence of the mts showed different variations according to the ethnic groups. however, dependence of the level of presence of the mts on tribal classifications was only statistically significant in the trifurcations; dock and others mt at p<0.001, while the dependence was significant at p<0.01 for overlap and return mts (table 4). while the tribal classification dependence on gender was not found to be significant for any of the classes of tmc (figure 2), the analysis of the distribution showed that gender dependence was significant for the distribution of some mts within the study population. fig 2. profile plot of the interaction of ethnicity and gender in the distribution total minutiae counts (tmc); for both hands (a), the right hand (b) and the left hand (c). c u.u. akpan et al fingerprint minutiae of igbo and yoruba ethnic groups ruhuna journal of science vol 10(1): 77-87, june 2019 84 we went further to check the dependent association between gender classifications and the distribution of the mts (table 5). the mts distributions were found to be gender dependent in the bifurcations and convergence (bc), ridge ending (er), fragment/point or dot (fp), overlap (ol) and break (br) mts. furthermore, we also checked the tribal distributions of these mts and their corresponding gender dependent association. table 5. distribution of different minutiae types between the genders. male female pvalue (fischer’s test) present (%) mean ± sd present (%) mean ± sd bc igbo bc 100 100 132.36 ± 31.07 127.20 ± 25.23 100 100 138.03 ± 36.17 162.05 ± 13.43 0.008 0.004 er yoruba er 93.94 86.67 18.30 ± 16.40 16.12 ± 16.42 90.91 85.71 6.21 ± 5.23 5.41 ± 4.32 0.001 0.008 fp igbo fp 100 100 30.11 ± 15.08 30.34 ± 15.12 100 100 32.15 ± 11.53 39.52 ± 4.23 0.001 0.003 ol yoruba ol 93.94 86.67 11.80 ± 8.22 10.36 ± 9.68 81.82 71.43 7.21 ± 5.76 4.72 ± 3.56 0.003 0.007 br igbo br 90.91 88.89 11.01 ± 9.6 13.03 ± 9.51 90.91 100 5.23 ± 4.23 8.18 ± 5.44 0.0001 0.004 igbo en 100 10.09 ± 6.43 100 6.61 ± 1.57 0.007 igbo rt 88.89 8.01 ± 8.62 100 5.72 ± 1.03 0.005 4. discussion in the application of fingerprinting for forensic identifications, ridge count and pattern type are not the only features in used. the absolute position, direction, and type of minutiae (e.g. termination or bifurcation) are also used. as important as minutiae are, only a few human population studies have been carried out (gutierrez-redomero et al. 2012) while fewer none has been carried out in nigeria. the few studies on minutiae in nigeria did not set out to evaluate the population characteristics of pattern. in the use of minutiae feature of fingerprint in identification, the frequency of appearance in the population is an important factor to consider (gutiérrez-redomero et al. 2011). this is because the features that are frequent in the population will be less indicative or unique to individuals than the rare features in the population. gutierrez-redomero et al. (2012) stated that the frequency of minutiae change in relation to population. u.u. akpan et al fingerprint minutiae of igbo and yoruba ethnic groups ruhuna journal of science vol 10(1): 77-87, june 2019 85 the most frequent minutiae type in our study, i.e. bifurcations and convergence, which accounted for more than 54.85% of the study’s total minutiae counts (tmc) is way higher than the 28.23 % recorded in a spanish population (gutiérrez-redomero et al. 2011) for the same combined minutiae type. we collected our samples manually as against digital collection used in the case of the spanish population and the dichotomy in the sample sizes. the rarity of the bifurcations and convergence may have also buttressed its value for identification in spanish population, but not in nigerian population as they are more frequent. also ridge ending with a frequency of about 5% contradicts 65% frequency recorded in the same spanish population. therefore, further studies should be carried out to establish the variation in minutiae frequency in relation to population, especially using digital tools as used in the earlier minutiae studies. although okajima (1970), gutierrez-redomero et al. (2007) and gutierrez-redomero et al. (2011) found basis for statistical association between some minutiae types’ distribution and gender, our study was able to find this association for mts that showed no association in previous studies including fragment/point or dot (fp), overlap (ol) and break (br) minutiae. because there has been no work on the distribution study of minutiae on any nigerian population, the results of this study could not be compared with previous study on nigerians, hence further studies are still needed if the application of minutiae feature of fingerprint in forensic identification is going to be taken seriously in nigeria. 5 conclusions the results of this study have provided us with insights on more morphological differences that exist even between different ethnic populations of the same country especially in dermatoglyphics. this knowledge if well researched on and harnessed can be an efficient and cost effective tool in forensic identifications of individuals of different ethnic groups in developing countries like nigeria. references aase jm, lyons rb. 1971. technique for recording dermatoglyphics. lancet (london, england) 1(7696): 432–433. adamu l. 2017. sex variation in thumbprint minutiae among hausa lineage. journal of anatomical sciences 8: 39-47. adamu lh, ojo sa, danborno b, adebisi ss, taura mg. 2016a. sexual dimorphism in epidermal ridge density and thickness asymmetry indices among hausa population of kano state nigeria. nigerian journal of experimental and clinical biosciences 4:42-47. u.u. akpan et al fingerprint minutiae of igbo and yoruba ethnic groups ruhuna journal of science vol 10(1): 77-87, june 2019 86 adamu lh, ojo sa, danborno b, adebisi ss, taura mg. 2016b. the potential of thumbprint profiles in prediction of body mass index among hausa ethnic group of nigeria. bayero journal of biomedical science 1(1):69-77. adamu lh, ojo sa, danborno b, adebisi ss, taura mg. 2017a. evaluation of facial proportions and their association with thumbprint patterns among hausa ethnic group. journal of anthropology 2: 7-18. adamu lh, ojo sa, danborno b, adebisi ss, taura mg. 2017b. evaluation of asymmetry using thumbprint minutiae among hausa population of kano state, nigeria. bayero journal of pure and applied sciences 10(1): 39 – 46. adamu lh, taura mg. 2017. application of likelihood ratio and posterior probability density in sex estimation from level two fingerprint features among hausa ethnic group. egyptian journal of forensic sciences 7:25. adekoya ko, ahmed ra, oboh bo, alimba cg. 2013. relationships between dermatoglyphics and multiple intelligence among selected secondary school students in lagos state, nigeria. nigerian society for experimental biology (niseb) journal 13 (3& 4): 53-60. akingbade am, saalu lc, akunna gg, anderson le, olusolade fs. 2014. finger and palmar dermatoglyphic study among the yorubas in jos, nigeria. annals of bioanthropology 2:4953. blangero j. 1990. population structure analysis using polygenic traits: estimation of migration matrices. human biology 62(1): 27–48. champod c, lennard cj, margot p, stoilovic m. 2004. fingerprints and other ridge skin impressions. crc press, pp 418. crawford mh, duggirala r. 1992. digital dermatoglyphic patterns of eskimo and amerindian populations: relationships between geographic, dermatoglyphic, genetic, and linguistic distances. human biology 64(5): 683–704. dankmeijer j, waltman jm, de wilde ag. 1980. biological foundations for forensic identifications based on fingerprints. acta morphologica neerlando-scandinavica 18(1): 67– 83. dike eu, oladipo gs, okoh pd. 2012. a comparative study of the digital pattern, position of triradii, b-c and a-d palmar distances of diabetic subjects and essential hypertensive individuals in river state. international journal of advanced biotechnology and research 3(2): 615-620. durham nm, plato cc. 1990. trends in dermatoglyphic research. 1st ed. eds. n.m. durham, plato, c. c. netherlands: springer netherlands. pp724 eboh deo. 2013. fingerprint patterns in relation to gender and blood group. journal of clinical anatomy, 12(2): 82 – 86. faigman dl, saks mj, sanders j. 2008. modern scientific evidence: standards, statistics, and research methods. thomson/west. pp322. gutiérrez-redomero e, alonso-rodríguez c, hernández-hurtado le, rodríguez-villalba jl. 2011. distribution of the minutiae in the fingerprints of a sample of the spanish population. forensic science international 208(1): 79–90. gutiérrez-redomero e, alonso-rodríguez c, hernández-hurtado le, rodríguez-villalba jl. 2012. are there population differences in minutiae frequencies? a comparative study of two argentinian population samples and one spanish sample. forensic science international 222(1): 266–276. gutiérrez-redomero e, galera v, martínez jm, alonso-rodriguez c. 2007. biological variability of the minutiae in the fingerprints of a sample of the spanish population. forensic science international 172(2): 98–105. iloanusi on. 2011. comparison of the minutiae quadruplets and minutiae triplets techniques. nigerian journal of technology 30(3): 28-33. jaja bnr, olabiyi o, noronhhe, cc. (2011) dermatoglyphics of the ogoni of nigeria and its historiographic implications. anthropologiseher anzeyer, 68 (2): 175-183 u.u. akpan et al fingerprint minutiae of igbo and yoruba ethnic groups ruhuna journal of science vol 10(1): 77-87, june 2019 87 janti rl breheme h. 1978. finger and palmar dermatoglyphics of a yoruba (nigerian) sample. annals of human biology 5: 539-546 jantz, rl, brehme, h. 1978. finger and palmar dermatoglyphics of a yoruba (nigeria) sample. annals of human biology 5 (6): 539-546 karthick r. 2015. dermatoglyphics -a review. biomedical and pharmacology journal 8se (1905): 417–420. lestrange, md. 1953. les crates papillaires digitales de 1.491 noirs d'afrique occidentale. bulletin de l'institut franeais d'afrique noire, 15: 1278-1315. loesch dz. 1983. quantitative dermatoglyphics: classification, genetics, and pathology. oxford university press. pp 134 meier rj. 1980. anthropological dermatoglyphics: a review. american journal of physical anthropology 23(1 s): 147–178. midlo h, cummins c. 1943. finger prints, palms and soles; an introduction to dermatoglyphics. philadelphia: blakiston. pp: 332 mohammed b, garba sh, adeyemi lb. 2014. digital dermatoglyphics patterns of the kanuri ethnic group of north eastern nigeria. international journal of innovation and applied studies 9(2): 985-988. national population commission (npc). 2006. report on the final census result. npc press 2006. from: http://www.population.gov.ng accessed on march 22, 2018. ogunnaike po, owolabi jo, ogunsola ao, olanrewaju ja. 2014. human identification: assessment of interrelationships between sex, handedness and dermatoglyphics. journal of medical and applied biosciences 6 (2): 44 – 54. ojikutu ro. 1964. a qualitative and quantitative analysis of finger and palmar cutaneous dermatoglyphics in the nigerian population. homo 15: 160-164. okajima m. 1970. frequency of forks in epidermal-ridge minutiae in the finger print. american journal of physical anthropology 32(1): 41–48. oladipo gs, afolabi eo, esomonu c. 2010. dermatoglyphic patterns of obese versus normal weight in nigerian individuals. biomedicine international 1: 66-69. oladipo gs, osogbe lg, bobmanuel i, ugbonna haa, sapira mk, ekeke on. 2010. palmer dermatoglyphics in essential hypertension amongst rivers indigenes. australian journal of basic and applied sciences, 4(12): 6300 -6305. oladipo gs, paul cw, bob-manuel if, iboroma ad, fawehinmi h, edibamode e. 2009. study of digital and palmar dermatoglyphic patterns of nigerian women with malignant mammary neoplasm. journal of applied biosciences 15: 829 834. orike s, anireh vie, ademuyiwa, si .2016. a gender and ethnicity identification system in nigeria using the fingerprint technology. proceedings of the world congress on engineering 1:1-5. osunwoke ea, amah-tariah fs, sapiraand mk, onosigho a. 2010. dermatoglyphic patterns in congenital deaf and mute in south-south nigeria. african journal of medicine, physiology, biomed engineering & science 2: 98 – 101. otobo tm, tarimobo-otobo, r. 2016. digital and palmer dermatoglyphic characteristics of the ijaw ethnic group. international journal of forensic medical investigation. 2(1)25-30. pankanti s, prabhakar s, jain ak. 2002. on the individuality of fingerprints. ieee transactions on pattern analysis and machine intelligence 24(8): 1010–1025. schaumann b, alter m. 1976. dermatoglyphics in medical disorders. 1st ed. berlin: springer-verlag berlin heidelberg, pp 1075. segura-wang m, barrantes r. 2009. dermatoglyphic traits of six chibcha-speaking amerindians of costa rica, and an assessment of the genetic affinities among populations. revista de biologia tropical 57(suppl. 1): 357–369. umana ue, ahunna co, timbuak, ja, ibegbu ao, musa sa, hamman wo. 2013. dermatoglyphics and cheiloscopic patterns in cancer patients; a study in ahmadu bello university teaching hospital (abuth), zaria, nigeria. current research journal of biology sciences, 5(5): 220. ruhuna journal of science vol 13 (1): 52-60, june 2022 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v13i1.115 university of ruhuna faculty of science, university of ruhuna sri lanka 52 abundance and ecological classification of fish species: a case study of owalla reservoir, osun state, nigeria adams ovie iyiola1*, yetunde folasadetaiwo2 and berchie asiedu3 1department of fisheries and aquatic resources management, faculty of renewable natural resources management, college of agriculture, ejigbo campus, osun state university, nigeria 2 natural history museum, obafemi awolowo university, ile ife, nigeria 3 department of fisheries and water resources, school of natural resources university of energy and natural resources, sunyani, ghana *correspondence: adams.iyiola@uniosun.edu.ng; orcid: https://orcid.org/0000-0002-2166-7299 received: 2nd august 2021, revised: 18th january 2022, accepted: 24th april 2022 abstract globally, we are experiencing higher species extinction rates than ever before which can critically affect food and nutrition security. the ecological classification of fish species was assessed in owalla reservoir, okinni, osun state, nigeria. fish landings of fishermen were sampled for 12 months from september 2019 to august 2020). the species were identified and grouped into herbivores, carnivores, and omnivores, and the fish abundance was recorded. the foragecarnivore ratio (f/c ratio) was calculated to express the ecological structure of the reservoir. a total of 15 fish species (n=1035 fish) were identified. in the dry and wet seasons, 15 species (692 fish) and 11 species (343 fish) were recorded, respectively, and herbivores were more abundant than carnivores. chrysichthys nigrodigitatus (25.9%) was the most abundant species and cichlidae (55.5%) was the most abundant family. the overall f/c ratio during the combined, dry and wet seasons were 1.35, 1.01 and 2.57, respectively. the ecological structure as indicated by the f/c ratio was unbalanced, but in the wet season, it was close to being balanced. we proposed management measures such as fishing across the trophic levels, efficient monitoring and surveillance, regulation of mesh sizes, and registration of fishermen in addition to routine data collection. keywords: feeding classification, fish abundance, fish composition, owalla reservoir, sustainability. 1 introduction fish abundance and growth in tropical waters can be influenced by food availability, changes in environmental factors, fishing pressure, pollution, water depth (soyinka et al. 2010) and climate change. exploitation is a major tool that alters the fish species https://rjs.ruh.ac.lk/index.php/rjs/index https://creativecommons.org/licenses/by-nc/4.0/ mailto:adams.iyiola@uniosun.edu.ng https://orcid.org/0000-0002-2166-7299 a. o. iyiola et al. ecological classification of fish species of owalla reservoir ruhuna journal of science vol 13 (1): 52-60, june 2022 53 diversity, food webs, and species interaction in an ecosystem. fishing can cause changes in the food web interactions because it can reduce the components in the food web thereby altering the interactions and duration of interaction between fish species (nrc 2012). from an ecological view, all fish species have a particular function which is important in regulating various functions in an ecosystem. therefore, fishing may either reduce the lower trophic level from predation or reduce the availability of prey for higher trophic level predators. when this occurs, the fish species in an ecosystem may tend to have an unbalanced assemblage. over the years, fluctuations in fish abundance have been reported in owalla reservoir, nigeria (taiwo 2010) and it was attributed to human activities and exploitation. despite the small catchment area, owalla reservoir in osun state, nigeria has been reported to be rich in fish species. as a result, there have been intense fishing activities that might have caused the trophic balance of fish species in the reservoir. taiwo (2010) reported an unbalanced fish population with an abundance of piscivorous fish species. thus, there is a need to constantly monitor the fish resources in this water body. abban et al. (2001) reported that studies on fish abundance in african waters are inadequate, and information is scarce for decision-making. of all ecosystems, freshwaters are among the most threatened in the world (dudgeon et al., 2006). over the past two decades, research efforts have been tailored toward the enhancement of inland fisheries in nigeria (neiland and ladu 1998). due to the lack of documented information on the ecological structure of fish species in the owalla reservoir, the present study aims to investigate the current status of fish species, feeding categories and their ecological relationship in the reservoir. this will enable the development of appropriate management measures by the relevant agencies and stakeholders for the sustainability of the reservoir fisheries as well as the acceleration of nigeria's effort towards achieving the 2030 agenda of the sustainable development goals (goal 6 “clean water and sanitation; goal 14: life below water). 2 material and methods 2.1 study area owalla reservoir is located between latitudes 7°44ʹ30.44ʺ and 7°57ʹ00.79ʺ n, and longitudes 4°26ʹ21.71ʺand 4°41ʹ23.48ʺ east of the greenwich meridian in okinni town, osun state, nigeria (figure 1). the reservoir is situated in the erinle river basin, which is part of the osun river basin (ita 1993). it supplies water to osogbo and the environs. its elevation is between 250 and 400 meters above sea level with a maximum width of 3.5m. diverse human activities such as farming, laundry activities and dredging were observed around the reservoir. some parts of the reservoir were shaded by dense vegetation. three sampling sites were selected in the reservoir based on fishing activities. a. o. iyiola et al. ecological classification of fish species of owalla reservoir ruhuna journal of science vol 13 (1): 52-60, june 2022 54 fig 1. map of owalla reservoir, osun state, nigeria 2.2 fish sampling and identification fishermen’s landings were collected fortnightly from purposively identified points on the reservoir from september 2019 to august 2020. the gears used for fishing were monofilament gill nets of 25mm, 40mm, 45mm and 60mm mesh sizes. the fishing nets were set by the fishermen at 2000gmt and retrieved at 0700gmt. the fish species were sorted and identified with combination of guides by holden and reeds (1978) and olaosebikan and raji (2013). the relative abundance of fish species and families was recorded by months and seasons (dry and wet seasons). 2.3 feeding categories and ecological structure of sampled fish species the sampled fish species were grouped based on feeding habits as described by holden and reeds (1978) into three, namely; i. herbivores which are the primary consumers and are plant eaters ii. carnivores which are the secondary consumers and are flesh eaters and highly predatory iii. omnivores which fed on both plants and animals, although omnivory is defined as feeding at more than one trophic level (pimm and lawton 1974; williams and martinez 2004) a. o. iyiola et al. ecological classification of fish species of owalla reservoir ruhuna journal of science vol 13 (1): 52-60, june 2022 55 the abundance of fish species was computed based on their feeding habits and seasons. the forage/carnivore ratio (f/c) as stated by ipinmoroti (2013) was based on fish abundance and used to determine the ecological structure of the fish species. it was calculated as follows: 𝐹 𝐶 = 𝐻𝑒𝑟𝑏𝑖𝑣𝑜𝑟𝑒𝑠 𝐶𝑎𝑟𝑛𝑖𝑣𝑜𝑟𝑒𝑠+𝑂𝑚𝑛𝑖𝑣𝑜𝑟𝑒𝑠 2.4 statistical analysis the data collected on fish abundance were analyzed using descriptive statistics involving percentages. the comparison between mean fish abundance per feeding habits in the wet and dry seasons was done using student's t-test to determine their level of significance (p<0.05). minitab 17.0 statistical software was used for statistical analyses. 3 results 3.1 sampling and identification of fish table 1: relative abundance of fish species identified during the study. family species abundance relative abundance abundance (family) relative abundance (family) cichlidae oreochromis niloticus 214 20.7 coptodon marie 205 19.8 sarotherondon galileaus 44 4.3 coptodon zilli 54 5.2 coptodon aurea 5 0.5 hemichromis fasciatus 52 5.0 574 55.5 gymnarchidae gymnarchus niloticus 1 0.1 1 0.1 hepsetidae hepsetus akawo 9 0.9 9 0.9 mormyridae mormyrus rume rume 79 7.6 mormyrops anguilloides 40 3.9 119 11.5 schilbeidae schilbe mystus 13 1.3 13 1.3 clariidae clarias gariepinus 3 0.3 3 0.3 bagridae chrysichthys nigrodigitatus 268 25.9 268 25.9 mochokidae synodontis budgetti 27 2.6 27 2.6 clupeidae sardina pilchardus 21 2.0 21 2.0 the relative abundance of fish species and grouping by family is presented in table 1. a total of 1035 individuals comprising of 15 species belonging to 9 families were identified. cichlids were the most abundant family (55.5%) and comprised of 6 species with oreochromis niloticus being the most abundant species (20.7%). gymnarchidae a. o. iyiola et al. ecological classification of fish species of owalla reservoir ruhuna journal of science vol 13 (1): 52-60, june 2022 56 was the least abundant family (0.1%) which comprised of only 1 species. chrysichthys nigrodigitatus was the most abundant (25.9%) and gymnarchus niloticus was the least abundant (0.1%) fish species identified during the entire study. 3.2 ecological structure of fish species the sampled fish species during the study period were grouped based on their feeding habits into herbivores, carnivores and omnivores (table 2). table 2. trophic classifications of fish species encountered during the study period (abundance values are given for wet and dry seasons separately and combined). cra= combined relative abundance; ne = not encountered herbivores were the most abundant in the combined season with an average of 57.5%, and in the wet (72.0%) and dry (50.3%) seasons with 7 fish species, respectively. o. niloticus was the most abundant in the combined with an average of 36.0% and in the wet season (36.8%) while c. marie (44.0%) was the most abundant in the dry season. omnivores had a relative abundance of 40.3% (5 species), 22.7% (4 species) and 40.9% (5 species) for the combined, wet and dry seasons, respectively. c. nigrodigitatus was the most abundant omnivore with 64.3%, 59.9% and 83.3% in the trophic group/ species wet season dry season wet + dry seasons cra herbivores o. niloticus 91 123 214 36.0 c. marie 52 153 205 34.5 s. galileaus galileaus 36 8 44 7.4 c. zilli 48 6 54 9.1 h. fasciatus ne 52 52 8.7 c. aurea ne 5 5 0.8 s. pilchardus 20 1 21 3.5 total (herbivores) 247 348 595 100 carnivores g. niloticus ne 1 1 4.3 h. akawo 8 1 9 39.1 s. mystus 10 3 13 56.5 total (carnivores) 18 5 23 100 omnivores c. gariepinus 2 1 3 0.7 c. nigrodigitatus 65 203 268 64.3 s. budgetti 2 25 27 6.5 m. rume rume ne 79 79 18.9 m. anguilloides 9 31 40 9.6 total (omnivores) 78 339 417 100 total fish species 343 692 1035 a. o. iyiola et al. ecological classification of fish species of owalla reservoir ruhuna journal of science vol 13 (1): 52-60, june 2022 57 combined, wet and dry seasons, respectively. carnivores were the least abundant family with 2.2% (comprising of 3 species), 0.7% (comprising of 5 species) and 5.2% (comprising of 2 species) in the combined, wet and dry seasons, respectively. schilbe mystus (56.5%), hepsetus akawo (60.0%) and schilbe mystus (55.6%) were the most abundant carnivores in the combined, dry and wet seasons, respectively. the f/c ratio was calculated as 1.35, 1.01 and 2.57 for the combined, dry and wet seasons, respectively. the illustration on the trophic classification of sampled fish species is presented in fig 2. fig 2. trophic web locations of identified fish species in the reservoir (adopted from adeogun et al. 2020) 3.3 forage-carnivore ratio the f/c ratio calculated for the entire study (combined), dry season and wet seasons were 1.35, 1.01 and 2.57, respectively. the fish species richness identified in each feeding habit was high for herbivores, carnivores and omnivores in the dry season with 7, 3 and 5, respectively (table 3) whereas a significant value (p<0.05) was observed for the herbivore species between the wet and dry seasons. a. o. iyiola et al. ecological classification of fish species of owalla reservoir ruhuna journal of science vol 13 (1): 52-60, june 2022 58 table 3: summary of fish species richness in each feeding habit. feeding habit dry season wet season dry + wet (combined) a t-value a significance herbivores 7 5 7 -0.81 0.049* carnivores 3 2 3 2.51 0.087 omnivores 5 4 5 -2.15 0.098 total species 15 11 15 f/c ratio 1.01 1.35 1.35 a t-test comparison between dry season and wet season (*significance at p<0.05); f/c = forage (herbivores)/ carnivore + omnivores 4 discussion the results recorded for fish abundance (1035 individuals; 9 families and 15 species) deviated from studies by komolafe et al. (2014) and badejo and oriyomi (2015) who reported fish abundance of 547 individuals (8 species; 4 families) in osinmo reservoir and 561 individuals (12 species; 7 families) in erinle reservoir, respectively. both studies reported a reduction in fish abundance and diversity while taiwo et al. (2018) reported a higher abundance and diversity with 1915 individuals (17 species, 10 families) in opa reservoir. chrysichthys nigrodigitatus was observed to be the most abundant fish species with 25.9% while the least was gymnarchus niloticus with 0.1%. across the families, cichlidae was the most abundant family (55.5%) comprising of 6 fish species namely coptodon marie, c. zilli, c. aurea, hemichromis fasciatus, oreochromis niloticus and sarotherondon galileaus galileaus. the abundance of the family cichlidae has been reported from various reservoirs in osun state; ipinmoroti (2013) reported 61.1% abundance in asejire reservoir, komolafe et al., (2014) reported 83% in osinmo reservoir with 5 species while taiwo et al. (2018) reported 89.8% in opa reservoir with 7 species. this is due to the high prolific ability of cichlids (negi and mamgain, 2013). out of the total of 1035 fish samples identified during the study, herbivores were observed to be most abundant with 595 individuals (57.5%) from 7 fish species. o. niloticus was the most abundant species in the herbivores group (36.0%) whereas c. aurea was the least (0.8%). herbivores which comprised mostly of cichlids are at the base of the food chain and the most abundant in the ecosystem. these results corroborated the findings of iyiola et al. (2020), komolafe et al. (2016) and taiwo et al. (2018) that reported an abundance of herbivores in osun river as 34.14%, in abandoned goldmine reservoirs in igun as 86.2% and in opa river as 89.8%. the values reported by iyiola et al. (2020) were however lower than the results from the present study. the omnivores were next abundant with 417 individuals (40.3%) from five (5) species with c. nigrodigitatus being the most abundant fish in the group (64.3%) and the least was clarias gariepinus (0.7%). the least trophic group was the carnivores with 23 individuals (2.2%) comprising 3 fish species. when fishes with commercial value (in this case carnivores) are fished, there is more space for the herbivores. a. o. iyiola et al. ecological classification of fish species of owalla reservoir ruhuna journal of science vol 13 (1): 52-60, june 2022 59 herbivores function in balancing the number of microalgae and sharpening the aquatic system. based on this, their abundance in the reservoir indicates the availability of food for other trophic levels to attain ecological stability (icri 2021). the combined forage-carnivore ratio (f/c) calculated indicates an unbalanced population of fish species because it was not within the range of 3–6 as stated by swingle (1950). the f/c values from the current study were lower except for the wet season which was higher. the ratio value recorded during the wet season (2.57) close to 3 was expected because the carnivores and omnivores were a few when compared to herbivores which almost tripled the total abundance. for the dry season and the entire study period which had relatively low f/c values, the abundance of carnivores and omnivores almost equated to the herbivores. balayut (1983) suggested management measures such as fishing across the trophic levels when an unbalanced fish population occurs as observed during this study. based on this, fishing across the herbivores is suggested as a management measure for attaining a balanced population of fish species. similar results of f/c ratio indicating an unbalanced population were reported in osun river by iyiola et al. (2020) who stated a ratio of 0.64, 0.34 and 0.48 in the dry, wet and combined seasons, respectively. these ratios were lower when compared with the results of this study. the fish species richness identified in each feeding habit indicated that herbivores were abundant in the reservoir during the period of study. this is similar to the findings by taiwo (2010) and badejo and oriyomi (2015) who reported a similar abundance of herbivores in owalla and erinle reservoirs, respectively. a significant difference (p=0.049) was observed between the herbivorous fish species identified in the dry and wet seasons. the possible reason for this is the breeding activities fish species undergo during the wet season which reduced their abundance when compared to the dry season. 5 conclusions it was observed that the reservoir is rich in fish abundance and diversity when compared with other reservoirs within south-western nigeria. c. nigrodigitatus was the most abundant, followed by two cichlid species. the herbivores constituted the largest trophic group of fish species observed during the study. the f/c ratio observed during the entire study and dry season was not balanced while the ratio observed in the wet season was close to balanced. the unbalanced state of fish species is a result of the relative abundance of carnivores. management measures such as fishing across the trophic levels for carnivores, monitoring, surveillance, regulation of mesh sizes and registration of fishermen are proposed for a balanced ecological structure. acknowledgments the authors are grateful to the anonymous reviewers for their comments. a. o. iyiola et al. ecological classification of fish species of owalla reservoir ruhuna journal of science vol 13 (1): 52-60, june 2022 60 references abban ek, dankwa hr. 2001. amanzuri conservation and integrated development (acid) project. ghana wildlife society: fish biodiversity baseline study. adeogun ao, ibor or, khan ea, azubuike v, chukwuka av, omogbemi ed, arukwe a. 2020. detection and occurrence of microplastics in the stomach of commercial fish species from a municipal water supply lake in southwestern nigeria. environmental science and pollution research 27: 31035– 31045. https://doi.org/10.1007/s11356-020-09031-5 badejo oa, oriyomi o. 2015. seasonal variation, abundance and condition factor of fish species in erinle reservoir. american scientific research journal for engineering, technology and sciences (asrjets) 12(1): 136-142. balayut ea. 1983. stocking and introduction of fish in lakes and reservoirs in the asean countries. fao fisheries technical paper. p. 236. dudgeon d, arthington ah, gessner mo, kawabata zi, knowler dj, lévêque c, naiman rj, prieurrichard ah, soto d, stiassny ml. 2006. freshwater biodiversity: importance, threats, status and conservation challenges. biological reviews 81(2): 163–182. holden m, reeds w. 1978. west african freshwater fish. longman group ltd, london. isbn 0 582 60426 5. international coral reef initiative (icri) 2021. the importance of herbivorous fish. available at https://icriforum.org/the-importance-of-herbivorous-fish/ ipinmoroti mo. 2013. ichthyofauna diversity of lake asejire: ecological implications. international journal of fisheries and aquaculture 5(10): 248-252. http://dx.doi.org/10.5897/ijfa13.0379. iyiola ao, kolawole as, akanmu oa, ayanboye ao, ipinmoroti mo. 2020. food habit and ecological balance of fish species in osun river, nigeria. the proceedings of the nigerian academy of science 13(1): 57-67. komolafe oo, adedeji aa, fadairo b. 2014. assessment of the water quality parameters in relation to fish community of osinmo reservoir, ejigbo, osun state. international journal of biological and chemical sciences 8(2): 596-609. komolafe oo, olofinmehinti mo, adedeji aa, adewalem ha. 2016. the distribution, composition and abundance of fish species in two abandoned gold mine reservoirs, igun, osun state, nigeria. ife journal of science 18(1): 095-102. negi rk, mamgain s. 2013. species diversity, abundance and distribution of fish community and conservation status of tons river of uttarakhand state, india. journal of fisheries and aquatic science, 8: 617-626. neiland ae, ladu bmb. 1998. enhancement of inland fisheries in nigeria: the institutional context provided by traditional and modern systems fisheries management. inland fisheries enhancement, fao technical paper no 374: 371-393. nrc 2012. dynamic changes in marine ecosystems: fishing food webs, and future options: phase ii assessment of the extent of change and the implication for fishery. national research council. olaosebikan bd, raji a. 2013. field guide to nigerian freshwater fishes (revised edition), remi thomas press. isbn 978-34-760-0-9. pimm c, lawton a. 1978. on feeding on more than one trophic level. nature 275: 542–544. soyinka oo, kuton mp, ayo-olalusi ci. 2010. seasonal distribution and richness of fish species in the badagry lagoon, south-west nigeria. estonian journal of ecology 59: 147-157. swingle hs. 1950. relationships and dynamics of balanced and imbalanced fish populations. alabama agricultural experiment station. taiwo yf. 2010. fish diversity in two reservoirs in south western nigeria. proceedings of the fisheries society of nigeria (fison) ascon, badagry 25-29 october, 2010. fns-fb-0007; issn:1117-3149. taiwo yf, adedeji aa, adesakin aa. 2018. the ecological balance of the fish community in an aging reservoir. natural resources 9: 327-335. williams rj, martinez nd. 2004. limits to trophic levels and omnivory in complex food webs: theory and data. american naturalist 163: 458–468. https://doi.org/10.1007/s11356-020-09031-5 https://www.nature.com/ sv-lncs ruhuna journal of science vol. 7: 43-57, december 2016 eissn: 2536-8400  faculty of science university of ruhuna  faculty of science, university of ruhuna 43 sri lanka assessing the geo-electric characteristics of basement complex rocks and its implication for groundwater prospecting in ilorin metropolis, nigeria i. p. ifabiyi 1, e. d. ashaolu2 and o. omotosho3 department of geography and environmental management, university of ilorin, ilorin, nigeria correpondence: 1tokunifabiyi@yahoo.com, 2damash007@yahoo.com 3olumideomotosho47@gmail.com received: 6th october 2016, revised: 21st december 2016, accepted: 24th december 2016 abstract. in basement complex rocks where rainfall is seasonal, water provision in dry season depends on regolith aquifer. for effective exploitation of groundwater resources, it is reasonable that geophysical investigation be conducted before development of well. in many instances, geophysical surveys may be expensive or nonexistent. hence, there is a need for spatial analysis which might advise water engineers within such environments. vertical electrical soundings (ves) data of 53 locations conducted with abem sas-1000 terrameter using schlumberger electrode configuration were obtained from the hydrogeology department of kwara state ministry of water resources and lower niger river basin and rural development authority, ilorin. ves locational coordinates were recorded using handheld gps device. sound curves were evaluated by partial curve matching approach and computer iteration using winresist. the results depict six geo-electric regional successions, namely: top soil, lateritic clay, weathered basement, fairly-hard basement, thin fractured and hard basement. the geo-electric succession identified was plotted in surfer 12 environment, using kriging interpolation method to show spatial distribution pattern of this zone. the spatial pattern is expected to give an insight to the nature of spatial variability of geo-electric layers and assist drillers as well as water resources policy makers in their operations. keywords. geo-electric, groundwater, basement complex, geophysical, gis 1 introduction groundwater occurrence is often localized and confined to weathered or fractured zones in the basement complex region, and groundwater exploration in such terrain is always complex. the crystalline basement rocks mailto:2damash007@yahoo.com i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 44 vol. 7: 43-57, december 2016 have low porosity and permeability, hence, have no water storage capacity in their unaltered form; which makes their groundwater prospects to be limited and often restricted to features produced by weathering and tectonic processes (olayinka and olorunfemi, 1992; olorunfemi and fasuyi, 1993; oyedele et al., 2013). in tropical basement rock, weathering process creates superficial layers, with varying degrees of porosity and permeability. this unconsolidated superficial layers, if significantly thick, porous and permeable, makes good aquifer units. it is important to note that the concealed basement rock may contain faulted areas, incipient joints and fracture systems derived from earlier tectonic processes in such region. the detection and delineation of these hydro-geologic structures may facilitate the location of groundwater potential zones in a typical basement rock environment (omosuyi, et al., 2003; oyedele, et. al., 2013). olorunfemi and fasuyi (1993) submitted that the highest groundwater yield in basement terrains is found in areas where thick overburden overlies fractured zones. these fractured zones are often characterized by relatively low resistivity values. olayinka and olorunfemi (1992) argued that before a borehole is sited, a surface geophysical survey such as vertical electrical resistivity sounding (ves) should be conducted to identify the localized aquifer for a productive well. the vertical electrical resistivity sounding survey provides information about the subsurface that aid in aquifer delineation and identification of lithologic boundaries and geological structures (bose, et al., 1973; abiola, et al., 2013). vertical electrical resistivity sounding method has been used widely by scholars in groundwater prospecting especially in the basement complex terrains to get detailed information about hydrogeological settings for groundwater potentials (olorunfemi, 1990; olorunfemi and olayinka, 1992; olorunfemi and fasuyi, 1993; oladapo, et al., 2009; anohanmoran. 2013; ogundana and talabi, 2014). ves is used to determine the vertical variation of electrical resistivity below the earth surface and the potential field generated by the current, and this is because electrical resistivity of most rock depends on the amount of water in their pores. this method proved useful in groundwater studies because neither the structure nor the dynamics of the soil was disturbed (otobo and ifedili, 2005; adiat et al., 2009; ariyo and adeyemi, 2009; anomoharan, 2011; anomoharan, 2013). despite the importance of ves in groundwater prospecting in a basement complex terrain, a better interpretation of hydrogeological data generated from this method often requires that their spatial location be incorporated into the analysis (shahid and nath, 2002). this will reveal the spatial variation of different geo-electric section of basement complex rock which can give a i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 45 vol. 7: 43-57, december 2016 better understanding of the hydrogeological prospect, especially of a large area. consequently, the incorporation of geographical information system (gis) into studies of groundwater prospecting becomes imperative. according to shahid and nath (2002), in recent time, gis is widely used for spatial modeling of hydrogeological prospect of a large area with more reliability on groundwater exploration. further, gis has proved to be an efficient tool in groundwater researches and the inclusion of subsurface information deduced from geo-electric survey can give more realistic picture of groundwater potentiality of an area (saraf, et.al., 1998; krishnamurthy, et.al., 1996; murthy, 2000; shahid and nath, 2002; amaresh and ravi prakash, 2003). ilorin city is underlain by precambrian basement complex; comprising mostly gneiss, granite, schist, undifferentiated meta-sediments rocks and overburden that are composed mainly of clay, sand and silt soils. the residents of this area often augment the public water supply by the kwara state water corporation with groundwater (shallow and deep) because the supply is erratic and unreliable (ifabiyi and ahmed, 2011) and the coverage is limited to some areas (ifabiyi and ashaolu, 2013). the population of this city is rapidly increasing as new residential areas have sprung up and continue springing up in the last decade. all these new residential areas depend solely on groundwater for their domestic needs. on this basis, this study assessed and mapped the geo-electric characteristics of basement complex rock of ilorin, nigeria in order to identify their spatial variation and implication on groundwater prospects of the city. 2 material and methods 2.1 the study area ilorin the kwara state capital is located between latitude 08°24'n and 08°38'n of the equator, and longitude 04°26' e and 04°37'e of the greenwich meridian, and covers about 12km. ilorin is one of the fastest growing urban centers in nigeria. there has been a huge increase in the population of ilorin since it became the state capital in 1976. the population growth rate is much higher than other cities at 2.9 percent of the national growth rate. the 2006 census put the population of ilorin city to about 847,582 (npc, 2006 provisional results). ilorin has a tropical wet and dry climate. wet season is experienced from april to october and dry season from november to march. rainfall condition in ilorin exhibits greater variability both temporarily and spatial. the annual mean rainfall is about 1,200mm, exhibiting the double maximal pattern between april and october of every year. relative humidity varies seasonally with an average of 79.7%. i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 46 vol. 7: 43-57, december 2016 the city is underlain by precambrian basement complex, comprising mostly gneiss, granite, schist, undifferentiated meta-sediments rocks and overburden that are composed mainly of clay, sand and silt soils. the underlying pre-cambrian igneous-metamorphic rock of basement complex is neither porous nor permeable except in places where they are deeply weathered or have zones of weakness. some part of the town is also laid by sedimentary rocks, which contains both primary and secondary laterites and alluvial deposits. groundwater on the alluvium is recharged directly by rainfall or the adjoining overflowing river system. in the dry season, the alluvium sustains considerable subsurface groundwater flow. the alluvial deposits have been exploited, with successful wells and boreholes in ilorin metropolis and its surrounding. the drainage system of ilorin is dendritic in nature, and is dominated by asa river, which flows from south to north and divides the city into two parts, the western and eastern parts. the map of the study area is presented in figure 1. fig 1. geological map of the study area showing the sampled points i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 47 vol. 7: 43-57, december 2016 2.2 methods this study adopted the combination of secondary and primary (field work) data. the secondary data collected are the vertical electrical soundings (ves) data conducted by the kwara state ministry of water resources and lower niger river basin development authority, ilorin. all the vertical electrical soundings (ves) data collected were conducted with abem sas1000 terrameter using the schlumberger electrode configuration, and the electrode spacing (ab/2) varied from 0.1 m to 200m. the results of geophysical survey carried out in 53 different locations in ilorin were collected. the minimum number of ves carried out in each of these 53 locations was 9 and the maximum were 12. the sounding curves were evaluated by partial curve matching method and computer iterations using winresist. the field work was carried out to get the coordinates (locations) of the sampled points using handheld global positioning system device. this was carried out to get the coordinates of all the locations used in the study, which facilitated the plotting of the data in surfer 12 gis software. the resistivity values from the interpretation of the field data using curve matching were used to generate geo-electrical succession. the data on geoelectric layers and the point location collected were used to plot the geoelectric map of the study area. this was carried out in surfer 12 gis software using kriging interpolation method to determine the spatial distribution of the identified layers in the study area. 3 results and discussion 3.1 general pattern of geophysical characteristics of groundwater in ilorin the identified layers resistivity and thickness ranges across the sampled points in the study area are presented in table 1. from the table, six regional geo-electric patterns are discernable, namely: top lateritic sand, lateritic clay, weathered basement, fairly hard basement, thin fractured and hard basement. however, the study looks at the vertical variations of the electrical resistivity recorded from one point to another across the study area, hence, overlaps observed in the reported ves range. the first layer consists of the top soil with resistivity values ranging from 30-3000 ohm-m and thickness ranging from 0.2-1.0 m. the lateritic clay layer is where resistivity ranges between 231400 ohm-m and thickness from 2-20 m. the third lithologic layer is characterized by highly weathered basement with resistivity values between 22-1000 ohm-m and thickness of 2-45 m. electrical resistivity values here are controlled by the degree of water saturation (odunsanya and amadi, 1990; oladipo et al., 2009). i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 48 vol. 7: 43-57, december 2016 table 1: ranges of geo-electric succession in ilorin, nigeria sn layers resistivity (ohms) thickness (m) 1 top lateritic soil 30-3000 0.2-1.0 2 lateritic clay 23-1400 2-20 3 weathered basement 25-1000 2-45 4 fairly hard basement 22-600 6-42 5 thin fractured 55-145 20-40 6 hard basement 35-780 15-42 above the fourth layer is a fairly hard basement weathered and resistivity ranging from 22-600 ohm-m and thickness between 6-42m. the fifth layer represents a thin fractured zone and resistivity ranging from 55145 ohm-m and thickness between 20-40m. the sixth layer represents the hard basement with resistivity values ranging from 35780 ohm-m and thickness between 1542m and above across the study area. figure 2 and table 2 show the resistivity value of each geo-electric succession in the 53 sampled locations across the study area. fig 2. geo-electric section of vertical electrical sounding in ilorin metropolis i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 49 vol. 7: 43-57, december 2016 table 2: geo-electric succession in the study area (res: resistivity ω-m; t: thickness, m) sn description top lateritic soil lateritic clay weathered basement fairly hard basement thin fractured hard basement res. (ω-m) t (m) res. ( ω-m) t (m) res. (ω-m) t (m) res. ( ω-m) t (m) res. ( ω-m) t (m) res. ( ω-m) t (m) 1 apata yakuba 1500 0-1 300 2-3 80 3-20 160 20-40 0 0 270 40-above 2 ganiki sango 85 0-1 130 2-6 90 6-20 190 20-35 0 0 250 35-above 3 royal valley 400 0-1 320 2-3 145 3-15 200 15-30 0 0 400 30-above 4 elekoyangan 120 0-1 na na 180 2-15 250 15-20 0 0 600 20-above 5 sango area 100 0-1 50 2-10 65 10-30 52 30-40 0 0 68 40-above 6 oyun area 100 0-1 120 2-3 65 3-10 200 10-30 0 0 350 30-above 7 alagbado 120 0-1 250 2-6 70 6-20 150 20-35 145 35-40 180 40-above 8 okelele 30 0-1 68 2-8 52 8-25 60 25-35 0 0 66 35-above 9 olorunsogo 170 0-1 160 2-3 35 3-20 40 20-40 0 0 55 40-above 10 agbabiaka area 400 0-1 700 2-15 170 15-30 180 30-40 0 0 190 42-above 11 kilanko area 270 0-1 na na 35 2-20 110 20-40 0 0 150 40-above 12 tanke area 110 0-1 80 2-20 200 20-40 365 40-40 0 0 380 42-above 13 danialu 280 0-1 300 2-3 225 3-20 260 20-30 0 0 450 30-above 14 fate tanke 80 0-1 100 2-5 90 5-10 105 10-20 55 20-30 600 15-30 15 wonderland chapel 200 0-1 1400 2-15 0 0 600 15-30 0 0 780 30-above 16 oloje area 270 0-1 160 2-5 32 5-6 58 20-30 60 30-35 85 35-above 17 asileke 290 0-1 220 2-15 0 0 250 15-25 0 0 450 25-above 18 jooro 25 0-1 85 2-10 72 10-20 100 20-35 0 0 130 30-above 19 okolowo area 220 0-1 70 2-3 30 3-20 40 20-40 0 0 60 42-above 20 yebumot area 1200 0-1 900 2-10 150 10-35 70 35-42 0 0 180 42-above 21 bal. eng. area 300 0-1 80 2-6 52 6-30 65 30-40 0 0 80 40-above 22 hajj camp 340 0-1 450 2-6 110 6-20 200 20-40 0 0 0 0 23 alhikmah univ. area) 710 0-1 350 2-6 60 6-25 125 25-40 0 0 0 0 24 adewole area 120 0-1 200 2-10 50 10-30 80 30-40 0 0 0 0 25 egbejila 550 0-1 700 2-8 260 8-45 300 15-25 0 0 520 25-above 26 asa dam 650 0-1 1200 2-8 100 8-20 150 20-40 0 0 180 42-above 27 pakata 80 0-1 30 2-6 25 6-15 35 15-30 0 0 40 30-above 28 iqra college 3000 0-1 280 25 0 0 120 8-40 0 0 160 40-above 29 baboko market 38 0-1 130 2-5 13 5-30 22 30-40 0 0 40 40-above 30 banni area 80 0-1 130 2-6 0 0 200 6-23 0 0 400 25-above 31 fufu str. sabooke 32 0-1 74 2-10 50 10-30 92 30-35 0 0 130 35-above 32 post office area 200 0-1 92 2-10 70 10-30 85 30-35 0 0 100 35-above 33 min. of agric area 260 0-1 150 2-10 45 10-30 50 30-35 above 0 0 60 35-above 34 ododosowapo amilegbe 340 0-1 100 2-10 56 10-30 43 30-35 0 0 68 35-above 35 onikanga gra 1200 0-1 1000 2-10 1000 10-30 55 30-35 0 0 60 35-above 36 sakama niger rd 20 0-1 35 2-10 330 10-30 220 30-35 0 0 270 35-above 37 okaka niger rd 720 0-1 1102 2-10 180 10-30 50 30-30 0 0 99 30-above 38 agbadam lake rd 50 0-1 65 2-10 90 10-30 80 30-35 above 0 0 110 30-40 39 abdulrasaq rd gra 200 0-1 180 2-8 140 8-30 170 30-35 above 0 0 175 35-above 40 unilorin garden 650 0-1 300 2-8 50 8-30 145 30-35 above 0 0 160 40-above i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 50 vol. 7: 43-57, december 2016 41 tipper garage area 200 0-1 310 2-8 170 8-30 300 30-35 above 0 0 350 30-above 42 behind fed sec. 1000 0-1 350 2-10 10-30 128 30-35 0 0 165 35-above 43 odofin lane basin 100 0-11 260 30-35 above 0 0 450 40-above 44 water view area 100 0-1 380 2-8 240 8-30 300 30-35 0 0 370 35-above 45 adelodun rd fate 140 0-1 90 2-8 55 8-30 70 30-35 0 0 92 35-above 46 edun area 265 0-1 130 2-8 55 8-30 78 30-35 0 0 82 35-above 47 behind amusement park 35 0-1 23 2-10 33 10-30 40 30-35 0 0 45 35-above 48 mustapha idiagbede basin 140 0-1 100 2-10 70 10-30 95 30-35 0 0 120 35-above 49 adeleye str. gra 800 0-1 300 2-10 55 10-30 120 30-35 0 0 130 35-above 50 gss area 80 0-1 2-10 28 10-30 40 30-35 0 0 60 35-above 51 kwara adp area 300 0-1 360 2-10 55 10-30 60 30-35 0 0 100 35-above 52 ile eleru ojagbooro 40 0-1 2-10 35 10-30 30 30-35 115 35 35-above 53 office rd gra 300 0-1 97 2-10 32 10-30 51 30-35 above 0 0 55 40-above 3.2 geo-electric succession and their characteristics in ilorin metropolis 3.2.1 top lateritic soil and lateritic clay the average resistivity and thickness values of top lateritic soil in apata yakuba, ganiki sango and royal valley are 662 ohm-m and 1m respectively which indicate lower clay proportion. the average resistivity and thickness values of top lateritic soil in elekoyangan, sango area and oyun area are 106 ohm-m and 2m respectively which indicated that the predominant composition of the top soil is lateritic clay. in apata yakuba, ganiki sango and royal valley, the average resistivity values and thickness of lateritic layer are 250ohm-m and 4m which implies that the layers composed of clayey sand. also, the average resistivity and thickness values of top lateritic soil in alagbado, okelele and olorunshogo areas are 107 ohm-m and 2m respectively which indicates that the predominant composition of the top soil is lateritic clay. the average resistivity and thickness values of top lateritic soil in agbabiaka area, kilanko area and tanke area are 107 ohm-m and 2m respectively which is an indication that the predominant composition of the top soil is lateritic clay. figure 3 shows map of the spatial distribution pattern of lateritic clay in the study area as it varies from one location to another. i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 51 vol. 7: 43-57, december 2016 fig 3. distribution pattern of lateritic clay depth (m) in ilorin 3.2.2 weathered basement in apata yakuba, ganiki sango and royal valley, the weathered basement average resistivity and thickness are 105ohm-m and 18m which indicates a saturation characterized by a moderately low resistivity layer. the weathered basement average resistivity and thickness are 103ohm-m and 18m in elekoyangan, sango area and oyun area, suggesting some level of saturation characterized a moderately lower resistivity layer. in alagbado, okelele and olorunshogo, the weathered basement average resistivity and thickness are 51ohm-m and 22m, respectively. this agrees with the findings of ogunlana and talabi (2014), indicating that the material composition is largely clay, sandy clay and clayey sand. this is evidenced in the high degree of water logging of this area, particularly in dry season. in agbabiaka, kilanko and tanke area of the city. the weathered basement average resistivity and thickness are 135ohm-m and 30m respectively, which indicates existence of some degree of fractures and water saturation weathered basement. the weathered basement average resistivity and thickness for danialu and fate tanke are 158ohm-m and 15m respectively, indicating a saturation characterized by a moderately low resistivity layer. figure 4 shows map of the spatial distribution pattern of weathered basement in the study area as it varies from one location to another. i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 52 vol. 7: 43-57, december 2016 fig 4. distribution pattern of weathered basement depth (m) in ilorin 3.2.3 fairly hard basement the fairly hard basement in elekoyangan, sango area and oyun area, average and thickness resistivity are 167ohm-m and 30m. this falls within the weathered and fresh basement rock which is also characterized by potential aquiferous units and hard rock in the areas. also, its resistivity and thickness values are 339ohm-m and 30m, the result agree with oyedele and olayinka (2012) that the groundwater potential of the aquifer may be significantly enhanced if the geo-electric basement has a fairly low resistivity. relatively low values of geo-electric basement resistivity (200 – 640 ωm) are indicative of good groundwater potential. the fairly low bedrock resistivity confirms the presence of fractures and hence water contained within the fissures (becson and jones, 1988; olayinka and olorunfemi, 1992; ayodele and olayinka, 2012). figure 5 shows map of the spatial distribution pattern of fairly hard basement in the study area as it varies from one location to another. i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 53 vol. 7: 43-57, december 2016 fig 5. distribution pattern of fairly weathered basement depth (m) in ilorin 3.2.4 thin fractured in alagbado, there is a presence of thin fracture zone which indicates a good potential aquiferous units in the area. fate tanke area also show thin fractured zone in the study area which served as reservoir in the area. oloje area have thin fractured zone. the resistivity and the thickness values are 60ohm-m and 35m which indicates that the area are water bearing zone due to its low resistivity values. the coloured part of figure 6, shows map of the spatial distribution pattern of thin fractured in ilorin, while the plain area are the locations without thin fracture zone in the study area. i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 54 vol. 7: 43-57, december 2016 fig 6. distribution pattern of thin fractured depth (m) in ilorin 3.2.5 hard basement the hard basement in alagbado, okelele and olorunshogo, also varies and the average resistivity and thickness values are 100ohm-m and 38m which indicate fine grained with intercalation of sandy clay. fresh hard rock is the last layer in apata yakuba, ganiki sango and royal valley, the section is relatively deep in the area and average resistivity and thickness values are 307ohm-m and 35m, the resistivity values are somehow high because of its crystalline nature. the hard basement in agbabiaka area, kilanko area and tanke area also varies and the average resistivity and thickness values are 240ohm-m and 41m which indicate fine grained with intercalation of sandy clay. the hard rock average resistivity and thickness values for danialu, fate tanke and wonderland chapel area, are 439ohm-m and 30m. the result agrees with oyedele and olayinka (2012), that the groundwater potential of the aquifer may be significantly enhanced if the geo-electric basement has a fairly low resistivity. relatively low values of geo-electric basement resistivity are indicative of good groundwater potential and the fairly low i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 55 vol. 7: 43-57, december 2016 bedrock resistivity confirms the presence of fractures, which shows that water is contained within the fissures (becson and jones, 1988, and olayinka and olorunfemi, 1992; ayodele and olayinka, 2012). figure 7 shows the spatial distribution pattern of hard basement in the study area as it varies from one location to another. fig 7. spatial distribution pattern of hard basement depth (m) in ilorin 4 conclusion in conclusion, understanding the spatial distribution of geo-electric succession, especially in a large area is a key to maximizing groundwater prospecting. this study has revealed four to six geo-electric successions in ilorin metropolis and took a step further to present a spatial distribution pattern of the identified layers in the study area. as revealed by the study, from the third layer to the sixth layer shows good groundwater potential. in addition, the resistivity values obtained shows that there are economically viable groundwater resources in the study area, especially in the weathered basement, fairly hard basement, thin fractured and hard basement lithology. although, the level of groundwater availability varies from one location to the i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 56 vol. 7: 43-57, december 2016 other as the depth and resistivity of these layers varies from place to place. the spatial distribution pattern of the geo-electric succession presented in this study would aid driller in siting boreholes in ilorin metropolis, without necessarily conducting geophysical survey, which will in turn reduce the cost of drilling borehole in the study area. acknowledgements the authors appreciate the managements of kwara state ministry of water resources and lower niger river basin and rural development authority, ilorin for providing us the data for this work. references abiola o, ogunribido tht, omoniyi ba, ikue po. 2013. geo-electric assessment of groundwater prospects in supare estate, supare akoko, south western, nigeria. journal of geosciences 3(1):23-33. adiat kan, olayanju, gm., omosuyi, go., ako bd. 2009. electromagnetic profiling and electrical resistivity soundings in groundwater investigation of a typical basement complex-a case study of oda town southwestern nigeria. ozean journal of applied sciences, 2(4): 333-359. amaresh, kr. s. and ravi prakash, s. 2003. an integrated approach of remote sensing, geophysics and gis to evaluation of groundwater potentiality of ojhala subwatershed, mirzapur district, u.p., india. water resources, map india conference 2003. anomoharan, o. 2011. determination of groundwater potential in asaba, nigeria using surface geo-electric sounding, international journal of physical sciences, 6: 7651-7656. anomoharan, o. 2013. geophysical investigation of groundwater potential in ukelegbe, nigeria. journal of applied sciences, 13(1): 119-125. bose kn., chatterjee d., and sen ak. 1973. electrical resistivity surveys for groundwater in the aurangabad sub-division, gaya district, bihar, indian pp. 171-181. ifabiyi, ip. and ahmed, ya. 2011. determination of household water demand in a traditional city: examples from western axis of ilorin nigeria. asianafrican journal of econometrics and economics, 11(2):395-408. ifabiyi, ip. and ashaolu, ed. 2013. analysis of the impacts of rainfall variability on public water supply in ilorin, nigeria. journal of meteorology and climate science, 11(1):18-26. krishnamurthy, j., venkataesa kumar, n., jayraman, v. and manivel, m. 1996. an approach to demarcate groundwater potential zones through remote sensing and gis. international journal of remote sensing, 17(10):1867-1884. murthy, ksr. 2000. groundwater potential in a semi-arid region of andhra pradesh: a geographical information system approach, international journal of remote sensing, 21(9):1867-1884. ogundana, ak. and talabi, ao. 2014. geo-electric characterization of aquiferous units and its implication on groundwater potential of owo, southwestern nigeria. american journal of water resources, 2 (2): 37-40. oladapo, mi., adeoye-oladapo, oo. and mogaji, ka. 2009. hydrogeophysical study of the groundwater potential of ilara-mokin southwestern nigeria. global journal of pure and applied sciences, 15(2):195-204. olayinka ai, and olorunfemi mo. 1992. determination of geo-electric characteristics in okene area and implications for borehole siting. journal of mining and geology. 28(2):403-411. olorunfemi mo. 1990. the hydrogeological implication of topographic variation with overburden thickness in basement complex area of sw nigeria. journal and mining and geology. 26 (1) i. p. ifabiyi et al. geo-electric characteristics of basement complex rocks ruhuna journal of science 57 vol. 7: 43-57, december 2016 olorunfemi mo. and fasuyi sa. 1993. aqufier types and geo-electric/hydrogeologic characteristics of part of central basement terrain of nigeria (niger state). journal of africa earth sciences, 16 (3): 309-317. omosuyi, go., ojo, js. and enikanselu, pa. 2003. geophysical investigation for groundwater around obanla – obakekere in akure area within the basement complex of south-western nigeria. journal of mining and geology. 39(2):109-116. otobo, e. and ifedili, so. 2005. the vertical electrical sounding: a viable tool for the investigation of fresh groundwater in the saline water environment with particular references to the community along warri river, journal of nigerian association of mathematics and physics, 5: 437-442. oyedele, eaa. and olayinka, ai. 2012. statistical evaluation of groundwater potential of adoekiti, south western nigeria. translational journal of science and technology, 12(6):110127. oyedele, eaa., oyedele, t. and oyedele, k. 2013. geo-electrical data analysis to demarcate groundwater pockets in ado-ekiti, southwest, nigeria. international journal of water resources and environmental engineering, 5(11): 609-615. saraf, ak. and chaudhary, pr. 1998. integrated remote sensing and gis for groundwater exploration and identification of artificial recharges sites, international journal of remote sensing, 19(10):1825-1841. shahid, s. and nath, sk. 2002. gis integration of remote sensing and electrical sounding data for hydrogeological exploration, journal of spatial hydrology, 2(1): 1-12. rjs-2007-karunawardena-wijayasiri.dvi ruhuna journal of science vol. 1, september 2007, pp. 41–47 http://www.ruh.ac.lk/rjs/ i s s n 1800-279x © 2007 faculty of science university of ruhuna. slowly rotating dust sphere for free space in general relativity with uniform matter distribution k.m.e.m. karunawardana and m.p.a. wijayasiri department of mathematics, university of ruhuna, matara, sri lanka erandi@maths.ruh.ac.lk, wijaya@maths.ruh.ac.lk abstract. einstein field equations for a charged dusty universe have already investigated. in this paper we present a new class of analytical solutions in terms of canonical coordinates for einstein’s field equations; assuming that the spacetime is spherically symmetric, formed by non-charged dust with uniform matter distribution. the metric we considered is of the form, ds2 = e2νdt 2 − e2λdr2 − r2dθ2 − r2 sin2 θ(dφ − ωdt)2 , where ν, λ and ω are functions of the radial coordinate r only. our model has only a space singularity at r = 0 and the solutions are well behaved for r > 0. in addition, we assume that the proper density ρ is constant. ω(r), the angular velocity of the inertial frame; can be an arbitrary function of r, which satisfies required boundary conditions to be a slow rotation. key words : spherical symmetry, slowly rotation, free-space 1. introduction the general relativity theory still dwells at the highest place among the gravitational theories; but the physical meaning of many exact solutions of einstein’s equations are unknown, or only partially understood bonnor (1992). one of the major difficulties is the, either long time taken to find a experimental verifications to prove the solutions have real physical meaning. however general relativity is central to the understanding of frontier astrophysical phenomena such as black holes, pulsars, the big bang and the universe itself milner (2000), ltartle (2003). it is interesting to note that every object in the space exhibits some form of rotation. for further details see, bayin (1981). therefore, during the last decades, rotating objects have been studied quite extensively tiwari et al. (1986), bayin (1981). in this paper our aim is to study the case of slowly rotating dust sphere in free space. we consider two types of rotations in our equation, ω(r), which represents the dragging of inertial frames and ω, which represents the angular velocity of dust distribution, along the coordinate axis φ, i.e., ω = dφ/dt. in the approximation of slow rotation (1) leads to a system of equation in the first order terms, and then new analytical solutions for λ, ν, ω, and ω are found subjected to the assumptions ρ = constant and time independent. the boundary conditions for ω are lim r→∞ ω(r) → 0, lim r→∞ (∂ω(r)/∂r) → 0 and lim r→0 ω(r) → 0, lim r→0 (∂ω(r)/∂r) → 0. the same boundary conditions should be satisfied by ω. we take the cosmological constant λ to be equal to zero for simplicity and consider a static metric. 41 karunawardana and wijayasiri: slowly rotating dust sphere ... 42 ruhuna journal of science 1, pp. 41–47, (2007) 2. field equations we consider the static, spherically symmetric metric in the standard form chandrasekher (1910) ds2 = gi jdx idx j; i, j = 0, 1, 2, 3 ds2 = e2νdt 2 − e2λdr2 − r2dθ2 (1) − r2 sin2 θ(dφ − ωdt)2; where ν, λ and ω are functions of the radial coordinate r only. the coordinates x0, x1, x2 and x3 correspond to t, r, θ and φ respectively. the function ω represents the angular velocity of the inertial frames along the rotation axis. the einstein field equations are, chandrasekher (1910), tiwari et al. (1986), gi j = ri j − 1 2 rgi j = 8πti j. and ricci tensor ri j is defined by tiwari et al. (1986) , ri j = g kl rik jl . (2) the ricci tensor components in tetrad form for slowly rotating sphere, related with the metric (1) are as follows: r(00) = −e −2λ(νrr − λr νr + ν2r + 2νr r ), (3) r(11) = e −2λ(νrr − λr νr + ν2r − 2 r λr), r(22) = ( νr 2 − λr r + 1 r2 )e−2λ − 1 r2 , r(03) = − 1 2 sin θe−(2λ+ν)[ωrrr + 4ωr − ωr r(λr + νr)]; where, ωr and ωr denote differentiation with respect to r. the energy momentum tensor t(i j) of dust is defined by tiwari et al. (1986), t(i j) = ρu(i)u( j); where ρ is the proper density of the universe, and u i are the four velocity components of the matter distribution along the each coordinate axis. the components u i are given by, u 0(r) = dx0 ds , u 1(r) = u 2(r) = 0, u 3(r) = dφ dx0 dx0 ds = ωu 0. karunawardana and wijayasiri: slowly rotating dust sphere ... ruhuna journal of science 1, pp. 41–47, (2007) 43 we use the tetrad frame ei(α) which is associated with the metric (1), eα(0) = (e −ν, 0, 0, 0), eα(1) = (0, e −λ, 0, 0), eα(2) = (0, 0, 1 r , 0), eα(3) = (0, 0, 0, 1 r sin θ). ei(α)e(β)i = η(αβ) =     1 0 0 0 0 −1 0 0 0 0 −1 0 0 0 0 −1     and ei(α)e (α) j = δ i j (an index within parentheses denotes a tetrad index). for the case of slowly rotation, tetrad components for four velocity are obtained as follows, u(0) = 1, u(1) = 0, u(2) = 0, u(3) = re −ν sin θ(ω − ω). the field equations for dust universe in the tetrad frame can be expressed in the form r(i j) = −8π(t(i j) − 1 2 g(i j)t ). (4) the tetrad components for energy momentum tensor are as follows: t(00) = ρ (5) t(01) = t(02) = t(11) = t(12) = t(22) = t(13) = t(33) = 0, t(03) = ρe−νr sin θ(ω − ω). from (4)and (5) the ricci tensor components can be obtained as follows: r(01) = r(02) = r(12) = r(13) = r(23) = 0, (6) r(00) = r(11) = r(22) = r(33) = −4πρ, r(03) = −8πρr sin θ(ω − ω). from (3)and (6) the following nonlinear system of equations can be obtained, 4πρ = e−2λ(νrr − λrνr + ν2r + 2νr r ), (7) −4πρ = e−2λ(νrr − λrνr + ν2r − 2 r λr), −4πρ = ( νr 2 − λr r + 1 r2 )e−2λ − 1 r2 , 8πρ sin θ(ω − ω) = 1 2 sin θe−(2λ+ν)(ωrrr + 4ωr − ωrr(λr + νr)). karunawardana and wijayasiri: slowly rotating dust sphere ... 44 ruhuna journal of science 1, pp. 41–47, (2007) 3. the solution all the solutions are obtained for ρ = constant. the solutions for e2λ and e2ν can be obtained by solving the first three of system of equations (7). we have, e2λ = ( 3 6 + 8πρr2 ) e2ν = 4 ( (12 + 2kr2)5/4 r3/2 ) ; k = 8πρ. (8) we have to assume a value for ω which is a function of r only that satisfies the given boundary conditions. therefore, ω is arbitrary and in this paper we choose functions for ω as a function of r only. let us take, ω(1) = r3 (2 + r2)4 + r(4r2 + r3) (r2 + 1)5 . then ω(r) can be found as follows, ω = ω − e−(2λ+ν) 2k [ωrrr + 4ωr − ωrr(λr + νr)] . (9) it is clear that both ω(r) and ω(r) satisfy the boundary conditions as r → 0 and r → ∞. 1 2 3 4 5 r 0.05 0.1 0.15 0.2 whrl figure 1 variation of dragging velocity of inertial frames in spacetime with increasing r (r, the distance from origin of the universe to the imaginated position); the dragging inertial frames have a maximal velocity in 0 < r < 1 (r, measured in light years) fig. 1 shows the ω has a maximum value 0 6 r < 1 , meaning that the inertial frames drag with a maximum velocity in 0 6 r < 1. karunawardana and wijayasiri: slowly rotating dust sphere ... ruhuna journal of science 1, pp. 41–47, (2007) 45 1 2 3 4 5 r 0.05 0.1 0.15 0.2 ωhrl figure 2 the variation of angular velocity of coordinate axis φ, with increasing r (r, distance from origin of the universe to the imaginary point) we choose a different ω as follows, ω(2) = r3 (2r5 + 4)2 sin ( 4r2 (r2 + 5)2 ) . 1 2 3 4 5 r 0.001 0.002 0.003 0.004 0.005 0.006 0.007 whrl figure 3 variation of dragging velocity of inertial frames in spacetime with increasing r, shows one maximum rotation 0 < r < 1, value of it less than to ω1 shape of the graph similar to that of fig. 1, but the maximum value for rotations in a different range and the value is completely different from that is given in ω1. using (9) we can find new ω for ω2 . 1 2 3 4 5 r 0.001 0.002 0.003 0.004 0.005 0.006 0.007 ωhrl figure 4 the variation of angular velocity of axis φ, with increasing r, shows maximum rotation 1 < r < 2 karunawardana and wijayasiri: slowly rotating dust sphere ... 46 ruhuna journal of science 1, pp. 41–47, (2007) by substituting value e2λ, e2ν and ω in (1) we can obtain two different models for the universe as follows: ds2 = 4 ( (12 + 2kr2)5/4 r3/2 ) dt 2 − ( 3 6 + 8πρr2 ) dr2 − r2dθ2 − r2 sin2 θ ( dφ − ( r3 (2 + r2)4 + r(4r2 + r3) (r2 + 1)5 ) dt 2 )2 . ds2 = 4 ( (12 + 2kr2)5/4 r3/2 ) dt 2 − ( 3 6 + 8πρr2 ) dr2 − r2dθ2 − r2 sin2 θ ( dφ − r3 (2r5 + 4)2 sin ( 4r2 (r2 + 5)2 ) dt 2 )2 . 4. discussion and conclusion in our work, we like to obtain a family of spherical symmetric cosmological models for a non-stationary rotating dust distribution. it is a fact that the motions of the planets were investigated theoretically to a greater degree of accuracy, the theory of the motion of a rigid body was developed and applied to the problem of the rotation of the earth. however, applications of general relativity are few indeed, a state of affairs may perhaps be accounted for by its extreme mathematical complexity; investigators have found specific problems mathematically difficult and have soon turned away in discouragement. we developed our metric for the particular case when ρ = constant, for uniform matter distribution. to satisfy the slow rotation condition we consider terms of the first order in ω. we make the following observations. 1. the model has a space singularity at r = 0 2. infinite number of solutions can be found for ω and ω. considering two expressions for ω the dragging velocity of the inertial frames, we find that it maximixed at at least one time 0 6 r < ∞. the angular velocity of dust distribution along the axis φ, also has one minimum value in 0 6 r < ∞. although, the solutions are satisfy the boundary conditions. all the solutions shows maximum and minimum rotations which make us wonder because according to our assumptions no external force has been exist exception the gravity, and it is rather difficult to give high accuracy physical explanations to solutions. to sum up, solutions for einstein field equations are not unique and linear, it is not easy to give physical explanations. 5. acknowledgments we are thanking to dr.j.r.wedegadera making valuable comments and suggestions. references bayin, selcuk s. 1981. slowly rotating fluid fluid spheres in general relativity with and without radiation. phys. rev.d. 24 2056 – 2065. bonnor, wb. 1992. physical interpretation of vacum solutions of einstein’s equations. par i. timeindependent solutions. general relativity and gravitation 24. karunawardana and wijayasiri: slowly rotating dust sphere ... ruhuna journal of science 1, pp. 41–47, (2007) 47 chandrasekher, s. 1910. the mathematical theory of blackholes. oxfered univ. press, ny. ltartle, james b. 2003. gravity an introduction to einstein’s general relativity. pearson education pte, ltd. indian branch. milner, brayan. 2000. cosmology. cambridge univ. press,cambridge. tiwari, rn., jr. rao, , rr. kanakamedala. 1986. slowly rotating charged fluid spheres in general relativity. phys. rev.d. 34 327 – 330. sv-lncs ruhuna journal of science vol 9(2): 127-139, december 2018 eissn: 2536-8400  faculty of science http://doi.org/10.4038/rjs.v9i2.40 university of ruhuna  faculty of science, university of ruhuna 127 rainfall trends and variability over onitsha, nigeria a.j. oloruntade*, k.o. mogaji and o.b. imoukhuede department of agricultural and bio-environmental engineering technology, rufus giwa polytechnic, owo, ondo state, nigeria *correspondence: johntades1@yahoo.com; orcid: 0000-0001-6156-6823 received: 27th march 2018, revised: 15th october 2018, accepted: 25th october 2018 abstract. analysis of trend and variability in rainfall can provide the necessary information required for water resources planning and management in any geographical setting. therefore, the present study applied standard tests to investigate rainfall trends and variability using monthly, seasonal and annual series over onitsha, nigeria between 1971 and 2008. rainfall variability index revealed the year 1983 as the driest (-2.38) and 1997, the wettest (+2.0), with more dry years observed between 2000 and 2008. variability was relatively low annually as compared to seasonal and monthly series. september (15.8%) has the highest contributions to total annual rainfall, while january contributed the least (0.6%). seasonally, about 40% of the annual rainfall was received in june-july-august (jja), while the lowest rainfall was during december-january-february (djf) (3.75%). trends were mostly insignificant on monthly basis with 5 of the 12 months exhibiting negative trends, while only january depicted positive significant trend (p< 0.05). similarly, only jja exhibited insignificant upward trend while other seasons showed downward trends that are also not significant. on the annual basis, an insignificant negative trend was observed for the period under study. hence, farmers and water resources managers may need to develop appropriate management strategies which include construction of more water storage and diversion structures such as reservoirs and dams to combat recurrent flooding during summer seasons and potential future water scarcity in the area. keywords. onitsha, rainfall, trend analysis, variability index 1 introduction information regarding rainfall trend and variability is an important requirement for the planning and management of water resources. irrigation planning and management is an area of water resources engineering that a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 128 vol 9(2): 127-139, december 2018 requires adequate knowledge of rainfall variation and its temporal pattern. given that recent changes in climate have intensified global variability of the hydrological cycle, creating uncertainties regarding the prediction of future climatic conditions and the associated impacts, studies of long-term climate series have become increasingly necessary (houghton et al. 1996). moreover, the planning, design and operation of most water storage reservoirs are based on the historical pattern of water availability, quality and demand, with the assumption of normal climatic behaviour which can no longer hold under a changing climate (abdul aziz and burn 2006). nevertheless, oguntunde et al. (2011) reported that understanding the behaviour of rainfall as a major component of the hydrological cycle, may be of profound social and economic significance. in this regard, hydrologists and water resources engineers need up-to-date knowledge of the behaviour of rainfall through statistical analysis. besides, reported cases of extreme rainfall events such as flood across the country in recent years call for a clear understanding of the variability and trends of rainfall. this is even more important considering the huge socioeconomic losses that accompany flood events. for instance, during the flood disaster of the year 2012 alone, a newspaper report showed that properties valued at over 40 billion naira (219.6 million usd) were lost in about nine local government areas of kogi state including lokoja, while the inhabitants of the 332 communities affected by the floods were also put at a further risk of diseases outbreak as a result of their exposure to contaminated water and water-borne pathogens (‘the punch’ 2012). also, for a country where agriculture plays a major role in employment and food availability, meeting rising future demands for food and potable water, requires a more judicious use of water in both irrigated and rain-fed agriculture (smith 2000). in addition, sustainability of food production will be contingent upon the effective allocation of water resources, given that fresh water for human consumption and agriculture is gradually becoming scarce by the day. several studies that have been carried out on rainfall time series across the world show both negative and positive trends. nevertheless, analysis of trend is a leading step towards attributing changes in climate to such factors as climate variability, greenhouse gases (ghg) and changes in the built-up environment (blake et al. 2011). many studies have examined the monotonic trend of rainfall and its abrupt changes at different locations across the world (e.g. xu et al. 2010, wang et al. 2011, anghileri et al. 2014). moreover, analysis of time series at different time scales around the world have revealed that rainfall is either decreasing or increasing, depending on the location (mondal et al. 2012). shahid (2010) examined the trends of annual and seasonal rainfall of bangladesh during 1958–2007 period and reported significant increases in the average annual and pre-monsoon rainfall. it was found that while the number of wet months increased, dry months decreased over the greater parts of the country. jones et al. (2015) analysed variability of a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 129 vol 9(2): 127-139, december 2018 precipitation in the upper tennessee river basin and observed statistically significant increasing or decreasing trends in 11% of the 78 sub-basins during the 50-year period covered. although, there were variations in trends of monthly precipitation volumes, significant increases were most pronounced in the summer and autumn. in europe, an overall insignificant decreasing trend was recorded over the lower altitudes stations in the agricultural zone of pieria and aison river basin, greece, while a decrease of precipitation was detected in spring on the basis of station and regional analysis (karpouzos et al. 2010). however, a few significant positive trends were detected only in remote stations during march, april and october in a study conducted in turkey by yavuz and erdogan (2012) using data from more than 100 gauging stations spanning through the period 1975–2009. studies on variability and trend in rainfall at both spatial and temporal scales are not many over africa and particularly in the west africa subregion. across nigeria, different trends have been reported by numerous scientists (e.g. adefolalu 2007, abaje et al. 2010, akinsanola and ogunjobi 2014), subject to the agro-climatic zone, while some of the studies have also testified to the occurrence of critical climatic events in forms of flood and drought, which are compatible to the prevailing changes in climate. for instance, abaje et al. (2010) in a study over kafancha in the guinea savanna ecological belt of nigeria found significantly drier conditions in the months of june and october for the sub-periods 1974-1983 and 1999-2008, respectively. it was further revealed that the reduction in annual rainfall was predominantly as a result of the substantial decline in july, september and october rainfall, which are the critical months for agricultural production in the area. oguntunde et al. (2011) analysed rainfall trends over nigeria (1900-2000) and reported that about 90% of the entire landscape of the country exhibited negative rainfall trends but only 22% showed significant changes at 5% level. furthermore, akinsanola and ogunjobi (2014) studied rainfall and temperature variability over nigeria and observed alternately decreasing and increasing trends in mean annual precipitation. however, in view of the geographical diversity of nigeria, trend analyses that are specific to any subregional setting in the country is a worthwhile scientific undertaking. meanwhile, considering that rain-fed agriculture is important for food production and employment in west africa, planning of water resources and agricultural projects usually depends on long-term records of many rainfall variables (tarhule and woo 1998). hence, the objective of the present study is to examine trend and variability of rainfall over onitsha, nigeria, with a view of suggesting ways for flood mitigation and agricultural water management. a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 130 vol 9(2): 127-139, december 2018 2 materials and methods 2.1 the study area onitsha is a city in south-eastern nigeria located on latitude 6°10′ n and longitude 6°47′ e (figure 1), with a population of about 561,000 (national population commission, 2006), spread over a land mass of about 36.19 km2. it falls within the rain forest belt and it is the last gauging station along the niger river basin before its final exit to the atlantic ocean. the city of onitsha enjoys the typical tropical climate of nigeria that is usually influenced by three main wind currents the tropical maritime (mt) air mass, the tropical continental (ct) air mass and the equatorial easterlies (ojo, 1977). rainfall commences around march/april, reaching its peak during july to september and finally stopping in november/december. mean annual rainfall ranged between 1200 mm and 2000 mm with a maximum temperature of 34.7 oc and a relative humidity of 65%. as the commercial nerve center of the eastern part of the country, both natural landscape and hydrological processes of the city have been severely altered by human activities, which include urbanization, river dredging and road constructions. in recent times, destructive floods of high magnitudes have been recorded in the area which many have attributed to climate and possibly, changes in land use. apart from the usual buying and selling activities in the city, fishery, agriculture with the cultivation of cassava, maize and other cereal crops form the main livelihoods of the people. fig. 1. map of nigeria showing onitsha in anambra state a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 131 vol 9(2): 127-139, december 2018 2.2 dataset this study used rainfall data obtained from the new high resolution global gridded rainfall data set (cru ts 2.1) provided by the climate impacts link project of climate research unit (cru), university of east anglia, uk (mitchell and jones, 2005). the cru dataset comprises of monthly 0.5o latitude/longitude gridded series of climatic parameters over the periods 19012008. new et al. (2000) and conway et al. (2009) provide in-depth information on the quality control and interpretation of cru data, given the general poor spatial and temporal coverage of meteorological stations in africa. moreover, the cru dataset is preferred for the present study because it has been used in many previous studies over nigeria with reliable results (e.g. oguntunde et al. 2011, 2012, oloruntade et al. 2017). kahya and kalayci (2004) posited that for a reliable examination of climatic trend, a period of 30 years is usually taken as adequate to arrive at an acceptable conclusion. hence, the use of 38 year data (1971-2008) is assumed adequate for the present study. 2.3 data analysis exploratory data analysis and descriptive statistics the use of exploratory data analysis (eda) helps to gain an understanding of the direction and mode of change in hydro-climatic variables. apart from the well-known mathematical techniques, it is commonly included in extensive trend detection (anghileri et al. 2014). eda refers to any technique of data analysis besides formal statistical methods. it employs graphical tools such as time series and scatter plots, and it is intended to ensure improved understanding of the existing data and the fundamental processes involved in its changes. in addition, simple descriptive statistics which include the mean, standard deviation (sd) and coefficient of variation (cv) are also calculated to obtain an initial understanding of the data. rainfall variability index rainfall index which is computed as the standardized departure of precipitation, helps to categorize the rainfall time series into various climatic periods such as very dry year, normal year and wet or very wet years. it was computed as, (1) where is rainfall variability index for year , is annual rainfall for year and are the mean annual rainfall and the standard deviation for the period between 1948 and 2008. moreover, following the world meteorological a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 132 vol 9(2): 127-139, december 2018 organization (wmo 1975), rainfall time series can be classified into various climatic regimes (table 1). table 1. classification of climate regime based on variability index (wmo 1975) performance rainfall regime extremely dry dry normal wet non-parametric trend test the rank-based nonparametric mann–kendall (m-k) test has been widely used to assess the significance of monotonic trends in hydro-meteorological time series (e.g. gan 1998, kumar et al. 2010, anghileri et al. 2014). it has been suggested that, for the assessment of trends in climatic data, the m-k test should be applied (wmo 1988). it is commonly chosen above other methods in view of its skill in dealing with non-normally distributed data, outliers and missing values in series and its high asymptotic efficiency (gan 1998). to estimate the true slope of an existing trend, the sen’s non-parametric technique, well-adjudged for its skillfulness (zhao et al. 2008) as proposed by sen (1968) was used and more details of the method are given in xu et al. (2007). trend free pre-whitening application of mann-kendall (m–k) test requires that the time series does not contain any serial correlation. significant positive serial correlation influences the power of m–k and thereby leads to major source of uncertainty. to eliminate or minimize this effect, pre-whitening of the original dataset before using the m–k test is recommended (abdul aziz and burn 2006). the procedure is adequately outlined in kumar et al. (2010) and many other authors. the m–k test was thereafter applied to identify trend in the final (or pre-whitened) series. 3 results and discussion 3.1 summary of eda and descriptive statistics annual and seasonal time series plots depict the pattern of rainfall over onitsha for the entire period of study (figure 2). inspection of the plots a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 133 vol 9(2): 127-139, december 2018 revealed a general decline in the annual rainfall over the study period. this indicates a possible drying condition which may cause a reduction in available water in the study area for domestic, industrial and agricultural water uses, most especially irrigated and rain-fed agriculture. fig.2. annual and seasonal time series of rainfall over onitsha, nigeria muhire et al. (2014) suggested that a decline in rainfall leads to a shortage of water for agriculture and, therefore, reduction in crop production. however, to confirm the present level of water availability in the area, additional study which may require further analysis of runoff/ discharge is needed. on the positive side, reduction in annual rainfall can lead to fewer occurrences of flood events. similarly, when rainfall is considered by seasons, decemberjanuary-february (djf), march-april-may (mam), june-july-august (jja) and september-october-november (son) – apart from jja which displayed increasing/positive trend, decreasing rainfall trends have been observed during the remaining seasons. similar to the present result, salami et al. (2014) had earlier analysed trends in hydro-meteorological variables covering about six stations in nigeria and found a significant reduction in rainfall and runoff in five stations. consequently, the study concluded that the reduction may have led to decreasing water resources availability. nevertheless, increasing summer rainfall exacerbates flooding which can lead to loss of lives and properties, including farmlands (oloruntade et al. 2017). however, the decrease in rainfall was most obvious in annual series followed by mam, a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 134 vol 9(2): 127-139, december 2018 son and djf, respectively. meanwhile, reduction of rainfall during the spring season (mam) may imply delay in rainfall onset which can also affect the resumption of agricultural activities and shortening of the growing season. a summary of the descriptive statistic of the long-term (temporal) series also showed that mean monthly rainfall varied from 329 mm/yr (september) to 12 mm/yr (january), seasonal mean ranged from 904 mm/yr (jja) to 78 mm/yr (djf), while the annual mean was 2081 mm/yr (table 2). likewise, standard deviation (sd) ranged between 117 mm/yr (august) and 18 mm/yr (january), it also hovered between 184 mm/yr (jja) and 61 mm/yr (djf), whereas the sd for the annual series was 292 mm/yr. with respect to the monthly coefficient of variation (cv), the highest was 147 % (january) while the lowest was 22% (june), seasonal cv ranged from 79% (djf) to 20% (jja), while the annual cv was 14%. the result showed that rainfall has been more varied during the month of january and in the summer (djf) season. in addition, rainfall variability was higher both on monthly and seasonal bases than inter-annually over onitsha. this has implications for rain-fed agriculture in the area, as farmers may have to evolve better water resources management plans to ensure sustainable agricultural production. meanwhile, high rainfall variability will alter the quantity, quality and supply of river water (wilby et al. 2006, whitehead et al. 2009), and can have implications for hydropower generation (hamududu and killingtveit 2016, oliveira et al. 2017). fig.3. percentage contribution of monthly rainfall to the annual series over onitsha, nigeria moreover, the highest monthly contributions to the annual total rainfall (figure 3) are in september (15.8%), followed by june (15.4%) and august (14.5%). this is contrary to the results of the study by oloruntade et al. (2017) over the niger-south basin using similar data (1948-2008), who reported a higher percentage of rainfall contribution from august rainfall as compared to september. however, the present result may be due to the possible delay in rainfall onset in the area which ultimately resulted in the forward shift of the summer rainfall. in addition, annual rainfall was dominated by the amount received between march and november, with about a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 135 vol 9(2): 127-139, december 2018 80% of the rainfall received during the period. surprisingly, the percentage contribution of february (2.07%) to the annual rainfall is higher than that of january (0.60%), the least amongst the months. on seasonal basis, jja had about 43.43%, mam, 22.64%, djf contributed the least (3.75%), while son recorded (30.18%). overall, jja seasonal intensity dominated the rainfall of the basin on the long-term basis. hence, it can be concluded that the flood events that have been recorded in the study area during september and october months may have been caused by the high rainfall intensity which usually persists from july to september. besides, high rainfall variability during the summer months may require additional efforts towards effective strategies for flood mitigation in the area. 3.2 rainfall variability (δ) annual rainfall variability indices covering 1971 to 2008 period are presented in figure 4. the result show three distinct periods which may be described for the station as: (1) from 1971 to 1982 (12 years) a seemingly random succession of 9 ‘‘normal’’ years, 2 “dry” years and 1 “wet” year, (2) from 1983 to 1998 (16 years), a series of 1 “extremely dry” year, 2 “dry” years, 10 ‘‘normal’’ years, and 3 “wet” years and (3) from 1999 to 2008 (10 years) of 3 “dry” years, 5 ‘‘normal’’ years and 2 “wet” years. the results are in agreement with the findings in earlier studies (e.g. nicholson et al. 2000, oguntunde et al. 2011). fig. 4. annual rainfall variability over onitsha, nigeria. moreover, following the classification of rainfall regimes by wmo (1975) as shown in table 1, only 1983 can be classified as “extremely dry” while 7 other years (18%) were “dry” among the 38 years. the “wet” years are 1980, 1990, 1995, 1997, 2003 and 2004 amounting to about 16% of the entire series. on the other hand, the remaining 24 years exhibited “normal” conditions which constituted about 63% of the whole series. the results showed that rainfall has been normal over onitsha for many years during the period under study and this may likely persist to the future. however, the occurrence of a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 136 vol 9(2): 127-139, december 2018 occasional wet years during which rainfall was above normal calls for adequate planning against summer flooding. 3.3 m-k trend analysis for the analysis of the trend using mann-kendall test, a spreadsheet (makesen 1.0) developed at the finnish meteorological institute (salmi et al. 2002) was used. the results showed insignificant negative trends in five of the 12 months in a year (table 2). table 2. descriptive statistics and trend test for monthly, seasonal and annual rainfall series over onitsha. rainfall series mean (mm/yr) sd (mm/yr) cv (%) test z (mm/yr) m-k (mm/yr) change (mm/yr) january 12.41 18.22 146.79 2.06 0.15 5.70 february 43.02 40.71 94.62 -0.82 -0.37 -13.88 march 80.57 44.99 55.84 -0.43 -0.20 -7.60 april 152.62 60.46 39.61 0.00 0.02 0.95 may 237.93 55.37 23.27 -0.98 -0.86 -32.51 june 281.88 62.03 22.00 0.60 0.58 21.92 july 320.17 90.78 28.35 0.38 0.44 16.83 august 301.52 117.05 38.82 -0.68 -1.70 -64.60 september 328.93 90.51 27.52 -0.63 -0.71 -26.93 october 236.73 75.04 31.70 0.25 0.25 9.61 november 62.31 62.89 100.93 1.38 0.61 23.05 december 22.63 30.74 135.88 0.82 0.08 3.17 djf 78.06 61.45 78.73 -0.03 -0.01 -0.38 mam 471.12 119.67 25.40 -0.55 -0.88 -33.25 jja 903.58 184.23 20.39 0.20 0.49 18.63 son 627.97 130.81 20.83 -0.52 -1.14 -43.46 annual 2080.73 291.61 14.01 -0.83 -4.00 -152.00 bold font is significant at p< 0.05 trends slope ranged from 0.61 mm/yr (november) to -1.70 mm/yr (august), but significant (p< 0.05) positive trend (slope= 0.15) was observed in january. the magnitude of change ranged between 23.05 mm/yr (november) and 64.60 mm/yr (august). generally, insignificant trends in either way show that rainfall on monthly basis has been relatively uniform over the period of study. on the seasonal basis, apart from jja which recorded an upward trend (slope= 0.49 mm/yr), trends in other seasons were downward; however, there was no significant trend in all cases. insignificant annual rainfall trends may be due to high inter-annual variability. the magnitude of change ranged between -18.63 mm/yr (jja) and -43.46 mm/yr (mam). on the annual basis, trend was insignificantly negative (slope = -0.83 mm/yr) and the magnitude of change was -152.00 mm/yr. this shows that over the study area, there has a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 137 vol 9(2): 127-139, december 2018 been reduction in the amount of annual rainfall received, indicating a drying condition. decreasing rainfall trends may lead to depletion in water availability for domestic, agricultural and industrial purposes. this might also cause reduction in the volume of water in surface water bodies such as streams, rivers and lakes in the area, thereby triggering increasing or overexploitation of groundwater resources. this condition may also hamper river navigation and hence impedes water transportation and conveyance of harvested fishes and other agricultural products. on the other hand, the insignificant upward trend in summer rainfall presents an opportunity for rainwater harvesting which can be utilized for small-scale irrigation farming during the winter season (goud et al. 2015). miao et al. (2012) in a similar study over the tropical climate of beijing, china observed an increasing summer rainfall. 4 conclusions the study investigated trends and variability in monthly, seasonal and annual rainfall time series during the period 1971 to 2008 over onitsha. the general result of this study indicated that it has been mostly drying over the whole landscape since the previous three decades of the last century and the situation may have continued in the 21st century. generally, increased rainfall in the summer months especially september has also been persistent with the likelihoods of floods and gully erosion in the area. in this regard, we suggest construction of additional water storage and diversion structures such as reservoirs and dams to mitigate recurrent flooding during summer seasons. meanwhile, given the projection of the intergovernmental panel on climate change (ipcc, 2007) for the present century, the decreasing trend observed in this study is attributable to both direct and indirect impacts of climate change. however, in view of the rapid urbanization rates in the study area, the intermittent floods being presently witnessed may have also been hastened by changes in land use/cover. nevertheless, the results of the present study could serve as a source of information needed by farmers and water resources managers for effective planning and as well provide the basis for in-depth climate change impacts studies on water resources in the area. acknowledgements the authors are grateful to dr. b. j. abiodun of environmental and geographical sciences department, university of cape town, south africa for providing the data used in this study. comments of anonymous reviewers on the initial manuscript are acknowledged. a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 138 vol 9(2): 127-139, december 2018 references abaje ib, ati of, iguisi eo. 2010. an analysis of rainfall trends in kafanchan, kaduna state, nigeria. research journal of environmental and earth sciences 2 (2): 89-96. abdul aziz oi, burn dh. 2006. trends and variability in the hydrological regime of the mackenzie river basin. journal of hydrology 319 (1): 282-294. adefolalu do. 2007. climate change and economic sustainability in nigeria. in: paper presented at the international conference on climate change and economic sustainability held at nnamdi azikiwe university, enugu, nigeria, 12–14 june, 2007. akinsanola aa, ogunjobi ko. 2014. analysis of rainfall and temperature variability over nigeria. global journal of human-social science (b) geography,geo-sciences, environmental disaster management 14 (3): 10-28. anghileri d, pianosi f, soncini-sessa r. 2014. trend detection in seasonal data: from hydrology to water resources. journal of hydrology 511: 171–179. blake r, grimm a, ichinose t, horton r, gaffin s, jiong s, bader d, cecil ld. 2011. urban climate: processes, trends, and projections. climate change and cities: first assessment report of the urban climate change research network. in: rosenzweig c, solecki wd, hammer s, mehrotra as. eds., cambridge university press, cambridge, uk, 43–81 pp. conway d, persechino a, ardoin-bardin s, hamandawana h, dieulin c, mahé g. 2009. rainfall and water resources variability in sub-saharan africa during the twentieth century. journal of hydrometeorology 10: 41-59. gan ty. 1998. hydroclimatic trends and possible climatic warming in the canadian prairies. water resources research 34 (11): 3009–3015. goud s, biradar v, janagoudar bs, shanwad uk, sreenivas ag. 2015. implications of rainfall distribution pattern and its variability for crop planning of crops: a study in north eastern transitional zone of karnataka. karnataka journal of agricultural sciences spl. issue 28 (5): 746-752. hamududu bh, killingtveit a. 2016. hydropower production in future climate scenarios: the case for the zambezi river. energies 9 (502): 1–18. houghton jt, meiro filho lg, callander ba, harris n, kattenburg a, maskell k. 1996. climate change 1995: the science of climate change. climatic change 584 pp. ipcc 2007. ipcc fourth assessment report: climate change 2007. geneva, switzerland. jones jr, schwartz js, ellis kn, hathaway jm, jawdy cm. 2015. temporal variability of precipitation in the upper tennessee valley. journal of hydrology: regional studies 3: 125-138. kahya e, kalayci, s. 2004. trend analysis of streamflow in turkey. journal of hydrology 289: 128–144. karpouzos dk, kavalieratou s, babajimopoulos c. 2010. trend analysis of precipitation data in pieria region (greece). european water 30: 31-40. kumar v, jain sk, singh y. 2010. analysis of long-term rainfall trends in india. hydrological sciences journal 55: 484–496. miao li, jun x, dejuan m. 2012. long-term trend analysis of seasonal precipitation for beijing, china. journal of resources and ecology 3 (1): 064-072. mitchell td, jones pd. 2005. an improved method of constructing a database of monthly climate observations and associated high-resolution grids. international journal of climatology 25: 693–712. mondal a, kundu s, mukhopadhyay a. 2012. rainfall trend analysis by mann-kendall test: a case study of north-eastern part of cuttack district, orrissa. international journal of geology, earth and environmental sciences 2 (1): 70-78. muhire i, ahmed f, abutaleb k. 2014. spatio-temporal trend analysis of projected temperature and precipitation data over rwanda. journal of meteorology and climate science 12 (1): 70-89. a.j. oloruntade et al. rainfall trends and variability over onitsha, nigeria ruhuna journal of science 139 vol 9(2): 127-139, december 2018 national population commission 2006. the nigeria population commission (npc) official result for 2006 population and housing census figures, bureau for national statistics, abuja. http://www.nigerianstat.gov.ng, accessed 23rd january, 2016. new m, hulme m, jones pd. 2000. representing twentieth century space-time climate variability. part ii: development of a 1961–90 mean monthly terrestrial climatology. journal of climate 13: 2217–2238. nicholson se, some b, kone b. 2000. an analysis of recent rainfall conditions in west africa, including the rainy seasons of the 1997 el nino and the 1998 la nina years. journal of climate 13 (14): 2628-2640. oguntunde pg, abiodun bj, lischeid g. 2011. rainfall trends in nigeria, 1901–2000. journal of hydrology 411: 207-218. oguntunde pg, abiodun bj, lischeid g. 2012. spatial and temporal temperature trends in nigeria, 1901–2000. meteorology and atmospheric physics 118 (1-2): 95-105. oliveira va, de mello cr, viola r, srinivasan r. 2017. assessment of climate change impacts on streamflow and hydropower potential in the headwater region of the grande river basin, southeastern brazil. international journal of climatology 37 (15): 5005-5023. oloruntade aj, mohammed ta, aimrun w. 2017. rainfall trends in the niger-south basin, nigeria. pertanika journal of science and technology 25 (2): 479 496. ojo o. 1977. the climates of west africa. heineman, london. salmi t, maatta a, anttila p, ruoho-airola t, amnell t. 2002. detecting trends of annual values of atmospheric pollutants by the mann-kendall test and sen’s slope estimates. publications on air quality, no. 31. helsinki, finland. sen pk. 1968. estimates of the regression coefficient based on kendall’s tau. journal american statistical association 63: 1379-1389. shahid s. 2010. rainfall variability and the trends of wet and dry periods in bangladesh. international journal of climatology 30 (15): 2299-2313. smith m, 2000. the application of climatic data for planning and management of sustainable rainfed and irrigated crop production. agriculture and forest meteorology 103: 99–108. tarhule a, woo mk. 1998. changes in rainfall characteristics in northern nigeria. international journal of climatology 18: 1261-1271. the punch 2012. kogi lost n40bn to floods. the punch newspaper, 28th september, 2012. wang d, hejazi m, cai x, valocchi aj. 2011. climate change impact on meteorological, agricultural, and hydrological drought in central illinois. water resources research 47: w09527. whitehead pg, wilby rl, battarbee rw, kernan m, wade aj. 2009. a review of the potential impacts of climate change on surface water quality. hydrological sciences journal 54 (1): 101-123. wilby rl, whitehead pg, wade aj, butterfield d, davis rj, watts g. 2006. integrated modelling of climate change impacts on water resources and quality in a lowland catchment: river kennet, uk. journal of hydrology 330 (1): 204-220. wmo 1975. drought and agriculture. wmo/ tn 138, geneva. wmo 1988. analyzing long time series of hydrological data with respect to climate variability. world meteorological organization: wcap-3, wmo/td-no: 224, switzerland, 1–12 pp. xu zx, li jy, liu cm. 2007. long-term trend analysis for major climate variables in the yellow river basin. hydrological processes 21: 1935–1948. xu z, liu z, fu g, chen y. 2010. trends of major hydroclimatic variables in the tarim river basin during the past 50 years. journal of arid environment 74: 256–267. yavuz h, erdogan s. 2012. spatial analysis of monthly and annual precipitation trends in turkey. water resources management 26: 609-621. zhao ff, xu zx, huang jx, li jy. 2008. monotonic trend and abrupt changes for major climate variables in the headwater catchment of the yellow river basin. hydrological processes 22 (23): 4587-4599.  © 2009 faculty of science university of ruhuna ruhuna journal of science vol. 4, september 2009, pp 13-20 http://www.ruh.ac.lk/rjs/ issn 1800-279x vajira p. bulugahapitiya1, syed ghulam musharaff2 1department of chemistry, faculty of science, university of ruhuna, matara, sri lanka 2 h.e.j research institute of chemistry, international centre for chemical sciences, university of karachi, pakistan abstrac.: microbial transformation is an effective tool for the structural modification of bioactive natural and synthetic compounds leading to synthesis of more potent derivatives. its application in asymmetric synthesis is increasing due to its versatility and ease. this article presents biotransformation of sesquiterpenoid ketone, (+)-nootkatone (1) by m. phaseolina, a plant pathogenic fungus. the transformation afforded four main compounds. they were determined to be 1:6 stereoisomeric mixture of 11,12-dihydroxy11,12-dihydronootkatone (2, 3), 13-hydroxynootkaone (4) and 12-hydroxy-11,12 dihydronootkatone (5) with the help of ei-ms, hr-fab-ms(pos), hr-fab-ms (neg), 1h-nmr, 13cnmr, cosy-450, noesy, hmbc, hmqc spectral analyses. the compound 4 was firstchandana-amarasingha-samayawardana-avifauna-bundala-1.1-28.07 identified as nootkatone metabolites in this study. further, the parental compound (1) and the transformed products 4 and 5 were found to be present significant antiprotozoal activity. key words : microbial transformation, nootkatone, m. phaseolina 1. introduction microbial transformation of secondary metabolites such as terpenoids, steroids and aromatic compounds from crude drugs and liverworts to obtain potent biologically active compounds such as drugs, pheromones and aromatics have been paid much interest since recently (choudhary et al., 2004: furusawa et al., 2005). microbial transformation involves the use of different microbes to perform chemical reactions in which the starting substances and product are of comparable chemic al complexity. enzymes in microbes are capable of performing a diverse range of reactions including the insertion of oxygen into c-h and c-c bonds, the addition of oxygen to alkenes, transfer of acetyl or sugar units from one substrate to another, the hydrolysis or formation of amides, epoxides, esters and nitriles. hydrogenation, hydrolysis and elimination of small units, epimerization and isomarization reactions and the formation of c-c, c-o, c-s and c-n bonds etc. (devkota et al., 2007). as enzymes are chiral biomolecules, these transformativajira13 microbial transformation of sesquitepenoid ketone, (+) nootkatone by macrophomia phaseolina http://www.ruh.ac.lk/rjs/rjs.html bulugahapitiya and musharaff ruhuna journal of science,3,pp53-60 (2008) microbial transformation of sesquitepenoid ketone... manuscript-rjs-reopened-nov-2011ons take place with high stereo and enantioselectivity. biological properties of the compounds mainly depend vajira-manuscript-rjs-reopenednov-2011on the specific configuration of the one or more chiral centers in such microbial transformations get much attention as support in structural modification of natural and synthetic compounds. moreover, it is capable of inserting functionalities into the inaccess ible site of the molecules in which cannot be reached by chemical syntheses and it is low cost and ease of handling. terpene hydrocarbons and their oxyfunctionalized derivatives, the terpenoids are the most diverse class of natural compounds whereas terpenoids are extensively applied in industry as fragrances and flavours, moreover those are important as chiral synthones for chemical synthesis (shaw, 1981). several terpenoids are easily available in large amount from plants or chemical syntheses. terpene skeleton is most favourable for structural modification by microbes. numerous publications described already mi crobial transformation of terpenoids (bock et al., 1988; bock et al., 2006; ishida, 2005; joglekar et al., 1969; welf-rainer-abraham et al., 1986; hieda et al., 1983). nootkatone (1) is a sesquiterpenoid kvajira-manuscript-rjs-reopened-nov2011etone and naturally available in grapefruit essential oil, and it has high demand in cosmetic and fibre industries. it has been reported that expensive aromatics, nootkatone (1) is capable of decreasing somatic fat ratio (furusawa et al., 2005) and posses some repellent activities towards termite ( zhu et al.,2001). the cheap aromatic, velencene from valencia orange has been successfully converted to nootkatone (1) via biotransformation using different microbes ( furusawa et al., 2005: del rio et al., 1991: wilson et al., 1978). there are few reports on microbial transformation of nootkatone (1). recentlly (furusawa et al. 2005) have obtained structurally interesting metabolites from nootkatone by the action of aspegillus niger. fusarium culmorum, botryosphaeria dothida (furusawa et al., 2005), this article deals with identification and testing of biological activities of the meta bolites formed by m.phaseolina, a plant pathogenchandana-amarasingha-samayawardanaavifauna-bundala-1.1-28.07 in its transformation of nootkatone (1). 2. experimental 2.1 preparation of fermentation medium for biotransformation m.phaseolina was inoculated and cultivated in saouraud-glucose-agar and stored in refrigerator at 40c for 2 days. the medium for biotransformation was prepared by dissolving glucose (30 g), peptone (15 g), kh2po4 (15 g), yeast extract (15 g), glycerol (30 ml) in 3 l of distilled water. the fermentation medium was distributed among 30 flasks of 250 ml capacity by adding 100 ml in each and autoclaved at 1210c for 20 min. the spores of m. phaseolina were aseptically transferred into the two broth flasks (seed flasks) and cultivated in a shaking table at room temperature for 3 days. m. phaseolina in the seed flasks were used for inoculation of the rest of 28 broth flasks and incubated them in the shaking table at room temperature for 3 days. 2.2 biotransformation of nootkatone (1) 14 ruhuna journal of science,3,pp53-60 (2008) bulugahapitiya and musharaff microbial transformation of sesquitepenoid ketone... a commercial sample of (+)-nootkatone (1) (425 mg, fluka, mf c15h2o, mw 218.338) was dissolved in distilled acetone (15 ml). the solution was evenly distributed among 29 flasks (14 mg / 0.5 ml in each flask) and further incubated for 4 days. 2.3 isolation of metabolites after four days of incubation, the culture media and the mycelium were separated by filtration and the filtrate was extracted with dichloromethane three times (1000 ml each case). the combined organic layers was washed with brine and dried over anhydrous mgso4 and then evaporated under vaccuo. the crude extract was purified by column chromatography (n-hexane: dicholomethane gradient) to afford pure metabolites their structures were elucidated with the help of ei-ms, hr-fab-ms(pos), hr-fabms (neg), 1h-nmr, 13c-nmr, cosy-450, noesy, hmbc, hmqc. 2.4 testing of biological activities commercial sample of (+)-nootkatone (1) and its metabolites 4 and 5 were tested for antibacterial, antiprotozoal, phytotoxic, insecticidal and enzyme inhibition activities. fi gure 1: (+)-noot kat one and its metaboli tes 3. results and discussion mainly four compounds were recognized as metabolites (figure 1); c-11 stereoisomeric mixture of 11, 12-dihydroxy-11, 12-dihydronootkatone (2, 3) (yellow oil, 66% isolated yield), 13-hydroxynootkaone (4) (pale yellow oil, 12 % isolated yield), 12-hydroxy11, 12dihydronootkatone (5) (yellow oil, 15% isolated yield) respectively. the compound 4 and 5 were afforded as single stereoisomers. the relative configurations at stereogenic centres of each isomer were determined with the help of 2d nmr data. the stereoi 15 bulugahapitiya and musharaff ruhuna journal of science,3,pp53-60 (2008) microbial transformation of sesquitepenoid ketone... somereic ratio of compound 2 and 3 was determined to be 1:6 by using their 1h-nmr spectra. table 1: 600mhz 1h-nmr spectral data of compounds 2-5 in cdcl3 the spectral data of 1h-nmr and 13c-nmr of each compound are given in table 1 and table 2 respectively. the compounds 2, 3 and 5 have already been reported as nootkatone metabolites (mai furusawa et al., 2005) whereas the compound 4 was first identified as nootkatone metabolite in this study. (+)-nootkatone and the metabolite 5 showed to be possessed only antiprotozoal activity in their biological activities testing. 16 ruhuna journal of science,3,pp53-60 (2008) bulugahapitiya and musharaff microbial transformation of sesquitepenoid ketone... table 2: 400mhz 13c-nmr data of compound 2-5, in cdcl3 ctd.,vajiramanuscript-rjs-reopened-nov-2011 17 bulugahapitiya and musharaff ruhuna journal of science,3,pp53-60 (2008) microbial transformation of sesquitepenoid ketone... the possible reactions types in this transformation are shown in figure 2. in general, oxidation reactions and hydroxylation reactions are more common in terpenoid nuclei dur ing the enzymatic transformations. 18 ruhuna journal of science,3,pp53-60 (2008) bulugahapitiya and musharaff microbial transformation of sesquitepenoid ketone... fi gure 2 : pos sible react ion ty pes in biotrans for mat ion of nootkat one by m. phas eol ina 4. conclusion microbial transformation is a good tool for the structural modification of natural and synthetic compounds with higher stereoselectivity, leading to synthesis of potent derivatives of biological important compounds. this is an ease of handling, low cost, environmentally friendly process. the structure and the stereochemistry of the metabolites depend on the fungal species used in biotransformation. the action of m.phaseolina on nootkatone (1) produced four main compounds; the compounds 2, and 3 as an inseparable 1:6 mixture of stereoisomers and the compounds 4 and 5 as single stereoisomers. out of these four compounds, the compounds 4 was first identified as a 9 nootkatone metabolite in our study. the parental compounds (1) and the compounds 4 and 5 showed antiprotozoan activity. acknowledgements the authors highly appreciate nam s & t centre, india for providing financial support and hej research institute, international centre for chemical sciences, university of karachi, pakistan for providing necessary support for this work. 19 bulugahapitiya and musharaff ruhuna journal of science,3,pp53-60 (2008) microbial transformation of sesquitepenoid ketone... references bock, g., benda, i. and schreier, p. 1988. microbial transformation of geraniol and nerol by botrytis cinerea, applied microbiology and biochemistry, 27: 351-357 bock, g., benda, i., schreier, p. 2006. biotransformation of linalool by botrytis cinerea, journal food science, 51(3): 659-662 choudhary i.m., musharaff, s.g, sami, a. and atta-ur rahman, 2004. microbial transformation of sesquiterpenes, (-)-ambrox and (+)-sclareolide, helatica. chemica acta, 87: 2685-2693 devkota k.p., choudhary i.m., nawaz s.a., lannang, a.l., lenta b.n., fokoup p.a. and sewald n. 2007. microbial transformation of the steroidal alkaloid dictyophlebine by rhizopus stolonifer, chemical and pharmaceutical bulletin, 55(4): 682-684 furusawa m., hashimoto t., noma y. and asakawa y. 2005. biotransformation of citrus aromatics nootkatone and valencene by microorganism, chemical and pharmaceutical bulletin, 53(11): 1423-1429 furusawa m, hashimoto, t., noma y. and asakawa, y. 2005. highly efficient production of nootkatone, the grapefruit aroma from valencene by biotransformation, chemical and pharmaceutical bulletin, 53 (11): 1513-1514 hieda, t., mikami, y., obi, y. and kisaki t. 1983. microbial transformation of labdanes, cis-aienol and sclareal, agricultural & biological chemistry, 47(2): 243-250 ishida t., 2005. biotransformation of terpenoids by mammals, microorganism and plant-cultured cells, chemistry and biodiversity, 155(5): 569-590 joglkar, s.s., dhavlikar, r.s. 1969. identification of metabolic produced by a pseudomonas from citronella and citral, applied microbiology, 18(6): 1084-1087 rio d., 1991. accumulation of the sesquiterpene nootkatone and valencene by callus culture of citrus paradise, citrus limonia and citrus aurantium, plant cell reports 10:410-413 shawa e.d. 1981. importance of nootkatone to the aroma of grapefruit oil and flavor of grapefruit juice, journal of agriculture and food chemistry, 29: 677-679 welf-rainer-abraham, staut b. and kieslich k. 1986. microbial transformation of terpenes, applied microbiology and bioechnology, 24: 24-30 wilson iii c.w., 1978. synthesis of nootkatone fom valencene, journal of agricultural food chemistry, 26 (6):1430-1432 zhu, b.c.r, henderson, g., chen, i., maistrello, l., laine, r.a. 2001. nootkatone is a repellent for formosan subterranean termite (coptotermes farmesanus), journal of insect science, 9: 523-531 20 ruhuna journal of science vol 13 (1): 14-28, june 2022 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v13i1.112 university of ruhuna faculty of science, university of ruhuna sri lanka 14 generalized stochastic restricted lars algorithm manickavasagar kayanan1,2 and pushpakanthie wijekoon3 1postgraduate institute of science, university of peradeniya, peradeniya, sri lanka 2 department of physical science, university of vavuniya, vavuniya, sri lanka 3 department of statistics and computer science, university of peradeniya, peradeniya, sri lanka *correspondence: mgayanan@vau.jfn.ac.lk; orcid: https://orcid.org/0000-0003-2662-4383 received: 29th july 2021; revised: 21st april 2022; accepted: 17th may 2022 abstract the least absolute shrinkage and selection operator (lasso) is used to tackle both the multicollinearity issue and the variable selection concurrently in the linear regression model. the least angle regression (lars) algorithm has been used widely to produce lasso solutions. however, this algorithm is unreliable when high multicollinearity exists among regressor variables. one solution to improve the estimation of regression parameters when multicollinearity exists is adding preliminary information about the regression coefficient to the model as either exact linear restrictions or stochastic linear restrictions. based on this solution, this article proposed a generalized version of the stochastic restricted lars algorithm, which combines lasso with existing stochastic restricted estimators. further, we examined the performance of the proposed algorithm by employing a monte carlo simulation study and a numerical example. keywords: lasso, lars, stochastic linear restrictions. 1 introduction the biased estimators such as ridge estimator (re) (hoerl and kennard 1970), almost unbiased ridge estimator (aure) (singh et al. 1986), liu estimator (le) (liu 1993), almost unbiased liu estimator (aule) (akdeniz and kaçiranlar 1995), principle component regression estimator (pcre) (massy 1965), r-k class estimator (baye and parker 1984), r-d class estimator (kaçiranlar and sakallıoğlu 2001) and sample information optimal estimator (sioe) (kayanan and wijekoon 2019) have been widely used in literature to resolve multicollinearity issue in the linear regression model. however, these estimators yield high bias when the number of explanatory variables is high, and they do not consider about irrelevant variables while fitting models. for high dimensional data, having many variables in the model and multicollinearity are major issues. to tackle these matters, tibshirani (1996) introduced least absolute shrinkage and selection operator (lasso). the lasso https://rjs.ruh.ac.lk/index.php/rjs/index https://creativecommons.org/licenses/by-nc/4.0/ mailto:mgayanan@vau.jfn.ac.lk https://orcid.org/0000-0003-2662-4383 https://orcid.org/0000-0003-2662-4383 m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 15 is a shrinkage method that was originally used for regularization and variable selection in the linear regression model. the least angle regression (lars) (efron et al. 2004) algorithm has been used to obtain the estimates of lasso. zou and hastie (2005) have shown that the lasso is unsteady when severe multicollinearity exists between the explanatory variables. therefore, they suggested elastic net (enet) estimator by combining lasso and re as a solution for this issue. furthermore, they proposed lars-en algorithm to attain enet solutions, which is a modified version of the lars algorithm. besides, kayanan and wijekoon (2020b) proposed a generalized version of lars (glars) algorithm to combine lasso with re and the other biased estimators based on sample information such as aure, le, aule, pcre, r-k class estimator, and r-d class estimator. finally, they have shown that the glars algorithm performs well when it combines lasso with r-k class and r-d class estimators. according to literature, the parameter estimation can be strengthened if prior knowledge about the regression coefficient is applied. the prior information on regression coefficients can be defined in the form of exact linear restrictions or stochastic linear restrictions. many researchers proposed stochastic restricted estimators such as mixed regression estimator (mre) (theil and goldberger 1961) stochastic restricted ridge estimator (srre) (li and yang 2010), stochastic restricted almost unbiased ridge estimator (sraure) (jibo and hu 2014), stochastic restricted liu estimator (srle) (hubert and wijekoon 2006), stochastic restricted almost unbiased liu estimator (sraule) (jibo and hu 2014), stochastic restricted principle component regression estimator (srpcre) (he and wu 2014), stochastic restricted r-k class estimator (srrk) (jibo 2014), stochastic restricted r-d class estimator (srrd) (jibo 2014), and stochastic restricted optimal estimator (sroe) (kayanan and wijekoon 2019) to incorporate prior information to the regression coefficient. stochastic restricted estimators also have the same issue as biased estimators when the linear regression model contains numerous predictors. to handle this problem, kayanan and wijekoon (2020a) proposed a stochastic restricted lars (srlars) algorithm to combine lasso and mre, and showed the superiority of the srlars over lars algorithm. this article proposes a generalized version of the stochastic restricted lars algorithm, namely srglars, to combine lasso with other stochastic restricted estimators. the prediction performance of the srglars algorithm was examined by employing a monte-carlo simulation and using a real-world example in the root mean square error (rmse) criterion. 2 model specification and the estimators consider the linear regression model 𝒚 = 𝑿𝜷 + 𝜺, (2.1) where 𝑿 is 𝑛 × 𝑝 matrix of explanatory variables, 𝜷 be the 𝑝 × 1 vector of unknown coefficients, and 𝜺 be the 𝑛 × 1 vector of disturbances such that 𝜺 ∼ 𝑁 (0,𝜎2 𝑰). m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 16 assume that there exists prior information on β, which may be expressed as a stochastic linear restriction, as (theil & goldberger, 1961) 𝝋 = 𝑹𝜷 + 𝒗, (2.2) where φ be the 𝑞 × 1 vector, 𝑹 be the 𝑞 × 𝑝 matrix with rank 𝑞, 𝒗 be the 𝑞 × 1 vector of disturbances, such that 𝒗 ∼ 𝑁 (0,𝜎2𝑾 ), 𝑾 is positive definite, and 𝐸(𝒗𝜺′) = 0. to make the variable selection and handle multicollinearity issue by incorporating prior information defined in model (2.2), kayanan and wijekoon (2020a) suggested stochastic restricted lasso type estimator (srlasso) for model (2.1) as �̂�𝑆𝑅𝐿𝐴𝑆𝑆𝑂 = argmin 𝜷 {(𝒚 − 𝑿𝜷)′(𝒚 − 𝑿𝜷)} subject to ∑ |𝛽𝑗| 𝑝 𝑗=1 ≤ 𝑡 and 𝑹𝜷 = 𝝋 − 𝒗, (2.3) where 𝑡 > 0 is a turning parameter. further, kayanan and wijekoon (2020a) proposed stochastic restricted lars (srlars) algorithm to find the srlasso estimates. note that srlars combines lasso and mre to find the estimates. to improve the srlars solutions, this article proposes a generalized version of srlars (srglars) to combine lasso with other stochastic restricted estimators such as srre, sraure, srle, sraule, srpcre, srrk, srrd and sroe. kayanan and wijekoon (2018, 2019) proposed a generalized form to express the estimators mre, srre, sraure, srle, sraule, srpcre, srrk, srrd, and sroe as �̂�𝐺 = 𝑮(𝑿 ′𝑿 + 𝑹′𝑾−𝟏𝑹)−𝟏 (𝑿′𝒚 + 𝑹′𝑾−𝟏𝝋), (2.4) where �̂�𝐺 = { �̂�𝑀𝑅𝐸 if 𝑮 = 𝑰 �̂�𝑆𝑅𝑅𝐸 if 𝑮 = (𝑿 ′𝑿 + 𝑘𝑰)−1𝑿′𝑿 �̂�𝑆𝑅𝐴𝑈𝑅𝐸 if 𝑮 = (𝑰 − 𝑘 2(𝑿′𝑿 + 𝑘𝑰)−2) �̂�𝑆𝑅𝐿𝐸 if 𝑮 = (𝑿 ′𝑿 + 𝑰)−1(𝑿′𝑿 + 𝑑𝑰) �̂�𝑆𝑅𝐴𝑈𝐿𝐸 if 𝑮 = (𝑰 − (1 − 𝑑) 2(𝑿′𝑿 + 𝑰)−2) �̂�𝑆𝑅𝑃𝐶𝑅𝐸 if 𝑮 = 𝑻ℎ𝑻ℎ ′ �̂�𝑆𝑅𝑟𝑘 if 𝑮 = 𝑻ℎ𝑻ℎ ′ (𝑿′𝑿 + 𝑘𝑰)−1𝑿′𝑿 �̂�𝑆𝑅𝑟𝑑 if 𝑮 = 𝑻ℎ𝑻ℎ ′ (𝑿′𝑿 + 𝑰)−1(𝑿′𝑿 + 𝑑𝑰) �̂�𝑆𝑅𝑂𝐸 if 𝑮 = 𝜷∗𝜷∗ ′(𝜎2(𝑿′𝑿)−𝟏 + 𝜷∗𝜷∗ ′) −1 note that 𝑘 > 0 and 0 < 𝑑 < 1 are the shrinkage/regularization parameters, 𝑰 is the 𝑝 × 𝑝 identity matrix, 𝑻ℎ = (𝑡1, 𝑡2. . . 𝑡ℎ)′ is the first ℎ columns of the eigenvectors of 𝑿′𝑿, and 𝜷∗ is the normalized eigenvector corresponding to the largest eigenvalue of 𝑿′𝑿. m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 17 3 srglars algorithm based on kayanan and wijekoon (2021a) and equation (2.4), the srglars algorithm for model (2.1) is outlined below: algorithm 1: srglars 1: standardize 𝑿 to have a mean zero with a standard deviation of one, and center the 𝒚 to have a mean zero. 2: start with all estimates of the coefficients �̂� = 0 with the residuals 𝒓 = �̂� and 𝝉 = �̂�. 3: find the predictor 𝑿𝒋 most correlated with 𝒓; 𝑗 = 1,2, . . . ,𝑝. 4: move the estimate of �̂�𝑗 from 0 towards the �̂�𝐺 direction until some other predictor 𝑿𝒌 has as large a correlation with the current residual as 𝑿𝒋 does. 5: move �̂�𝑗 and �̂�𝑘 in the direction defined by their joint �̂�𝐺 direction of the current residual on (𝑿𝒋,𝑿𝒌), until some other predictor 𝑿𝒍 eventually earns its way into the active set. 6: if a non-zero coefficient hits zero, drop its variable from the active set of variables and recomputed the current joint �̂�𝐺 direction. 7: repeat the steps 5 and 6 until srglars conditions attained. the mathematical details of the srlars algorithm are as follows: let us assume that the estimates of the coefficients �̂� and residuals 𝒓 and 𝝉 are (�̂�) 0 = [(𝛽1)0, (𝛽2)0,…,(𝛽𝑝)0 ] ′ = 𝟎, 𝒓0 = 𝒚 and 𝝉0 = 𝝋. find the predictor (𝑋𝑗)1 most correlated with 𝒓0. (𝑋𝑗)1 = max 𝑗 |𝐶𝑜𝑟(𝑋𝑗,𝒓0)|; 𝑗 ∈ (1,2,…,𝑝) (3.1) then, increase the estimate of respective regression coefficient (𝛽𝑗)1 from 0 unto any other predictor (𝑋𝑗)2 has a high correlation with 𝒓1 as (𝑋𝑗)1 does. at this stage, srglars moves in the equiangular direction between (𝑋𝑗)1 and (𝑋𝑗)2 rather than proceeding the path based on (𝑋𝑗)1. similarly, in 𝑖𝑡ℎ run, the variable (𝑋𝑗)1 eventually acquires its path in the active set, and then srglars moves in the equiangular direction among (𝑿)𝑖 = [(𝑋𝑗)1 , (𝑋𝑗)2 ,…,(𝑋𝑗)𝑖 ] ′ . proceed to add variables to the active set in this way, running in the path established by the least angle direction. during this process, (�̂�) 𝑖 = [(𝛽𝑗)1 ,(𝛽𝑗)2 ,…,(𝛽𝑗)𝑖 ] ′ is updating using the following formula: (�̂�) 𝑖 = (�̂�) 𝑖−1 + 𝛼𝑖𝒖𝑖, (3.2) m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 18 where 𝛼𝑖 ∈ [0,1] which signifies how long the estimate runs in the path before another predictor enters the model and the path turns anew, and 𝒖𝑖 is the equiangular vector. the path vector 𝒖𝑖 is computed using the formula given below based on the generalized form defined in equation (2.4): 𝒖𝑖 = 𝑮𝐸((𝑬)𝒊 ′(𝑿′𝑿 + 𝑹′𝑾−𝟏𝑹)(𝑬)𝒊) −𝟏(𝑬)𝒊 ′(𝑿′𝒓𝒊−𝟏 + 𝑹′𝑾 −𝟏𝝉𝒊−𝟏), (3.3) where (𝑬)𝒊 = [(𝒆𝑗)1 ,(𝒆𝑗)2 ,… ,(𝒆𝑗)𝑖 ], and (𝑒𝑗)𝑖 be the 𝑗 th standard unit vector in 𝑅𝑝, which has the record of selected variables in every succeeding steps, and 𝑮𝐸 is a generalized matrix that can be changed by respective expressions for any of stochastic restricted estimators of our interest as outlined in table 1. then, 𝛼𝑖 be calculated as follows: 𝛼𝑖 = min {𝛼𝑖 +,𝛼𝑖 −,𝛼𝑖 ∗}, (3.4) where 𝛼𝑖 ± = 𝐶𝑜𝑟((𝑋𝒋)𝑖 ,𝒓𝒊−𝟏)±𝐶𝑜𝑟(𝑋𝑗,𝒓𝒊−𝟏) 𝐶𝑜𝑟((𝑋𝒋)𝑖 ,𝒓𝒊−𝟏)±𝐶𝑜𝑟(𝑋𝑗,(𝑿)𝑖𝒖𝑖) (3.5) for any 𝑗 such that (𝛽𝑗)𝑖−1 = 0, and 𝛼𝑖 ∗ = − (�̂�) 𝑖−1 𝒖𝑖 (3.6) for any 𝑗 such that (𝛽𝑗)𝑖−1 ≠ 0. if 𝛼𝑖 = 𝛼𝑖 ∗, then (𝑬)𝒊 is reformed by deleting the column 𝑒𝑗 from (𝑬)𝒊−𝟏. then 𝒓𝒊 and 𝝉𝒊 can be calculated as 𝒓𝒊 = 𝒓𝒊−𝟏 − 𝛼𝑖(𝑿)𝑖𝒖𝑖 and (3.7) 𝝉𝒊 = 𝝉𝒊−𝟏 − 𝛼𝑖(𝑹)𝑖𝒖𝑖, (3.8) where (𝑹)𝑖 = [(𝑹𝒋)1 ,(𝑹𝒋)2 ,…,(𝑹𝒋)𝑖 ]. then proceed to the next step where (𝑗)𝑖+1 is the value of 𝑗 such that 𝛼𝑖 = 𝛼𝑖 +or 𝛼𝑖 = 𝛼𝑖 − or 𝛼𝑖 = 𝛼𝑖 ∗ . proceed with the algorithm until𝛼𝑖 = 1. table 1: 𝐺𝐸 of the estimators for srglars. estimators 𝑮𝐸 mre (𝑬)𝒊 srre (𝑬)𝒊((𝑬)𝒊 ′ (𝑿′𝑿 + 𝑘𝑰)(𝑬)𝒊) −1(𝑬)𝒊 ′ 𝑿′𝑿(𝑬)𝒊 sraure (𝑬)𝒊(𝑰𝑃𝐸 − 𝑘 2((𝑬)𝒊 ′ (𝑿′𝑿 + 𝑘𝑰)(𝑬)𝒊) −2) srle (𝑬)𝒊((𝑬)𝒊 ′ (𝑿′𝑿 + 𝑰)(𝑬)𝒊) −1(𝑬)𝒊 ′ (𝑿′𝑿 + 𝑑𝑰)(𝑬)𝒊 sraule (𝑬)𝒊(𝑰𝑃𝐸 − (1 − 𝑑) 2((𝑬)𝒊 ′ (𝑿′𝑿 + 𝑰)(𝑬)𝒊) −2) srpcre 𝑻ℎ𝐸𝑻ℎ𝐸 ′ (𝑬)𝒊 srrk 𝑻ℎ𝐸𝑻ℎ𝐸 ′ (𝑬)𝒊((𝑬)𝒊 ′ (𝑿′𝑿 + 𝑘𝑰)(𝑬)𝒊) −1(𝑬)𝒊 ′ 𝑿′𝑿(𝑬)𝒊 srrd 𝑻ℎ𝐸𝑻ℎ𝐸 ′ (𝑬)𝒊((𝑬)𝒊 ′ (𝑿′𝑿 + 𝑰)(𝑬)𝒊) −1(𝑬)𝒊 ′ (𝑿′𝑿 + 𝑑𝑰)(𝑬)𝒊 sroe (𝑬)𝒊((𝑬)𝒊 ′ 𝜷∗𝜷∗ ′(𝑬)𝒊(𝜎 2((𝑬)𝒊 ′ 𝑿′𝑿(𝑬)𝒊) −𝟏 + (𝑬)𝒊 ′ 𝜷∗𝜷∗ ′(𝑬)𝒊) −1 m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 19 in table 1, 𝑰𝑃𝐸 is the 𝑝𝐸 × 𝑝𝐸 identity matrix, 𝑝𝐸 is the amount of selected variables in each succeeding step, and 𝑻ℎ𝐸 = (𝑡1, 𝑡2. . . 𝑡ℎ𝐸) is the first ℎ𝐸 column of the standardized eigenvectors of (𝑬)𝒊 ′ 𝑿′𝑿(𝑬)𝒊. we can apply srglars to consolidate lasso and any of the stochastic restricted estimators listed in table 1. the suitable value of regularization parameter 𝑘 or 𝑑 of the proposed algorithms can be chosen by 10-fold cross-validation for every 𝑡 as outlined in appendix c. we can get a separate algorithm for each stochastic restricted estimator by referring srglars as lars-mre, lars-srre, lars-sraure, lars-srle, larssraule, lars-srpcre, lars-srrk, lars-srrd, and lars-sroe when ge equals to the corresponding expression of mre, srre, sraure, srle, sraule, srpcre, srrk, srrd, and sroe, respectively. 4 discussion the performance of srglars algorithm by considering different combinations of lasso and stochastic restricted estimators were examined using rmse criterion, which is the expected prediction error of the algorithm for each estimator, and is defined as 𝑅𝑀𝑆𝐸(�̂�) = √ 1 𝑛∗ (𝒚∗ − 𝑿∗�̂�) ′ (𝒚∗ − 𝑿∗�̂�), (4.1) where (𝒚∗,𝑿∗) are the new observations, 𝑛∗ is the size of new observations, and �̂� is the estimated value of 𝜷 by the corresponding algorithm. note that we cannot provide theoretical conditions for the superiority of particular algorithms in variable selection methods. therefore, we use a monte carlo simulation and a real-world example to compare the srglars algorithms. 4.1 simulation study we used the following formula to generate the explanatory variables based on mcdonald and galarneau (1975): 𝑥𝑖,𝑗 = √(1 − 𝜌 2)𝑧𝑖,𝑗 + 𝜌𝑧𝑖,𝑚+1; 𝑖 = 1,2, . . . ,𝑛. 𝑗 = 1,2, . . . ,𝑚. (4.2) where 𝑧𝑖,𝑗 is an independent standard normal pseudo-random number, and 𝜌 is the correlation among any two explanatory variables. in this study, we have used 100 observations with 20 explanatory variables, in which 70 observations were used to fit the model and 30 observations were used to compute the rmse. a dependent variable is generated by using the following equation, 𝑦𝑖 = 𝛽1𝑥𝑖,1 + 𝛽2𝑥𝑖,2 + ⋯+ 𝛽20𝑥𝑖,20 + 𝜀𝑖, (4.3) where 𝜀𝑖 is a normal pseudo-random number with 𝐸(𝜀𝑖) = 0 and 𝑉 (𝜀𝑖) = 𝜎 2 = 1. m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 20 we choose 𝛽 = (𝛽1,𝛽2,…,𝛽20) as the normalized eigenvector corresponding to the largest eigenvalue of 𝑿′𝑿 for which 𝜷′𝜷 = 1 (mcdonald and galarneau 1975). prior information was defined based on nagar and kakwani’s (1964) approach, which is described in appendix b. further, we have assumed the first four elements of olse estimates as 𝒃 (see appendix b). to study the effects of various degrees of multicollinearity on the data, we pick 𝜌 = (0.5, 0.7, 0.9), which signifies weak, moderate, and high multicollinearity, respectively. since the execution time for the algorithm simulation is long due to cross-validation, we simulated only 50 datasets. future simulation studies will be implemented with a greater number of simulations using cluster programming. figures 1-3 and tables 2-4 show the cross-validated rmse and the median cross-validated rmse of the srglars algorithms, respectively, for 50 simulated data. fig 1. cross-validated rmse values of the srglars algorithms when ρ = 0.5 fig 2. cross-validated rmse values of the srglars algorithms when ρ = 0.7 m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 21 fig 3. cross-validated rmse values of the srglars algorithms when ρ = 0.9 table 2: median cross-validated rmse values of the srglars algorithms when ρ = 0.5. algorithms rmse (k, d) t selected variables lars-mre 3.2805 – 6.1630 15 lars-srre 3.3531 0.3 6.1777 16 lars-sraure 3.2832 0.1 6.1631 16 lars-srle 3.3531 0.7 6.1773 16 lars-sraule 3.3012 0.99 6.1630 16 lars-srpcre 3.3070 – 6.4343 16 lars-srrk 3.3122 0.1 6.9493 17 lars-srrd 3.2976 0.9 6.9497 16 lars-sroe 3.0182 – 4.1102 18 table 3: median cross-validated rmse values of the srglars algorithms when ρ = 0.7. algorithms rmse (k, d) t selected variables lars-mre 3.2041 – 6.8458 15 lars-srre 3.3575 0.1 7.5043 16 lars-sraure 3.3884 0.8 6.7785 16 lars-srle 3.3537 0.9 7.5049 17 lars-sraule 3.3886 0.2 6.7786 16 lars-srpcre 3.3135 – 7.7712 16 lars-srrk 3.2525 1.0 8.5411 17 lars-srrd 3.2534 0.5 8.1097 17 lars-sroe 3.0324 – 4.3864 18 m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 22 table 4: median cross-validated rmse values of the srglars algorithms when ρ = 0.9. algorithms rmse (k, d) t selected variables lars-mre 3.3391 – 11.6628 16 lars-srre 3.2722 0.1 11.5897 17 lars-sraure 3.3354 0.1 11.6606 16 lars-srle 3.3110 0.99 11.6628 17 lars-sraule 3.3107 0.99 11.6628 16 lars-srpcre 3.3782 – 12.2288 16 lars-srrk 3.3531 0.7 12.9540 17 lars-srrd 3.3276 0.3 13.0038 17 lars-sroe 2.9878 – 4.5654 18 from figure 1-3, we can observe that the lars-sroe algorithm outperformed other srglars algorithms in rmse criterion under all degrees of multicollinearity. from tables 2-4, we observe that the sroe included more variables than the other srglars algorithms, although it has minimum rmse. therefore, in practical situations, if a researcher intends to reduce the number of variables from the model, they may consider other srglars algorithms based on the interested variables and prediction performance using the plot of coefficient paths discussed in the real-world example. 4.2 real-world example as a real-world data set, we considered the prostate cancer data (stamey et al. 1989), which is used by tibshirani (1996), efron et al. (2004) and zou and hastie (2005) to study the performance of lasso, lars algorithm and enet. the prostate cancer data contains 97 observations and 8 predictors such as log cancer volume (lcavol), log prostate weight (lweight), age, log of the amount of benign prostatic hyperplasia (lbph), seminal vesicle invasion (svi), log capsular penetration (lcp), gleason score (gleason) and percentage gleason score 4 or 5 (pgg45). the dependent variable is the log of prostate-specific antigen (lpsa). the variance inflation factor (vif) values of the predictor variables are 3.09, 2.97, 2.47, 2.05, 1.95, 1.37, 1.36 and 1.32, and the condition number is 243, which exposes high multicollinearity between the predictor variables. this data set is associated in “lasso2” r package. we have used 67 observations to fitting the model, and 30 observations to compute the rmse. we have assumed (nagar and kakwani 1964) that the first three olse estimates of prostate cancer data are unbiased, and we defined the prior information for this data based on nagar and kakwani’s (1964) approach, as described in appendix b. the cross-validated rmse of the srglars algorithms are shown in table 5, and coefficient paths of respective srglars algorithm are shown in figure 4 in appendix a. m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 23 table 5: cross-validated rmse values of prostate cancer data using srglars. algorithms rmse (k, d) t selected variables lars-mre 0.77784 – 1.5632 6 lars-srre 0.73567 0.38 1.7329 8 lars-sraure 0.77784 0.01 1.5632 6 lars-srle 0.73566 0.62 1.7331 8 lars-sraule 0.77784 0.99 1.5632 6 lars-srpcre 0.74084 – 1.5656 7 lars-srrk 0.74083 0.04 1.5643 7 lars-srrd 0.74084 0.96 1.5643 7 lars-sroe 0.69897 – 1.1226 8 table 5 shows that the lars-sroe algorithm outperforms other algorithms on prostate cancer data, which is agreed the results obtained in the simulation study. further, we can note that lars-srre, lars-srle, and lars-sroe did not make any variable selections on prostate cancer data. figure 4 (see appendix a) shows that the choice of variables is different for each srglars algorithms. note that in all graphs, the order of selection of variables is the same except figure 4(i). further, the coefficients relevant to the variables lcp, age and gleason are very small for fig 4((a)-(h)) and those coefficients are almost zero in fig 4(i). this indicates that the contribution from those three variables is very low for the fitted model. 5 conclusions this study presented a generalized version of stochastic restricted lars (srglars) algorithm to combine lasso with existing stochastic restricted estimators. we have shown the superiority of srglars when it combines lasso with sroe (larssroe) using a monte carlo simulation and a real-world example in rmse criterion. however, lars-sroe is less likely to make variable selections when looking at the results directly. however, by examining the size of the coefficients as shown in the example, we can drop some irrelevant variables form the model. if a researcher wanted to reduce some variables from the model, the other combinations of the srglars algorithm could be considered to fit the regression model by studying the rmse and coefficients estimates based on the coefficients path. acknowledgements anonymous reviewers are acknowledged for their comments on the initial manuscript. m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 24 references akdeniz f, kaçiranlar s. 1995. on the almost unbiased generalized liu estimator and unbiased estimation of the bias and mse. communication in statistics theory and methods 24(7):1789–1797. baye mr, parker df. 1984. combining ridge and principal component regression: a money demand illustration. communications in statistics-theory and methods 13(2):197–205. efron b, hastie t, johnstone i, tibshirani r. 2004. least angle regression. the annals of statistics 32(2):407–499. he d, wu y. 2014. a stochastic restricted principal components regression estimator in the linear model. the scientific world journal, vol. 2014, article id 231506, 6p. https://doi.org/10.1155/2014/231506 hoerl e, kennard rw. 1970. ridge regression: biased estimation for nonorthogonal problems. technometrics 12:55–67. hubert mh, wijekoon p. 2006. improvement of the liu estimator in linear regression model. statistical papers 47:471–479. jibo w. 2014. on the stochastic restricted r-k class estimator and stochastic restricted r-d class estimator in linear regression model. journal of applied mathematics, vol. 2014, article id 173836, 6 p. https://doi.org/10.1155/2014/173836 jibo w, hu y. 2014. on the stochastic restricted almost unbiased estimators in linear regression model. communications in statistics-simulation and computation 43(2):428–440. kaçiranlar s, sakallıoğlu s. 2001. combining the liu estimator and the principal component regression estimator. communications in statistics-theory and methods 30(12):2699–2705. kayanan m, wijekoon p. 2018. stochastic restricted biased estimators in misspecified regression model with incomplete prior information. journal of probability and statistics, vol. 2018, article id 1452181, 8 p. https://doi.org/10.1155/2018/1452181. kayanan m, wijekoon p. 2019. optimal estimators in misspecified linear regression model with an application to real-world data. communications in statistics: case studies, data analysis and applications 4(3-4):151–163. kayanan m, wijekoon p. 2020a. stochastic restricted lasso-type estimator in the linear regression model. journal of probability and statistics, vol. 2020, article id 7352097, 7 p. https://doi.org/10.1155/2020/7352097. kayanan m, wijekoon p. 2020b. variable selection via biased estimators in the linear regression model. open journal of statistics 10(1):113–126. li y, yang h. 2010. a new stochastic mixed ridge estimator in linear regression model. statistical papers 51:315–323. liu k. 1993. a new class of biased estimate in linear regression. communication in statistics theory and methods 22(2):393–402. massy wf. 1965. principal components regression in exploratory statistical research. journal of the american statistical association 60:234–266. mcdonald gc, galarneau di. 1975. a monte carlo evaluation of some ridge-type estimators. journal of the american statistical association 70:407–416. nagar al, kakwani nc. 1964. the bias and moment matrix of a mixed regression estimator. econometrica 32:174–182. singh b, chaubey yp, dwivedi td. 1986. an almost unbiased ridge estimator. the indian journal of statistics 48:342–346. stamey ta, kabalin jn, mcneal je, johnstone im, freiha f, redwine ea, yang n. 1989. prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate: ii. radical prostatectomy treated patients. journal of urology 141:1076–1083. theil h, goldberger as. 1961. on pure and mixed statistical estimation in economics. international economic review 2:65–78. tibshirani r. 1996. regression shrinkage and selection via the lasso. journal of the royal statistical society: series b (methodological) 58:267–288. zou h, hastie t. 2005. regularization and variable selection via the elastic net. journal of the royal statistical society: series b 67:301–320. https://doi.org/10.1155/2014/231506 m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 25 appendix a: figure 4 (a) (b) (c) (d) m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 26 (e) (f) (g) (h) m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 27 (i) fig 4. coefficient paths of the (a) lars-mre, (b) lars-srre, (c) larssraure, (d) lars-srle, (e) lars-sraule, (f) lars-srpcre, (g) lars-srrk (h) lars-srrd and (i) lars-sroe versus 𝒕 = ∑ |𝜷𝒋| 𝒑 𝒋=𝟏 for the prostate cancer data. m. kayanan and p. wijekoon generalized stochastic restricted lars algorithm ruhuna journal of science vol 13 (1): 14-28, june 2022 28 appendix b: selection of prior information according to nagar and kakwani (1964), we can define the prior information as follows: let 𝜷1 be a vector of some selected 𝑞 elements of 𝜷 and 𝜷2 is the rest of elements. assume that 𝒃 is the known unbiased estimates of 𝜷1. by using the “two sigma rule”, now we can write the range of 𝜷1 as 𝒃 ± 2𝑆𝐸(𝒃). based on that we can set the expressions of equation (2.2) as �̂� = 𝒃, �̂� = ( 1 0 ⋯ 0 ⋯ 0 0 1 ⋯ 0 ⋯ 0 ⋮ 0 ⋮ 0 ⋱ ⋯ ⋮ 1 ⋯ ⋯ 0 0 ) 𝑞×𝑝 , 𝜷 = ( 𝜷1 𝜷2 ) and 𝜎2�̂� = ( 𝑆𝐸(𝑏1) 0 ⋯ 0 0 ⋮ 0 𝑆𝐸(𝑏2) ⋮ 0 ⋯ ⋱ ⋯ 0 ⋮ 𝑆𝐸(𝑏𝑞) ) 𝑞×𝑞 . appendix c: k-fold cross-validation to estimate shrinkage parameters, (k, d) step 1: split the data set into 𝐾 groups. step 2: for each unique group 𝑖 = 1,2, . . . ,𝐾: • take one group as a test data set • take the remaining 𝐾 − 1 groups as a training data set • estimate the respective estimator �̂�(𝑘,𝑑) with shrinkage parameters (𝑘,𝑑) using the training data set. use intital values of 𝑘,𝑑 as 0.01, and then compute the rmse cross-validation errors separately as follows: 𝐶𝑉_𝑅𝑀𝑆𝐸𝑖(�̂�(𝑘,𝑑)) = √ 𝐾 𝑛 ∑ (𝑦𝑖 ∗ − ∑ 𝑥𝑖𝑗 ∗ �̂�𝑗(𝑘,𝑑) 𝑝 𝑗=1 ) 2 𝑛/𝐾 𝑖=1 , where 𝑦𝑖 ∗ and 𝑥𝑖𝑗 ∗ are the variables belongs to the test data set. step 3: then, find the overall cross-validation errors as follows: 𝐶𝑉_𝑅𝑀𝑆𝐸 (�̂�(𝑘,𝑑)) = 1 𝐾 ∑𝐶𝑉_𝑅𝑀𝑆𝐸𝑖 (�̂�(𝑘,𝑑)) 𝐾 𝑖=1 step 4: continue this procedure by increasing the values of 𝑘 or 𝑑 by a small increment, and choose the value of 𝑘 or 𝑑 that makes 𝐶𝑉_𝑅𝑀𝑆𝐸 (�̂�(𝑘,𝑑)) smallest. rjs-2008-lwsomathilake-final.dvi ruhuna journal of science vol. 4, september 2009, pp. 1–12 http://www.ruh.ac.lk/rjs/ i s s n 1800-279x © 2009 faculty of science university of ruhuna. numerical solutions of reaction-diffusion systems with coupled diffusion terms l.w. somathilake department of mathematics, university of ruhuna, matara, sri lanka. (correspondence: sthilake@maths.ruh.ac.lk) abstract. numerical solutions of some initial-boundary value problem associated with a particular reaction-diffusion systems namely gray-scott model responsible for spatial pattern formation are considered. the aim of this paper is to numerically solve the above system with coupled diffusion terms. firstly, using some linear transformations, a general form of diffusion coupled reactiondiffusion system is converted into reaction-diffusion system with uncoupled diffusion terms and then, some finite difference schemes (based on (hoff 1978)) are constructed to obtain the solutions. finally, the graphical representation of the numerical solutions are presented. key words : reaction-diffusion equations, pattern formations, finite difference methods. 1. introduction properties of analytical solutions of diffusion-coupled reaction-diffusion systems have been reported for instance in (kirane 1989, badraoui 2002, p. collet 1994, kirane 1(1993). for example consider the diffusion-coupled reaction-diffusion system ∂u ∂t = a∆u − uh(v) x ∈ ω t > 0, ∂v ∂t = b∆u + d∆v + uh(v) x ∈ ω t > 0,      (1) with initial conditions u(x, 0) = u0(x), v(x, 0) = v0(x), x ∈ ω (2) on a bounded domain ω ⊂ rn with neumann boundary conditions, b > 0, a 6= d, v0 ≥ bu0 a − d ≥ 0, h(s) is a differentiable nonnegative function on r. the major result of this has been regarded in (kirane 1989) while the existence of global solutions for the system (1) on unbounded domains has been reported in (badraoui 2002). the existence of global solutions in rn for (1) with h(s) = vm has been studied in (p. collet 1994). the quasilinear system of reaction-diffusion equations ∂u ∂t = ∇.(a(u)∇u) − uh(u)v x ∈ ω t > 0, ∂v ∂t = ∇.(b(v)∇v) + uh(u)v − λv x ∈ ω t > 0,      (3) 1 lw somathilake: numerical solutions of reaction-diffusion systems ... 2 ruhuna journal of science 4, pp. 1–12, (2009) with initial conditions u(x, 0) = u0(x), v(x, 0) = v0(x), x ∈ ω (4) and with neumann or dirichlet boundary conditions, is studied in (kirane 1(1993) where in particular, the existence of a globally bounded solution has been shown. moreover, the large time behavior of the solution has also been discussed. this type of mathematical models may arise when constructing mathematical models for the reaction-diffusion processes of substances (eg. chemicals, species, deceases etc.) such that diffusion of one substance depend on the diffusion of some other substance appear in precess. the turing instability condition which give rise to pattern formation in reaction-diffusion system are well known (murray 2003). some of such patterns formation reaction diffusion systems are gray-scott model(webpage ????a), fitzhugh-nagumo model(webpage ????b). many researchers have studied numerical solutions of these systems without coupled diffusion terms(seaid 2001, turk 1992, grindrod 1991). above investigations of analytical solutions of diffusion-coupled reaction-diffusion systems and numerical solutions of pattern formation reaction diffusion systems motivated us to investigate numerical solutions of diffusion-coupled pattern formation reaction-diffusion systems. we aim in investigating numerical solutions of diffusion-coupled pattern formation reaction-diffusion systems using finite difference techniques. the paper is organized as follows: in section 2, we consider a general form of diffusion-coupled reaction-diffusion system and its transformation into a reaction-diffusion system with uncoupled diffusion terms. in section 3, some finite difference schemes are constructed for a general form of reactiondiffusion system with uncoupled diffusion terms. in section 4, numerical solutions, which are obtained using constructed semi-implicit finite difference method, of gray-scott type diffusion-coupled reaction diffusion system are presented. when semi-implicit finite difference scheme is applied to a reaction diffusion system, a system of linear equations arise in each time step. in order to solve these linear systems conjugate-gradient method has been used in computer programs. 2. general form of a diffusion-coupled reaction-diffusion system let v1(x, t), v2(x, t) ∈ rm be variables explaining some characteristics of model system (eg. concentration, population size) in suitable units. consider the following general form of diffusion-coupled reaction-diffusion system: ∂v1 ∂t = a∆v1 + f(v1, v2), x ∈ ω, t > 0, ∂v2 ∂t = b∆v1 + d∆v2 + g(v1, v2), x ∈ ω, t > 0,      (5) with initial conditions v1(x, 0) = v10(x) for x ∈ ω v2(x, 0) = v20(x) for x ∈ ω } (6) lw somathilake: numerical solutions of reaction-diffusion systems ... ruhuna journal of science 4, pp. 1–12, (2009) 3 and boundary conditions ∂v1 ∂n = 0, for x ∈ ∂ω, t > 0. ∂v2 ∂n = 0, for x ∈ ∂ω, t > 0.          (7) here f = (f1, f2, ..., fm) t , g = (g1, g2, ..., gm) t , ω ⊂ rd is a connected, bounded open set with piecewise smooth boundary, and n is the outward unit normal vector to the boundary. a, b, and d are diagonal matrices of order m whose ith diagonal entries are respectively ai, bi, and di each of which is a real constants and ai 6= di for i = 1, 2, ..., m. 2.1. transformation of a diffusion-coupled reaction-diffusion system to a reaction-diffusion system with uncoupled diffusion terms let us define the linear transformations l : (v1, v2)−→ ( v1, v2 − (a − d) −1bv1 ) = (v, w). under l, the system of equations (5) is transformed to ∂v ∂t = a∆v + f1 ( v, w + (a − d)−1bv ) , x ∈ ω, t > 0, ∂ ∂t (w + (a − d)−1bv) = b∆v +d∆(w + (a − d)−1bv) +g1 (v, w + (a − d)−1bv) , x ∈ ω, t > 0,                  (8) by simplifying have ∂v ∂t = a∆v + f1 ( v, w + (a − d)−1bv ) , x ∈ ω, t > 0, ∂w ∂t = −(a − d)−1(ba − db)∆v + b∆v +d∆w + g1 ( v, w + (a − d)−1bv ) −(a − d)−1bf1 ( v, w + (a − d)−1bv ) , x ∈ ω, t > 0,                  (9) this system can be written in the form: ∂v ∂t = a∆v + f2(v, w), x ∈ ω, t > 0, ∂w ∂t = d∆w + g2(v, w), x ∈ ω, t > 0,          (10) lw somathilake: numerical solutions of reaction-diffusion systems ... 4 ruhuna journal of science 4, pp. 1–12, (2009) where f2(v, w) = f1 (v, w + (a − d)−1bv) , g2(v, w) = g1 (v, w + (a − d)−1bv) −(a − d)−1bf1 (v, w + (a − d)−1bv)        (11) the initial conditions are reduced to: v(x, 0) = v0(x) = v1(x, 0) = v10(x), for x ∈ ω w(x, 0) = w0(x) = v2(x, 0) − (a − d) −1bv1(x, 0) = v20(x) − (a − d) −1bv10(x), for x ∈ ω        (12) and boundary conditions are reduced to ∂v ∂n = 0, for x ∈ ∂ω, t > 0. ∂w ∂n = 0, for x ∈ ∂ω, t > 0.          (13) now take u = (v, w)t , c = ( a 0 0 d ) , and f = (f2(v, w), g2(v, w)) t . then the system is reduced to the form: ∂u ∂t = c∆u + f (u), x ∈ ω, t > 0, (14) with initial data u(x, 0) = u0(x) for x ∈ ω and boundary conditions ∂u ∂n = 0 for x ∈ ∂ω t > 0. this system has 2m number of equations. the reaction-diffusion system (5) with coupled diffusion terms is now reduced to the reaction-diffusion system (14) which has no coupled diffusion terms. this system can be solved numerically for approximate solutions of u = (v, w)t . finally, approximations for v1 and v2 can be obtained using l−1 transformation. 3. finite difference schemes for reaction-diffusion systems in this section finite difference schemes for reaction diffusion system (14) are constructed based on (hoff 1978). let x = (x1, x2, ..., xd ) ∈ r d , ∆xi = hi (i = 1, 2, ..., d) be an increment in xi (i = 1, 2, ..., d) and τ be an increment in t. also let xk = (k1h1, k2h2, ..., kd hd ) for k = (k1, k2, ..., kd ) ∈ z d and tn = nτ for n ∈ z. we shall approximate u(xk, tn) ≡ u nk for k ∈ i ; where i is an appropriate index set contained in zd such that k ∈ s implies xk ∈ ω. lw somathilake: numerical solutions of reaction-diffusion systems ... ruhuna journal of science 4, pp. 1–12, (2009) 5 let m = 2m, s = d ∏ i=1 [ai, bi] and y = {{uk}k∈i : uk ∈ r m} be a vector space of rm valued functions on s. suppose that the second order accurate operator ∆2j on y are constructed such that ∣ ∣ ∣ ∣ ( ∆2j u h2j ) k − ∂2u ∂x2j (xk) ∣ ∣ ∣ ∣ ∞ ≤ c1||u||4 h 2 j , j = 1, 2, ..., d ( ∆2j u h2j ) k = ∂2u ∂x2j (η) for some η ∈ ω j = 1, 2, ..., d where c1 is a constant which is independent of u and h j. assuming that i has been defined and operators ∆2j have been constructed, we replace the differential equation (14) with the finite difference equation u n+1k −u n k τ = d ∑ j=0 ∆2j h2j ( θu n+1k + (1 − θ)u n k ) + f (15) where f is evaluated at (xk, tn,u n k ). here 0 ≤ θ ≤ 1, and u n ∈ y is defined by u nk = u (xk, tn) which is the corresponding finite difference approximation of u at the point (xk, tn). let and β j = τ/h2j . then (15) may be written as: ( i − θ d ∑ j=1 (β jc∆2j ) ) u n+1k = ( i + (1 − θ) d ∑ j=1 β jc∆2 j ) u nk + τf (16) applying initial and boundary conditions the operator ∑dj=1 β j d(xk, t,uk)∆ 2 j can be decomposed. let n be the cardinality of i and define f : [0, ∞) × sn−→y by f (t,u )k = f(xk, t,uk). also let l : [0, ∞) × sn−→y and z : [0, ∞) × sn−→y be two mappings such that l is linear and [l(t,u )u + z(t,u )]k = { d ∑ j=1 [ β jc∆2j ] u } k , (17) where t ≥ 0, u ∈ sn , and u ∈ y . using these notations the difference scheme (16) can be written in the following simple form: (i − θl)u n+1 = (i + (1 − θ)l)u n + z + τf . (18) where l, z and f are evaluated at (tn,u n). when θ = 0 in (18) we get u n+1 = (i + l)u n + z + τf . since l, z, and f are evaluated at nth time level, u -values at (n+1)th time level can be evaluated explicitly by above equation. these types of finite difference schemes are called lw somathilake: numerical solutions of reaction-diffusion systems ... 6 ruhuna journal of science 4, pp. 1–12, (2009) fully explicit schemes. when θ 6= 0 in (18) u -values at (n + 1)th time level are expressed implicitly by u -value at nth time level. these types of finite difference schemes are called implicit schemes. when θ = 1/2 the scheme is called the crank-nicolsion finite difference scheme. the case θ = 1 is called semi-implicit scheme. this semi-implicit scheme is used in numerical simulations of this paper. in the following example the finite difference scheme is implemented on a domain in two dimensional space. example 1. consider two dimensional case with ω = (a1, b1) × (a2, b2). let i = {(i, j) ; (i = 1, 2, ..., n1), ( j = 1, 2, ...n2)}, where n1 and n2 are such that (n1 + 1)h1 = b1 − a1, (n2 + 1)h2 = b2 − a2 and usual second order approximation for the laplacian operator. then ∆22 =      a a . . . a      n2×n2 +z1 and ∆21 =        a1 b1 b1 a1 b1 . . . . . . . . . b1 a1 b1 b1 a1        n2×n2 +z2; where a =        −2i i i −2i i . . . . . . . . . i −2i i i −2i        n1×n1 , a1 =        −2i −2i . . . −2i −2i        n1×n1 , b1 =        i i . . . i i        n1×n1 ; lw somathilake: numerical solutions of reaction-diffusion systems ... ruhuna journal of science 4, pp. 1–12, (2009) 7 here i is an m × m identity matrix. z1 = (z1,1,z1,2, ...,z1,n2 ) t and z2 = (z2,1,z2,2, ...,z2,n2 ) t . in the abovez1i = (1, 0, ..., 0, 1) t 1×n1 for i = 1, 2, ..., n2 z2,i = { (1, 1, ..., 1, 1)t1×n1 if i = 1, n2 (0, 0, ..., 0, 0)t1×n1 otherwise. ¿from the equation (17) we get lu + z = (β1d∆21 + β2d∆ 2 2)u that is we get l(t,u ) =    l11 . . . l1n1 ... ... ln11 . . . ln1n1    n2×n2 where li j =    β1da1 + β1db1 + β2da if i = j β1db1 if i = j ± 1 0 otherwise also z = β1dz1 + β2dz2. 4. numerical experiments in this section we consider pattern formation of diffusion coupled gray-scott model. the gray-scott model includes the following two irreversible reactions: u + 2v −→ 3v v −→ p where u and v are two reacting specimens and p an inert precipitate. the mathematical model for this reaction-diffusion process is of the form: ∂u ∂t = d1∆u + u2v − (η + ζ)u ∂v ∂t = d2∆v − u2v + η(1 − v).      (19) where u and v are the concentrations of u and v respectively and d1 and d2 are their respective diffusion coefficients. η and ζ are dimensionless feed rates of first and second reaction respectively (webpage ????a). lw somathilake: numerical solutions of reaction-diffusion systems ... 8 ruhuna journal of science 4, pp. 1–12, (2009) figure 1 time evolution of v : surface plots of v at different time levels 4.1. gray-scott model with coupled diffusion terms we consider the following diffusion-coupled gray-scott model: ∂u ∂t = d1∆u + u2v − (η + ζ)u ∂v ∂t = d∆u + d2∆v − u2v + η(1 − v)      (20) lw somathilake: numerical solutions of reaction-diffusion systems ... ruhuna journal of science 4, pp. 1–12, (2009) 9 figure 2 time evolution of v : surface plots of v at different time levels here η, ζ, d, d1 and d2 are constants and d1 6= d2. we consider this reaction-diffusion system on the bounded domain [0, 1] × [0, 1] under no flux boundary conditions and under the parameters d1 = 1.0 × 10 −7, d2 = 8 × 10 −6, d = 2.5 × 10−6, ζ = 0.005, η = 0.0006. in this case initial conditions are: u(x, y, 0) = 0.101215; (x, y) ∈ (0, 1) × (0, 1) v(x, y, 0) = v0 + (rand(200)/100000.0)v0; (x, y) ∈ [0, 1] × [0, 1]; } (21) where v0 = 0.055328. lw somathilake: numerical solutions of reaction-diffusion systems ... 10 ruhuna journal of science 4, pp. 1–12, (2009) figure 3 time evolution of v : density plots of v at different time levels it can be shown that above initial state and parameters satisfy the turing instability conditions which are the conditions should satisfy by reaction diffusion systems in order to form spatial patterns(murray 2003, turing 1952). numerical solutions of the reaction-diffusion system (20) subject to no-flux boundary conditions under initial data (21) are obtained using above introduced semi-implicit finite difference scheme. surface plots of the v-component at different time levels of those numerical solutions are shown in figures 1 and 2 (v denotes the numerical solutions of v). density plots of the same are shown in figures 3 and 4. lw somathilake: numerical solutions of reaction-diffusion systems ... ruhuna journal of science 4, pp. 1–12, (2009) 11 figure 4 time evolution of v : density plots of v at different time levels according to these density plots it can be seen that these solutions form some spatial patterns. 5. discussion the constructed implicit finite difference scheme can be directly applied to solve the transformed (system with coupled diffusion terms to system without coupled diffusion terms) reaction diffusion system. after that, the solutions of the coupled-reaction diffusion system are obtained by applying inverse of the transformation. lw somathilake: numerical solutions of reaction-diffusion systems ... 12 ruhuna journal of science 4, pp. 1–12, (2009) however when transforming a diffusion-coupled reaction diffusion system to a reactiondiffusion system with uncoupled diffusion terms the reaction terms become more complicated. this may affect to convergence of the finite difference scheme. in order to get rid of this problem step sizes of time have to be shorter. again this may cause to increase computational errors and computational time. it is expected to estimate and compare computational errors and computational time of these two methods in my future work. references badraoui, s. 2002. existence of global solution for systems of reaction-diffusion equations on unbounded domain. electronic journal of differential equations 2002(2002) 1–10. grindrod, p. 1991. patterns and wavesthe theory and applications of reaction-diffusion equations. oxford applied mathematics and computing science series, oxford. hoff, david. 1978. stability and convergence of finite difference methods for systems of nonlinear reaction-diffusioon equations. siam j. numer. anal. 15 1161–1177. kirane, m. 1(1993). global pointwise a priori bounds and large time behavior for a nonlinear system describing the spread of infectious disease. applicationes mathematicae 1–9. kirane, m. 1989. global bounds and asymptotics for a system of reaction-diffusion equations. journal of mathematical analysis and applications 138 328–342. murray, j.d. 2003. mathematical biology: spatial models and biomedical applications, vol. ii. springer-verlag berlin heidelberg. p. collet, j. xin. 1994. global existence and large time asymptotic bounds of l∞ solutions of thermal diffusive combustion systems on rn. arxiv:chao-dyn/9412003 5. seaid, m. 2001. implicit-explicit approach for coupled systems of nonlineat reaction-diffusion equations in pattern formation. science letters 3. turing, a. m. 1952. the chemical basis of morphogenesis. philosophical transactions of the royal society of london 237 37–72. turk, g. 1992. texturing surfaces using reaction-diffusion. phd theses, university of north carolina, chapel hill. webpage. ????a. http://www-swiss.ai.mit.edu/projects/amorphous/grayscott/. webpage. ????b. http://www.scholarpedia.org/article/fitzhugh-nagumo−model. sv-lncs ruhuna journal of science vol. 5: 7-15, 2014 http://rjs.ruh.ac.lk/ issn: 1800-279x 7  faculty of science, university of ruhuna preliminary phytochemical studies in the leaf, stem and root extracts of the traditional medicinal plant species, thalictrum javanicum blume abinaya gurunathan and paulsamy subramanium* department of botany, kongunadu arts and science college, coimbatore, india *correspondence: paulsami@yahoo.com received: 28 april 2014, revised version accepted: 10 october 2014 abstract. the leaf, stem and root parts of the traditional medicinal plant species, thalictrum javanicum are effectively used against the disorder haematuria and the roots are prescribed as diuretic, purgative and tonic by the local healers of high hills of nilgiris, the western ghats, india. little works on the analysis of certain secondary metabolites of antioxidant properties have only been made in this species. hence, the present study was aimed at determining the available major secondary metabolites in the study species, t. javanicum by preliminary analysis including tlc. the results of the study show the presence of many major secondary metabolites viz., tannins, saponins, resins, flavonoids, alkaloids, glycosides, terpenoids, cardiac glycosides and triterpenoids in all the three plant parts studied. a comparison of extracts prepared with petroleum ether, chloroform, methanol and water showed that the methanol extract contained a greater number of secondary metabolites than the other extracts. the mobile phases for the separation of various secondary metabolites were also standardized. the presence of different varieties of secondary metabolites of medicinal importance helps to rationalize the use of t. javanicum for therapeutic purposes by local healers of high hills of nilgiris. keywords. secondary metabolites, thalictrum javanicum, tlc gurunathan and subramanium phytochemical studies of t. javanicum ruhuna journal of science (december 2014) 8 1 introduction thalictrum genus that belongs to the family, ranunculaceae is an extremely important medicinal plant source, and more than 200 species are available worldwide. thalictrum plants are rich in benzylisoquinoline derived alkaloids; atleast 250compounds of this category have been isolated from 60 species and most of them with strong biological activities. extracts and alkaloid isomers from thalictrum are known to exhibit various pharmacological activities, including antitumor, antimicrobial, antiamoebic and hiv antiviral activities (chen et al. 2003). thalictrum species are mainly growing in the temperate and cold zones of both hemispheres. due to the temperature stress, almost all species of this genus are reported to have rich variety of secondary metabolites particularly the quartenary alkaloids like oxyberberine, thalrugosaminine, rugosinone, thalisopine, berberine, thalifendine, palmatine, berberine, columbanine, jatrorrhizine, magnoflorine and demethyleneberberine (bahadur and shukla 1983; sahei et al. 1985; khamidullinia et al. 2006). t. javanicum, a rare sighted traditional medicinal plant species inhabiting the high hills of nilgiris (ca. 2400 m above msl, the western ghats, india) is being used by the rural communities against certain specific disorders like haematuria (tiwari et al., 2006), and the roots are used as diuretic, purgative and tonic (anonymous 1976; sharma 2009). the parts below ground are used for extracting dye (paulsamy 2009). as the phytochemicals are important factors for healing property of any species, study of their phytochemical profile is most essential. however, little information is available on the major phytochemical constituents of this species (abinaya et al. 2013) therefore, to address this lacuna, a preliminary phytochemical investigation of this plant was carried out. 2 materials and methods the leaf, stem and root of t. javanicum were collected from the forest margins in high hills of nilgiris were shade dried, powdered and extracted separately with the solvents, petroleum ether, chloroform, methanol and water for further studies. gurunathan and subramanium phytochemical studies of t. javanicum ruhuna journal of science (december 2014) 9 2.1 preliminary phytochemical analysis fifty (g) of leaf, stem and root powder were added separately into 300 ml of the solvents viz., petroleum ether, chloroform, methanol and water, and kept in soxhlet apparatus for hot extraction for 16 hours. the extracts were filtered separately and stored at 4˚c. the presence of a variety of phytochemical compounds in the leaf, stem and root extracts of different solvents was determined by using the methods of brindha et al. (1977) and harbone (1988). 2.2 thin layer chromatography the tlc studies were performed by following the method of wagner et al. (1984) by using silica gel-g thin layer chromatographic plate of 15x5 cm size with 3 mm thickness. for the separation of phytochemical compounds, only the methanolic extracts of leaf, stem and root of t. javanicum were spotted manually using capillary tube. the spotted plates were put into a solvent system to identify the suitable mobile phase as per the method of wagner et al. (1984). after the separation of phytochemicals, the spray reagents such as dragondorff’s reagent, vanillin hydrochloric acid and vanillin-phosphoric acid reagent were used to spot the compound. the colour of the spots was noted and the retention factor (rf) values were calculated by using the following formula: distance travelled by the respective solvent extract retention factor (rf) = distance travelled by the solvent 3 results and discussion 3.1 phytochemical screening the results of the study showed that depending on the polarity of the solvents, there was wide variation in the presence of the type of secondary metabolites observed (table 1). it has been observed that in comparison to other solvent extracts, methanol extracts of all the three parts of t. javanicum contained a greater variety of secondary metabolites viz., tannins, saponins, alkaloids, steroids, phenols and triterpenoids with higher quantity in leaf than that of the other parts (eloff 1998). gurunathan and subramanium phytochemical studies of t. javanicum ruhuna journal of science (december 2014) 10 table 1. preliminary phytochemical analysis of various extracts of leaf, stem and root parts of thalictrum javanicum. *pe-petroleum ether, chchloroform, memethanol, wa-water. the methanolic extracts of other two parts analysed were also contained many types of secondary metabolites (resins, glycosides, triterpenoids and tannins in stem and glycosides, steroids and terpenoids in root). on the other hand, the petroleum ether extracts of all the three parts contained a lesser number of major secondary metabolites. this difference may be due to the relatively high polarity of methanol compared to the other solvents use (eloff 1998). in addition, it has been noted that alkaloids were generally extracted by almost all solvents used. the species with high content of alkaloids are generally considered to have effective healing properties like diuretic, antispamodic, antiseptic and antidote actions (ming zhao et al. 2010; gonzales et al. 2014; melendez-camargo et al. 2014). 3.2 thin layer chromatography (tlc) many solvents and solvent mixtures were tried for the separation of alkaloids, flavonoids and terpenoids in the leaf, stem and root parts of the study species, t. javanicum. the results of the study revealed that the combination of relatively high polarity solvents viz., chloroform with methanol and water (chcl3:ch3oh:h2o) in different proportions was the most suitable solvent system for the separation of certain secondary metabolites like alkaloids secondary metabolite leaf stem root solvent* pe ch me wa pe ch me wa pe ch me wa tannins +++ +++ ++ +++ saponins +++ ++ + +++ + resins + + +++ +++ ++ flavonoids +++ +++ +++ +++ + + alkaloids +++ +++ +++ +++ +++ ++ ++ +++ +++ glycosides + + + +++ steroids + +++ + +++ +++ phenols ++ + terpenoids +++ +++ + + + cardiac glycosides +++ ++ +++ ++ +++ + triterpenoids ++ +++ + +++ gurunathan and subramanium phytochemical studies of t. javanicum ruhuna journal of science (december 2014) 11 (figure 1), flavonoids (figure 2) and terpenoids (figure 3). with respect to separation of alkaloids, in methanolic leaf, stem and root extracts, the solvent mixture chcl3:ch3oh:h2o in the ratio of 1:0.8:0.4, 1:0.7:0.2 and 1:0.6:0.2 respectively was the most suitable. fig 1. tlc for detection of alkaloids in methanolic leaf, stem and root extracts of thalictrum javanicum (ratio of chloroform: methanol: water for leaf 1:0.8:0.4, for stem 1:0.7:0.2, and for root 1:0.6:0.2). fig 2. tlc for detection of flavonoids in methanolic leaf, stem and root extracts of thalictrum javanicum (ratio of chloroform: methanol: water for leaf 1:0.6:0.3, for stem 1:0.5:0.3, and for root 1:0.5:0.4). gurunathan and subramanium phytochemical studies of t. javanicum ruhuna journal of science (december 2014) 12 fig 3. tlc for detection of terpinoids in methanolic leaf, stem and root extracts of thalictrum javanicum (ratio of chloroform: methanol: water for leaf 1:0.7:0.2, for stem 1:0.7:0.1, and for root 1.6:0.9:0.5). for the separation of flavonoids in the leaf, stem and root, the solvent system that was the most suitable mobile phase contained chcl3:ch3oh:h2o in the ratio of 1:0.6:0.3, 1:0.5:0.3 and 1:0.5:0.4 respectively. for the terpenoid separation, above solvents in the ratio of 1:0.7:0.2, 1:0.7:0.1 and 1.6:0.9:0.5 respectively was the most suitable. results indicate that the alcoholic solvents are most effective as mobile phases for the separation of major secondary metabolites in the leaf, stem and root of the study species, t. javanicum. the tlc studies confirmed that the methanolic extracts of the leaf, stem and root parts of t. javanicum contained the most important secondary metabolites of medicinal importance viz., alkaloids, flavonoids and terpenoids (table 2). according to allen and miller (1996), the species with more content of secondary metabolites such as alkaloids, flavonoids and triterpenoids naturally have antibacterial, anti-inflammatory, anti-allergic, antiviral, antineoplastic, antithrombotic, antioxidant, and vasotilatory properties etc. the presence of these phytochemical compounds help to support the traditional knowledge on medicinal usage of the species, t. javanicum by the local healers of high ranges of nilgiris, the western ghats, for the treatment of various ailments. the results also support as well as recent findings pertaining to medicinal properties of this plant species by other investigators such as abinaya et al. (2013, 2014). gurunathan and subramanium phytochemical studies of t. javanicum ruhuna journal of science (december 2014) 13 table 2. preliminary separation of phytochemicals in methanolic extracts of the leaf, stem and root parts of thalictrum javanicum. plant part mobile phase chcl3:ch3oh:h2o spray reagent colour of the spot/band rf value compound leaf 1: 0.8: 0.4 dragondorff’s reagent orange 0.57 alkaloids stem 1: 0.7: 0.2 dragondorff’s reagent orange 0.85 alkaloids root 1: 0.6: 0.2 dragondorff’s reagent orange 0.78 alkaloids leaf 1: 0.6: 0.3 vanillin hydrochloric acid yellowish spot 0.57 flavonoids stem 1: 0.5: 0.3 vanillin hydrochloric acid yellowish spot 0.28 flavanoids root 1: 0.5: 0.4 vanillin hydrochloric acid yellowish spot 0.43 flavonoids leaf 1: 0.7: 0.2 vanillin-phosphoric acid reagent (vpa) blue 0.90 terpenoids stem 1: 0.7: 0.1 vanillin-phosphoric acid reagent (vpa) blue 0.57 terpenoids root 1.6: 0.9: 0.5 vanillin-phosphoric acid reagent (vpa) blue 0.71 terpenoids 4 conclusion the prelimnary phytochemical analysis showed that the traditional medicinal plant species, t. javanicum inhabiting the high ranges of nilgiris, the western ghats, india contained rich variety of secondary metabolites of medicinal importance, thus which rationalizing the traditional uses of this plant species by local healers for a variety of medicinal purposes. however, further in depth pharmaceutical and pharmacological studies need to be carried out to confirm the therapeutic efficacy of this plant in the disease conditions for which local medical practitioners prescribe it. gurunathan and subramanium phytochemical studies of t. javanicum ruhuna journal of science (december 2014) 14 acknowledgement the first author is graciously acknowledging tamil nadu state counsel for science and technology, chennai for providing financial support to carry out the work (rc. no. 28696/k2/2012, dated: 19.11.2013). references anonymous 1976. the wealth of india, raw materials (vol. x, sp-w). publications and information directorate, csir, new delhi, 201-207. abinaya gurunathan, paulsamy subramaniam and saradha maran. 2013. antioxidant potential of leaf, stem and root extracts of thalictrum javanicum blume international journal of biological & pharmaceutical research. 4(12): 1217-1221. abinaya gurunathan, paulsamy subramaniam and saradha maran. 2014. quantification of certain secondary metabolites and evaluation of antioxidant properties in the traditional medicinal plant, thalictrum javanicum blume. world journal of pharmacy and pharmaceutical sciences 3(6): 1856-1873. allen l and miller n.d. 1996. antioxidant flavonoids: function and clinical usage . alternative medicine review, 1(2): 109-111. bahadur s and shuklaak, 1983. studies on medicinal plants i, the quartenary alkaloids of thalictrumjavanicum. journal of natural products. 1983; 46(4): 454-457. brindha pk, sasikala and purushoth k. 1977. prelimnary phytochemical studies in higher plants. ethnobotany3.84-96. chen, s., chen,s and xiao p. 2003. ethnopharmacological investigations on thalictrumplants in china. journal of asian natural product.research , 5(4); 263-271. eloff jnc. 1998. which extractant should be used for the screening and isolation of antimicrobial components from plants? journal. ethanopharmacology; 60:1-8. gonzales, maria victoria m., tolentino and angelina g. 2014 extraction and isolation of the alkaloids from the samanea saman (acacia) bark: its antiseptic potential international journal of scientific & technology research 3(1): 119-124.harbone jb. 1988. phytochemical methods. a guide to mordern technique of plant analysis (3 rd edn) chapman and hall, london.1-138. khamidullinia, e.a., gromova, a.s.,and lutsky vi owen nl. 2006. natural products from medicinal plants: non alkaloidal natural constituents of the thalictrum species. natural product reports. 23: 117-129. melendez-camargo me, contreras-león i and silva-torres r. 2014 diuretic effect of alkaloids fraction extracted from selaginella lepidophylla (hook. et grev.) spring boletín latinoamericano y del caribe de plantas medicinales y aromáticas 13 (1): 92 – 99. ming zhao,yan-fang xian, siu-po ip, harry hs fong and chun-tao che. 2010 a new and weakly antispasmodic protoberberine alkaloid from rhizoma coptidis phytotherapy research 24 (9): 1414–1416. paulsamy s. 2009. annual progress report, ministry of environment and forests, government of india, new delhi sponsored project, “evaluation of conservation strategies for the sustainable utilization of herbaceous bioresources in the sholas of nilgiris, the western ghats” (file no. 08/ 16/03.cs /br) . gurunathan and subramanium phytochemical studies of t. javanicum ruhuna journal of science (december 2014) 15 sahei, m., sinha, sc., ray,a b., chahopadhyay,s k., khalil, s a., slaticin,j d and schiff,p lj. 1985. addition alkaloids of thalictrumjavanicum. journal of natural products; 48(4): 669-669. sharma, o. p. 2009.plant taxonomy.2nd edition. tata mcgraw hill education private limited, new delhi. 2009; 564. tiwari, lalit and pande p. c. 2006. indigenous vetinary practices of darma valley of pithoragarh district, uttaranchal. indian journal of traditional knowledge, 5(2): 201-206. wagner hm, bladt s, zgainski em. 1998. plant drug analysis-a thin layer chromatography atlasspringer verlog berlin.1984.  © 2009 faculty of science university of ruhuna ruhuna journal of science vol. 4, september 2009, pp 21-27 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x abstract. due to the predicted threats of global warming and sea level rise, the salt tolerance and salt accumulative abilities of plants have become popular contentious topics. mangroves are one of the major groups of salt tolerant plants and several mechanisms are known as instrumental in their salt tolerance. salt excretion through leaf drop is given as one, but its validity is questioned by some recent works compelling the necessity for further studies. knowledge of the salt contents in different mangrove plants is a pre requisite for such studies. hence, this study aimed to quantify and compare the salt content in mature leaves of nine mangrove species in sri lanka., i.e. aegiceras corniculatum, avicennia marina, avicennia officinalis, bruguiera gymnorrhiza, bruguiera sexangula, ceriops tagal, excoecaria agallocha, lumnitzera racemosa, rhizophora apiculata and rhizophora mucronata which are growing in the same mangrove system; the rekawa lagoon in sri lanka. two species of non mangrove plants, gliricidia sepium and artocarpus heterophyllus, which were growing in inland areas were also selected for comparison. the concentration of na+ in leaves was considered as a measure of the salt concentration. the na+ in leaves was extracted by acid digestion and quantified by flame photometry. the salt content of mangroves was measured under two contrasting hydrological situations: at the highest and lowest water levels of the lagoon. rekawa lagoon can be considered as a ‘barrier built estuary’, the highest water level occurs when the lagoon mouth is blocked due to the formation of the sand bar and the water level is increased by fresh water inflow, inundating the total mangrove area and decreasing the soil/water salinity. the water level of the lagoon becomes lowest when the lagoon mouth is opened (naturally or by dredging) and lagoon water is flushed out to the sea. then the salinity of lagoon water becomes high due to sea water influx. the results showed that the concentration of na + in mangrove leaves was 3 to 12 times higher compared to that in leaves of selected non mangroves. statistical analysis revealed that the variations in na+ content in leaves of different mangrove species were same under both hydrological regimes. e. agallocha and r. mucronata showed the highest salt content whilst a. corniculatum b. sexangula showed the lowest salt content. the three species, a. marina, a. officinalis and l. racemosa, showed the second highest salt content and the remaining two species c. tagal and b. gymnorhiza, showed the second lowest salt content. apparently the interspecific variation in the concentration of na+ in mangrove leaves follow the interspecific variations in the salinity tolerance reported for the same species. key words; mangroves, salt accumulation, na+ in plants 21 assessment and comparison of salt content in mangrove plants in sri lanka n. p. dissanayake and k. m. c. amarasena department of botany, university of ruhuna, matara http://www.ruh.ac.lk/rjs/rjs.html dissanayake and amarasena ruhuna journal of science,4,pp21-27 (2009) assesment and comparison of salt content.... introduction mangroves are woody plants which are adapted to grow on muddy or sandy intertidal areas of lagoons, estuaries and sheltered bays in tropical and some sub tropical areas. more than 60 species of true or exclusive mangrove species are recorded in the world (kathiresan and bingham, 2001). according to jayatissa et al. (2002), 21 species of true mangroves and more than 18 species of mangrove associates occur in mangrove communities of sri lanka. mangroves are important to sri lanka as about one third of the country’s coastline in its original state has the potential to support mangroves and hence to provide ‘green barriers’ against ocean surges such as tsunami and cyclones (jayatissa et al., 2005). as a result of predicted global climate change and sea level rise, many of the low-lying landward areas could be added newly to the interface between the sea and the land. with this gradual change, the vegetation of those areas could also be shifted gradually from normal plants to salt tolerant plants. as salt tolerant plants mangroves can grow in such harsh conditions and may become dominant plant type in such areas. as a preparatory mission for the predicted sea level rise, mechanisms behind the capacity of mangrove plants to tolerate saline conditions should be understood well and attempts should be made to give such abilities to other important plants also through advance techniques. it will allow getting the maximum use from the limited extent of arable lands in the future. the salt content of sea water is dominated by high concentrations of na + and cl . however k+, mg ++, ca++ and so4- are also in significant concentrations (läuchli and lüttge, 2002).. mangroves are adapted to overcome high sediment salinities through several mechanisms. salt exclusion, salt secretion and salt accumulation are agreed as the three processes involved in that adaptability and mangroves appear to use a combination of the above processes to avoid heavy salt load (kathiresan and bingham, 2001). although there are some studies on the salt secretion and salt exclusion of mangroves indicating the salt concentration in each species (john et al., (2002), most of the studies had been done using few species. hence, a comparison of the salt content among many mangrove species cannot be done. therefore it is important to conduct studies to fill this gap and enrich the knowledge.the prime objective of this project was to determine and compare the salt contents in leaves of mangrove and some non mangrove plant species. materials and methods the nine true mangrove species selected from rekawa lagoon were aegiceras corniculatum, avicennia officinals, avicennia marina, bruguiera gymnorhiza, bruguiera sexangula, ceriops tagal, excoecarea agallocha, lumnitzera racemosa, and rhizophora mucronata, and the non halophytic species selected from inland areas were gliricidia sepium and artocarpus heterophyllus. three mature individuals of each mangrove species from different places of the lagoon were selected for sampling and considered as three replicates of the same treatment. when the water level of the lagoon is high, multiple leaf samples with mature leaves were 22 ruhuna journal of science,4, pp21-27 (2009) dissanayake and amarasena assesment and comparison of salt.... collected randomly from each individual by removing each leaf at their abscission zone. samples of the two non halophytic species selected from inland area were also collected following the same procedure. all samples were collected at the same time between 9.00 – 11.00 am on the same day with the help of several trained people and collected leaf samples were placed in separate plastic bags to transfer immediately to the laboratory. the same procedure was followed to collect samples from the same mangrove and non mangrove individuals when the water level of the lagoon was very low. as a result of sea spray and salt secretion in salt secreting species, there may be some salt deposited on the surface of leaves in the samples. therefore, in the laboratory, all the leaves were washed with de-ionized water to remove surface salts and other surface contaminants. then they were gently dried with absorbent papers and immediately transferred to the drying oven. the leaves were kept in the drying oven at 60 0c until they were completely dried. one gram of dried leaves of each sample was digested in 2 ml of concentrated hno 3. then hno3 acid was evaporated and the residue was dissolved in deionized water to get 25 ml of aquas salt extraction. the salt contents in extractions were analyzed by flame photometry. the soil salinity of the mangrove area at each sampling occasion was measured by using pore water. pores in the soil were made at the area just beyond the high water mark to collect water samples and salinity was measured using a refractometer. means and standard deviations of na+ content in leaves were calculated and one way anova test was performed to test significant interspecific differences in na + content of mangrove leaves. turkey’s post-hoc tests were used to have pair wise differences between species. all statistical analysis was carried out using the standard statistical software jmpin (3.2.6 version). results the sodicity or concentration of na+ in leaf extractions of the two species of nonhalophytes used in this study, i.e. gliricidia sepium and artocarpus heterophyllus, was less than 0.1375 mmol/g dry weight, whilst that in the 09 mangrove species varied from 0.3334 ±0.168 to 1.8341± 0.114 mmol/g dry weight. the results of the one way anova that was carried out for na + concentrations in leaf extractions at the high water level and low water level separately, revealed that the sodicity among the leaf extractions of different mangrove species varied in the same way under the both water regimes (figure 1a and 1b). the sodicity (ie. concentration of na+), of the nine mangrove species falls into four categories as highest, higher, lower, and lowest. e. agallocha and r. mucronata showed the highest salt content (1.79-1.83 mmol/g dry weight). the three species, a. marina, a. officinalis and l. racemosa, showed the next highest salt contents (1.44 1.20 mmol/g dry weight), c. tagal and b. gymnorhiza, showed lower salt contents (0.78 1.07 mmol/g dry weight) whilst a. corniculatum b. sexangula showed the lowest salt contents (0.33 0.34 mmol/g dry weight). 23 dissanayake and amarasena ruhuna journal of science,4,pp21-27 (2009) assesment and comparison of salt content.... 1a. 1b. na + amount 0 0.5 1 1.5 2 2.5 1 2 3 4 5 6 7 8 9 10 11 species n a + a m o u n t (m m o l/ g d ry w t) 24 na + amount 0 0.5 1 1.5 2 2.5 1 2 3 4 5 6 7 8 9 10 11 s pecies n a + a m o u n t (m m o l/ g d ry w t) a a b b b d c c d a d d c c b b b a a b ruhuna journal of science,4, pp21-27 (2009) dissanayake and amarasena assesment and comparison of salt.... figure 1. means and standard deviations of the content of na+ in leaves of diferent mangrove plants in rekawa lagoon, sri lanka at the high water level (1a) and at the low water level (1b). (same superscripts indicate species which are not significantly different in their na+ content.) 1-aegiceras corniculatum, 2avicennia marina, 3avicennia officinalis, 4-bruguiera gymnorrhiza, 5bruguiera sexangula, 6ceriops tagal, 7 excoecaria agallocha, 8lumnitzera racemosa, 9rhizophora mucronata, 10gliricidia sepium, 11artocarpus heterophyllus. the mean soil salinity of the mangrove area of rekawa lagoon at the high water level and low water level were 4.2 ± 0.4 and 7.4 ± 0.6 respectively. discussion the salinity that gives the salt content in a solution means the concentration of dissolved mineral salts present in the solution on a unit or weight basis. the major solutes comprising dissolved mineral salts in plant materials are the cations sodium (na +) potassium (k+), calcium (ca2+) and magnesium (mg2+), and the anions chloride (cl-), sulfate (so42-), bicarbonate (hco3 ), carbonate (co32-) and nitrate (no3-) (taiz and zeiger, 1991). however in halophytes, the content of na+ and clis comparatively very high as those ions inevitably enter into plants when they grow in saline soils (cram et al, 2002). moreover, under the situation with excess and very high concentration of these salt ions, the salinity (i.e. concentration of dissolved mineral salts), sodicity (ie. concentration of na+), and the chlorinity (i.e. concentration of cl-) in plant tissues could be in more or less equal molarities. therefore, the concentration of na + or clresulted in the analysis of plant tissues in this study is considered as an alternative for the concentration of salts in the relevant tissue. in this study, only two species of non-halophytic mesophytes were selected to measure and compare the salt content in leaves. it was considered that measurements from many species are not necessary as the fact that non halophytic plants posses very low salt concentrations (bowman and strain, 1986). the results of this study also indicated that the salt content in leaves of the non-halophytic species tested is very low and comparable with values given for other mesophytic species (bowman and strain, 1986). that comparability indicates that the measurements received for the na + concentration in mangrove leaves also in this study are reliable. the results of this study indicate that the salt concentration in different mangrove species vary remarkably and the pattern of those variations is almost same at the high water level as well as low water level of the lagoon. (many of the lagoons in sri lanka including rekawa lagoon, are actually ‘barrier built estuaries’ in which, the highest water level occurs by the fresh water inflow, when the lagoon mouth is blocked due to the formation of the sand bar. this highest water level could remain maximum for few weeks and during that time the whole mangrove area remains inundated by low saline water. then 25 dissanayake and amarasena ruhuna journal of science,4,pp21-27 (2009) assesment and comparison of salt content.... the lagoon mouth is naturally or cut open and lagoon water flushed out to the sea, reducing the water level and emerging the mangrove area. this lowest water level could keep for longer period particularly in drought periods and the salinity of lagoon water can be increased due to sea water influx through the lagoon mouth.) in other words, the pattern of variations in the salt content of mangroves appears to be independent from the soil salinity. however, the anaerobic condition of the soil in the intertidal area could be remarkably different during two occasions as the intertidal area was completely inundated at the highest water level and completely emerged and open to air at the lowest water level. hence, the results of this study indicate that anaerobic or aerobic condition of the soil has no much effect on the salt accumulation in mangrove leaves. these results also indicate that the na+ concentration in different mangrove species vary remarkably and those variations of the na+ concentration among different mangrove species are parallel to their salt tolerance reported by other studies. the salt tolerance of a plant can be defined as the plant’s capacity to endure the effects of excess salt in the medium of root growth (taiz and zeiger, 1991). the two species which showed the highest na+ concentration in this study, r. mucronata and e. agalocha, are reported as species which can perform better under high levels of soil salinity (jayatissa et al, 2008., kodikara, 2009). on the other hand, the salt tolerance of the two species, a. corniculatum and b. sexanguila, which showed the lowest na+ concentration, are reported as low saline species (jayatissa et al., 2008., kodikara, 2009). the remaining species i.e. c.tagal, b.gymnorrhiza, a.officinalis, a.marina, and l. racemosa showed intermediate levels in na+ as well as in salinity tolerance (jayatissa et al., 2008). this study may help to identify the species which can cope up with a high salt concentration and to calculate how much salt can be stored in the biomass of a mangrove forest. furthermore, this work and findings could be a motivation to study the capacity of different mangroves to excrete salts via leaf drop. references bowman, w.d. and strain, b.r. (1986). interaction between co2 enrichment and salinity stress in the c4 non-halophyte andropogon glomeratus (walter) bsp. plant, cell and environment. 10: 267270. cram, w.j., torr, p.g. and rose, d.a. (2002). salt allocation during leaf development and leaf fall in mangroves. trees, 16. pp 112-119. jayatissa, l. p., dahdouh-guebas, f. and koedam, n. (2002) a revision of the floral composition and distribution of mangroves in sri lanka. botanical journal of the linnaean society. 138, 29-43 jayatissa, l.p., lo seen, d., hettiarachi, s. and senanayake, g. (2005). nature’s protection against nature’s fury: a post tsunami assessment of the importance of 26 ruhuna journal of science,4, pp21-27 (2009) dissanayake and amarasena assesment and comparison of salt.... mangroves as a natural barrier against the wrath of the sea. proceedings of annual science symposium, faculty of science, university of ruhuna, matara, sri lanka. kathiresan1 k. and bingham2 b.l. (2001). biology of mangroves and mangrove ecosystem. advances in marine biology. 40: 81-251. kodikara, s. (2009). mangrove seedlings perform better in moderate salinity than high and low salinities. thesis submitted as a partial fulfilment for the b. sc. (special) degree in botany. department of botany, university of ruhuna. sri lanka. pp teiz, l. and zeiger, e. (1991). plant physiology. the benjamin/cumming publishing company, inc., uk., pp.565. 27 results discussion sv-lncs ruhuna journal of science vol 9: 13-31, june 2018 eissn: 2536-8400  faculty of science http://doi.org/10.4038/rjs.v9i1.38 university of ruhuna  faculty of science, university of ruhuna 13 sri lanka structural and functional responses of xylem in rhizophora mucronata lam. seedlings under drought and hypersaline conditions n.p. dissanayake, k.a.s. kodikara*, s. premachandra and l.p. jayatissa department of botany, faculty of science, university of ruhuna, matara, sri lanka correspondence: * sunandaruh@gmail.com ; orcid: 0000-0002-2493-3580 received: 5th april 2018, revised: 28th june 2018, accepted: 30th june 2018 abstract. water translocation in mangrove seedlings is often affected by water stress conditions such as drought, hyper-salinities and their frequent variations. this study was therefore aimed at studying the wood anatomical responses of xylem tissue and hydraulic conductivity of rhizophora mucronata lam., a common species in mangrove planting, under different levels of drought [25%, ~50% and ~100% of water holding capacity (whc)] and soil salinity [high salinity (35 psu), moderate salinity (15 psu) and freshwater (0 psu)]. as wood anatomical responses, significantly higher vessel density, vessel grouping (p<0.001) along with narrow vessel elements (p<0.001) were observed in plants grown in the 25% and 50% whcs and high salinity treatments. all these anatomical responses are more directed towards avoidance of vessel cavitation which is commonly found under water deficit conditions. the results showed that r. mucronata plants failed to maintain efficient transportation of water when the field capacity was 50% of whc or lower and the level of salinity was 35 psu or greater, as evident by the reduction of water conductive areas, vessel areas and hydraulic conductivity (p<0.05). overall, water use efficiency of r. mucronata seedlings under the imposed water stress conditions has remarkably reduced and it further indicated that such imposed stress conditions directly affect the survival of planted seedlings as depicted by the significantly low survival in 25% and 50% of whcs and high salinity. therefore, in-depth study on lagoon hydrology including inundation levels, water depth, salinity and the selection of correct tidal positioning is highly recommended as prerequisites in mangrove planting. keywords. hydraulic architecture, hydraulic conductivity, mangroves, restoration, water stress. 1 introduction mangrove forests are unique plant communities that grow in extreme environmental conditions such as high and changing salinity, frequent mailto:sunandaruh@gmail.com n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science 14 vol 9: 13-31, june 2018 inundation with associated hypoxia, low air humidity and high temperatures. mangrove ecosystems are restricted to intertidal areas of lagoons, estuaries, and sheltered bays in tropical and sub-tropical areas worldwide (tomlinson 1986; spalding et al. 2010; mukherjee et al. 2015). however, the destruction of mangrove ecosystems is increased at an alarming rate and hence, attempts for mangrove replanting are taking place as a common activity in such areas. mangrove planting in sri lanka has received a higher attention particularly after realizing the unprecedented mangrove loss in previous decades (polidoro et al. 2010; richards and friess 2016) and safeguarding action of mangroves against the tsunami in 2004 (dahdouh‐guebas et al. 2005b; lee et al. 2014; jayatissa et al. 2016; satyanarayana et al. 2017). however, majority of the planting attempts showed higher failure rates (ahmad 2012; kodikara et al. 2017a) as they were not supported by sound scientific knowledge. among the relevant causes for the failures, improper site selection without studying site history and planting inappropriate species have been identified as two major causes for the failures (primavera and esteban 2008; lewis 2009; kathiresan 2011; ahmad 2012; kodikara et al. 2017a). in many cases, mangrove planting practitioners tend to establish mangrove plantations out of intertidal range (i.e. natural mangrove growing area) of lagoon which is commonly known as “inappropriate tidal positioning” in mangrove planting (samson and rollon 2008; brown et al. 2014a). this often brings about inappropriate ecological conditions, for example hard soil, low water content, high salinity and hypoxia which are detrimental for the survival of mangrove plants (brown et al. 2014a). therefore, planted mangrove propagules/seedlings are vulnerable for drought stress (also known as physical water stress) and salt stress (also known as physiological water stress), in case of planting beyond intertidal zone (supralittoral) where soil water content is low (particularly in dry zone and dry season) and more deeper areas in lagoon water (infra-littoral) where high salinity (during dry season) and hypoxic (prolonged submergence) conditions are applied, respectively (hoppe-speer et al. 2011). however, in general, mangrove plants have robust adaptions to overcome prevailing extreme conditions (tomlinson 1994; shi et al. 2005) and it is therefore important to study their plasticity of tolerance under these created extremities. under the above stressed conditions, the water movement could be negatively affected which results in numerous variations in vessel anatomy as adaptations to ensure safe water movement (hacke et al. 2006). in case of water limiting conditions, air bubbles could be formed inside vessel elements (cavitation) which breaks the continuity of water column (cochard 2006; robert et al. 2009) disturbing the movement of water. in response to these conditions, anatomical adaptations like high vessel density, high vessel grouping index, small and short vessels have been recorded for some n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science vol 9: 13-31, june 2018 15 mesophytes (baas et al. 1983). however, such studies for mangroves are scarce. schmitz et al. (2006) and robert et al. (2009) studied the anatomical traits like vessel density, vessel grouping index and vessel size of avicennia marina and rhizophora mucronata in relation to prevailing environmental conditions with particular attention on water relations. several variations in size, density, and grouping of vessels have been observed. nevertheless, the wood anatomical studies, water translocation behavior (functional response) and their plasticity under created abiotic stress conditions such as drought, hypersaline conditions and permanent submergence have not been adequately addressed. further, in most of the studies, plant behavior under such stress conditions, are only discussed in term of vessel anatomy. however, we argue that hydraulic conductivity, i.e. the ease of water moving through a water transporting system, is a more reliable measure which reflects physiological response on the functional behavior of plants under water limiting conditions and it is more trustworthy, when accompanied with compatible anatomical data. thus, the following aspects were addressed in this study, a) anatomical responses of rhizophora mucronata plants under drought and hypersaline conditions, b) variations of hydraulic conductivity (kh) of r. mucronata seedlings under above conditions and c) the plasticity of r. mucronata plants in modifying their vessel anatomy and water translocation behavior to secure their survival under water stress conditions. 2. methodology the research was conducted in a plant-house, in the department of botany, university of ruhuna, matara, sri lanka. rhizophora mucronata lam. was selected for the study as the most commonly used mangrove species in planting programmes (~97%) in sri lanka (kodikara et al. 2017a). 2.1 experimental details in the plant-house experiment, mature propagules of rhizophora mucronata collected from the natural mangrove forest in pambala (07º31´n–79º49´e) in july, 2016 were kept floating in low saline (i.e. 2-3 psu) water for about one month. later, the propagules were transferred from the water-filled containers to the plant-house conditions and maintained in a nursery to be used as the planting material. the soil mixture was prepared using garden top soil, mangrove top soil, sand and compost in the proportion of 1:1:1:1. the pots were prepared using black polythene of 23 cm x 25 cm and filled with about 3.5 kg of the prepared soil mixture. thereafter, seedlings with the first two n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science 16 vol 9: 13-31, june 2018 unfurled leaves (i.e. same vigor) were selected for the experiment. for the two separated experiments (drought stress and salt stress), three treatment levels namely 25% whc (water holding capacity), 50% whc and 100% whc (control) were used for drought stress experiment (physical water stress) while another set of treatments for salt stress experiment (physiological water stress), high salinity (35 psu), moderate salinity (15 psu) and freshwater (0 psu) were used, keeping moderate salinity (15 psu) as the control since it is reported that r. mucronata performs its best under moderate salinity condition (jayatissa et al. 2008). the water holding capacity was calculated based on the volume of the water held in the oven-dried, 100 g of soil sample when 100 ml water is added and the retained water volume was considered as the field capacity (100% whc). half and quarter of the field capacity volume were taken as 50% whc and 25% whc of the soil respectively. the individual pots were treated with the respective water volume daily to keep the imposed stress levels. in addition, several soil samples, taken randomly from the pots, were tested for the water level by using oven-dried method for further confirmation (ncrs, 2010). three replicates for each treatment level were used with 12 seedlings per replicate. a total of 216 seedlings were subjected for the plant-house study. low saline water (i.e. 5 psu) that was prepared separately by mixing sea water and aged tap water (i.e. water kept in open containers for few days before the use in order to remove excess chlorine), was added to maintain the respective whcs and the same procedure was followed in preparation of moderate saline condition. pots with seedlings were placed individually in well-separated tanks according to a completely randomized design (crd) and this was done separately for the two experiments. based on our preliminary measurements, the salinity of water in the tanks was checked once in every two days using a hand refractometer (atago s/mill-e, japan) and adjusted the salinity when necessary (added aged tap water when the level of salinity was higher than the expected and seawater was added when the salinity was lower). commercially available fertilizer was also applied (in the form of pellets) once a month by providing the same amount per pot (adopted from jayatissa et al. 2008; dissanayake et al. 2014; kodikara et al. 2017b). 2.2 data collection data on level of survival of rhizophora mucronata seedlings in each treatment were recorded. as growth parameters, cumulative height (ch) (summation of the height of main stem and lengths of branches) of the r. mucronata seedlings, was measured once a month using a scaled ruler. average number of leaves was also recorded and total average leaf area n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science vol 9: 13-31, june 2018 17 (tala) of the seedlings in all treatments were measured once in a 2-week period by using a millimeter graph sheet. stomatal conductance (sc) was measured with the help of steady state porometer (sc-1; decagon devices, usa). two measurements were taken on the same day as one at the late morning (10:00-11:00 hrs) and the other in early afternoon (13:00-14:00 hrs). the youngest, fully expanded leaf exposed to full sunlight was selected and abaxial conductance was recorded. the same leaf was subjected to measure the stomatal conductance of r. mucronata plants to avoid the variability among the different leaves. fig.1 solitary and grouped vessels in a cross section of rhizophora mucronata stem (x 400). conductive area of a cross section of r. mucronata stem (under low power), considered for the calculation of the hydraulic conductivity vessel density (vd), vessel diameter (vdia), vessel grouping index (vgi), number of solitary vessels (sv) and conductive area (ca) and vessel area (va) were measured as anatomical parameters, by following the procedure given; the stems from the middle part of r. mucronata plants were collected and hand sectioned to obtain transverse sections of the stems. in case of hard stems, (about 4 months after planting), the collected samples were stored in a softening solution i.e. copenhague mixture: 70% ethanol; 28% de-ionised water and 2% glycerol in a 1:1:1 ratio) for about one month and then transverse sections were obtained using a microtome. sections were stained using safranin for 5 minutes and were then dehydrated by transferring through an ethanol series of 50%, 70%, 96% and 100% keeping 3 minutes in each. sections were mounted on glass slides using canada balsam. microscopic images of the sections were taken at different magnifications (25x, 40x, 100x) with an olympus microscope (bx60f-3, tokyo, japan) equipped with a digital camera (ruh-cam, university of ruhuna, sri lanka). vessel dimensions were measured using imagej software, a public solitary vessel fiber cells conductive area grouped vessels cortex region pith region n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science 18 vol 9: 13-31, june 2018 domain, java-based image processing program. afterwards, vdia was directly measured by image software while vd (number of vessels per mm2) and vgi (mean number of vessels per vessel group) and ca were calculated based on the vessel counts and area measurements. percentage conductive area was calculated by taking percentage of total conductive area to the total area of the cross section whereas percentage vessel area was obtained by taking percentage of total vessel area to the total conductive area (fig. 1). hydraulic conductivity of the r. mucronata plants was measured at the 6th month after planting, using manually designed “choatometer” [see fig. 2 adopted from choat et al. 2005]. fig.2 diagrammatic illustration of the manually designed choatometer used in the experiment [adopted from choat et al. 2005] (“h” = height of the pressure head). segments of about 15 cm long and about 1 cm diameter were cut under water, from the stems (at the middle part), for hydraulic measurements according to choat et al. (2005). a seawater solution (1%) filtered with a syringe filter of 0.22 μm, was used as perfusion solution for the experiment. the ends of the branch segments were cut smoothly with a razor blade and the mass of the branch segments i.e. mass start (m0), was recorded, before wrapping the segment ends with parafilm. thereafter, the branch segments were placed in the tubing system, connected at one end to a solution reservoir creating a pressure head (h) and to a pipette, placed on millimeter paper and filled with coloured liquid at the other end. the movement of the liquid in the pipette was followed by a camera taking pictures with a 3 seconds time interval (in total 30 seconds) and small video clips were also kept for further confirmation. after taking ten pictures, the mass of the branch segments was n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science vol 9: 13-31, june 2018 19 weighed again i.e. mass end (m1). furthermore, the length of the branch, the diameter of the branch, the number of nodes per branch segment, the air temperature and the relative humidity during the experiment were also recorded. the bark, pith and xylem area were measured on pictures of a transverse section of the branch segments, using imagej software. hydraulic conductivity (kh) was then calculated using the equation (1). ……………………………………….…… (1) where kh is hydraulic conductivity, f is flow rate, l is branch length, and p is pressure drop. 2.3 statistical analyses cumulative height (ch), total average leaf area (tala), stomatal conductance (sc), hydraulic conductivity (kh), and vessel diameter (vdia) were measured and treated as the continuous variables while vessel density (vd), vessel grouping index (vgi), conductive area (ca), vessel area (va) number of solitary vessels (sv) was taken as the count data (proportional and percentage). mean and standard deviation were calculated as descriptive statistics. three hypotheses, normality (shapiro test; p>0.05), linearity (scatter-plot) and homogeneity of variance (levene’s test; p>0.05) were tested and all conditions were met for the data. parametric tests were, therefore, performed. the mean differences of the aforementioned continuous variables were checked using one-way anova test (at a significance level of p<0.05), taken the “levels of stress” as the “factor” (for the same species rhizophora mucronata) followed by the tukeykramer multiple comparison test to check all pairwise differences. chi-square test was conducted for the count data. all statistical analyses were performed using the r-3.2.2 statistical software. 3 results the level of survival of rhizophora mucronata seedlings decreased over time in all treatments except the control treatments. at the end of 18th week, 61% of r. mucronata plants survived in the 50% whc level while all the seedlings died in the 25% whc level by the end of 16th week (fig. 3). in salt stress experiment, 21% of seedlings out of the initial number was able to survive in the high salinity level and only a few seedlings died in both moderate and p lf kh   pith region n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science 20 vol 9: 13-31, june 2018 freshwater treatments showing no significant difference in level of survival between two treatments (fig. 3). fig. 3 percentage survival of rhizophora mucronata seedlings over the study period. (top) drought stress experiment and (bottom) salt stress experiment. error bars indicate the standard deviation (±). in drought stress experiment, both ch and tala of rhizophora seedlings grown in the 50% whc level was significantly lower (p<0.05) than that of the seedlings in the 100% whc (fig. 4 & fig. 5). interestingly, there was no significant difference in growth performances of the seedlings grown in freshwater and moderate saline condition whilst ch and tala of the r. time (weeks) p e rc e n ta g e l e v e l o f su rv iv a l (% ) 100% whc 50% whc 25% whc moderate salinity freshwater high salinity n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science vol 9: 13-31, june 2018 21 mucronata seedlings in high salinity were remarkably lower (p<0.05) when compared with that in the freshwater and the moderate saline conditions (fig. 4 & 5). the sc also showed noteworthy variation among the treatments. it was significantly lower (p<0.05) in the 50% whc treatment level when compared with that of the control treatment (fig. 6). however, there was no significant difference in sc between the moderate saline and freshwater conditions. it was significantly lower (p<0.001) in the high salinity treatment (fig. 6). fig. 4 cumulative height of rhizophora mucronata seedlings over the study period, in drought stress experiment (top) and salt stress experiment (bottom). error bars indicate the standard deviation (±). 100% whc 50% whc 25% whc moderate salinity freshwater high salinity c u m u la ti v e h e ig h t (c m ) time (weeks) n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science 22 vol 9: 13-31, june 2018 fig.5 total average leaf area (cm2) of rhizophora mucronata seedlings at the end of the study period in (top) drought stress experiment and (bottom) salt stress experiment. vertical bars indicate the standard deviation (±) while significant level by different letters at 95% confidence level. anatomical responses and hydraulic conductivity rhizophora mucronata seedlings showed remarkable variation in their anatomy and kh under the given stress conditions which indicated that all the stress factors tested had a significant effect on vessel anatomy and kh. t o ta l a v e ra g e l e a f a re a ( cm 2 ) x y p p q n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science vol 9: 13-31, june 2018 23 fig.6 stomatal conductance (mmol m-2s-1) of rhizophora mucronata seedlings over the study period in (top) drought stress experiment and (bottom) salt stress experiment. vertical bars indicate the standard deviation (±) while significant level by different letters at 95% confidence level. according to the results, percentage cover of ca of the stems of rhizophora mucronata in both 50% whc and high salinity treatments were significantly lower (p<0.001) when compared with those of the stems, harvested from their respective controls i.e., 100% whc and moderate saline condition (table 1). however, no significant difference in the percentage cover of ca was observed among freshwater and moderate saline conditions. when considering the vessel characters, both vd and vgi were significantly higher x q y p p s to m a ta l co n d u ct a n ce ( m m o l m -2 s1 ) n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science 24 vol 9: 13-31, june 2018 (p<0.001) in the mangrove plants grown in the 50% whc when compared with that of the respective control, 100% whc (table 1). table 1. anatomical responses of vessel elements of rhizophora mucronata seedlings under drought and salt stress conditions (tukey multiple comparison; 95% confidence level, p<0.05; significant differences among treatments within drought stress and salt stress experiments separately for the same parameter are shown by different superscript letters) (n: number of cross sections & n*: number of measurements). parameter drought stress experiment salt stress experiment 100% whc 50% whc 25% whc moderate saline fresh water high saline conductive area (%) (n= 84) 11.7±2.2a 4.1±1.7b a ll p la n ts d ie d a ft e r 1 4 w e e k s 12.9±3.1a 13.1±2.4a 6.1±1.5b vessel area (%) (n=84) 12.8±1.1a 5.2±0.9b 13.8±2.6a 14.4±1.9a 6.7±1.8b vessel density (mm-2 xylem area) (n= x) (n= 84) 25.8±1.6a 66.2± 3.9b 24.5 ±2.9a 25.9± 1.7a 68.2 ±4.1b vessel grouping index (n=84) 2.0 ±0.5a 5.1± 0.6b 2.4± 1.8a 1.8± 1.0a 6.1± 1.4b solitary vessels (%) (n=84) 72.4±4.2a 33.4±3.1b 75.6±1.9a 83.2±2.2a 38.9±4.1b vessel diameter (μm) (n*=1680) 82.6±13.6a 37.2±2.7b 88.7±11.8a 92.1±12.8a 34.8±3.8b in contrast, vdia of r. mucronata plants in the 50% whc showed significant decrease (p<0.05) when compared that of the 100% whc treatment. in salt stress experiment, increased vd and vgi were found in the high salinity treatment (p<0.05) while the vdia was significantly lower (p<0.05) as compared to that of the moderate saline condition. interestingly, vessel anatomy of r. mucronata seedlings grown in the freshwater condition showed the pattern as same as the control plants i.e. low vd and vgi while having larger vdia. percentage of sv elements in conductive area of r. mucronata seedlings grown in the 50% whc was significantly low (p<0.001) when compared with the control. in the same way, percentage of sv in r. mucronata stems grown under the high salinity condition was remarkably low (p<0.001) as compared to the moderate saline condition. n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science vol 9: 13-31, june 2018 25 fig.7 boxplot graphs of hydraulic conductivity (mmol/ms mpa) of rhizophora mucronata plants in (a) drought stress experiment and (b) salt stress experiment. tukey-kramer multiple comparison; 95% confidence level, p<0.05. significance level shows by different letters and were used separately for two treatments. further, we did not observe significant difference in the percentage of sv between the plants grown under the freshwater and moderate saline conditions. in spite of a high variability in vessel anatomy, kh also varied among the treatments. the kh of r. mucronata seedlings grown in the 50% whc level and the high salinity treatment were significantly lower (p<0.05) as compared to their respective controls (fig. 7) i.e. 100% whc and moderate saline conditions respectively. the highest kh was observed in the freshwater treatment which was not remarkably different from the control plants. 4 discussion selection of incorrect tidal positioning in mangrove planting due to ignorance of the accepted technical guidelines such as emr (ecological mangrove restoration) methodology (lewis 2005; lewis and brown 2014) and modified protocols (bosire et al. 2008), often leads to create stress conditions like drought (or negative soil water potential) and hypersaline conditions for planted seedlings (kodikara et al. 2017a). mangrove seedlings have therefore no choice rather than facing to such stress conditions. although, mangrove 100%whc 50%whc 25%whc moderate sal. freshwater high salinity h y d ra u li c co n d u ct iv it y ( m m o l /m s m p a ) x y p p treatments q a b n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science 26 vol 9: 13-31, june 2018 plants are well-armed with various strategies to overcome extreme conditions such as high saline conditions and frequent inundation (alongi 2002), the level of tolerance secondarily depends on the type and intensity of stress (hsiao 1973; larcher 2003). we further emphasize that created stress conditions may have relatively more deleterious effects on planted mangrove seedlings, particularly in the early life stage, than naturally prevailing extreme conditions. under natural conditions, both level of stress and exposure time are relatively low since the effects of stresses are naturally moderated, for example, effect of mangrove soil salinity increase is often diluted with periodic inundation. however, according to the results, survival, growth performances, stomatal conductance, hydraulic traits and hydraulic conductivity of rhizophora mucronata plants were affected in different ways and to different degrees under aforementioned two stress conditions. in this study, the stressed r. mucronata plants in both drought and salt stress conditions showed similar responses in many aspects. the decreased level of survival and growth performances of the mangrove plants grown in the 50% whc, 25% whc and high salinity conditions, could be due to low water potential and osmotic stress (munns and tester 2008). reduced cell turgor pressure is expected under both drought and hypersaline conditions due to which cell expansion is hampered (naidoo 2006). also, accumulation of salt ions like nacl in cell vacuoles further intensifies negative water potential under hypersaline condition (werner and stelzer 1990) where it needs prompt osmoregulation to cope up with such conditions (naidoo 1985). allocation of energy for osmoregulation may be prioritized over plant growth and development which may ultimately restrict energy allocation for growth, showing the reduction of growth parameters such as plant height, leaf production and total leaf area (naidoo 2006; hoppe-speer et al. 2011). when r. mucronata plants run out of metabolic energy in case of exposing to intense stress conditions for a longer period, the plant may end up with early mortality. in that sense, hypersaline condition has made more lethal effects on r. mucronata plants as the level of survival was reduced up to 26% whilst the survival was 61% under drought conditions (50%whc). several anatomical responses could also be observed under the imposed stress conditions, and those responses could be considered as modifications which may be contributory in securing their plant functionality and survival. however, decreased ca in both 50% whc and high salinity treatments is disadvantageous as it reduces water transporting area inside the plants. cell shrinkage which has resulted in due to the reduced cell turgor pressure under drought and hypersaline conditions may be the cause for the reduced ca (akram et al. 2002). the common observations i.e. increased vd and vgi as well as narrow vessel elements of the mangrove plants with the enhanced q n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science vol 9: 13-31, june 2018 27 drought (e.g., 50% whc and 25% whc) and salinity (>35 psu) can be explained in different ways. the formation of vessel conduits with different vessel diameters primarily depends on the differentiation rates of conduits from the cambium. it has showed that negative water potential under water stress conditions caused plants to form small cambial derivates resulting in narrow vessels (arend and fromm 2007). moreover, aloni (2004) has suggested that the formation of narrow vessels under water stress is a result of the elevated auxin concentrations in the plant body which induces the early secondary wall formation and lignification preventing further cell expansion. however, on the other hand, such anatomical modifications, have several associated functional significances. as reported by robert et al. (2009), the described anatomical features like increased vd, vgi and decreased vdia are known to be the adaptations which support the plant to overcome risk of cavitation as air bubbles are easily formed inside the water filled vessel elements under negative water potential obstructing the flow of water. therefore, high vd and vgi under the water stress conditions help to serve more efficient bypass (being closer) of embolized vessels which minimizes the number of dysfunctional vessel elements due to cavitation (schmitz et al. 2006). probably this could be the reason why higher number of solitary vessels was favored in non-stress conditions, for example, 100% whc, freshwater and moderate saline conditions, since risk of cavitation is comparatively low under sufficient water availability. however, the chance for cavitation under negative water potential cannot be totally avoided by the anatomical modifications in conductive tissues (robert et al. 2009). this is clearly evident from the fact that hydraulic conductivity of r. mucronata seedlings grown in the 50% whc and high salinity conditions, has remarkably reduced even when occurring the suitable anatomical modifications in conductive tissue. therefore, we propose that this would happen either due to the presence of a higher number of dysfunctional vessels through which a resistance to water flow is created or any internal adjustment to water flow, done by the plants themselves in case of higher water demand. the reduced hydraulic conductivity could secondarily depend on sc of the plants as sc of the r plants grown in the 50% whc and high salinity was significantly lower while the highest was recorded in the plants grown in the freshwater treatment. several studies reported that high stomatal conductance in r. mucronata was recorded in freshwater condition and it was reduced with the increase of salinity and similar results were recorded for other mangrove species as well (naidoo 1985; aziz and khan 2000; kodikara et al. 2017b). moreover smaller leaf area is also a strategy to reduce water loss from plants under water deficit conditions (ball 1988; brugnoli and lauteri 1991). interestingly, parker and pallardy (1985) have observed a strong positive correlation of leaf area with vessel diameter, stem conductive area and total xylem area for seedlings of several species. we also observed n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science 28 vol 9: 13-31, june 2018 the reduced ca in both experiments of our study. however, this indicates that water use efficiency (wue) decreases with the increased negative water potential under substrate drought and hypersaline conditions. according to the observations, when the rhizophora mucronata plants face water stress conditions (substrate drought or hypersaline condition), both anatomy of the xylem tissues and the hydraulic conductivity were modified as a response to stress. further, the plants make anatomical modifications such as increased vessel density, vessel grouping index with narrow vessel elements which are more directed toward the avoidance of vessel cavitation. parallel to these anatomical changes, stomatal conductance and leaf area also decreased, most probably as a measure of water conservation under water stress conditions. the trade-off between the energy allocation for the above tolerant mechanisms and plant growth is achieved through retarded growth of r. mucronata plants as evident by the plants grown in the 50% whc, 25% whc and high salinity treatments. at extreme level, r. mucronata plants ended up in early mortality (e.g. 25% whc), more likely due to physiological imbalances that occur with increased water and energy demand under such prolonged water stress conditions. on the other hand, r. mucronata seedlings could build up some safety mechanisms inside the plants with anatomical modifications though, efficiency of water transportation was not wellmaintained as reflected by decreased hydraulic conductivity of the plants under highly stressed conditions. further, it clearly demonstrated that the water use efficiency of the r. mucronata plants also reduces under the stress conditions. the results of the study demonstrate that, under abiotic stress conditions, rhizophora mucronata seedlings are able to survive for a certain period (up to some critical level) through some structural and physiological adjustments. nevertheless, if these highly stressed conditions prevail for a longer period (prolonged exposure), the r. mucronata plants are unable to secure their survival. this clearly depicts that the impacts of stress conditions, created on mangrove plants particularly in early life stage due to the inappropriate planting practices, are more detrimental when compared with the natural fluctuations of salinity and availability of water in mangrove ecosystems. this emphasizes the need of understanding the lagoon history with special attention to behavior of lagoon hydrology including inundation levels, water depth, salinity and selection of suitable tidal position accordingly for mangrove planting. acknowledgement we kindly acknowledge the department of botany, university of ruhuna, sri lanka for giving a working place for the laboratory experiments. also, the authors would like to thank mrs. sk madarasinghe, post-graduate student, university of ruhuna for her assistance in taking anatomical measurements and n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science vol 9: 13-31, june 2018 29 performing descriptive analyses. two anonymous reviewers are acknowledged for their comments on the initial draft. references ahmad iu. 2012. status of mangrove plantations in the living delta: an overview of the coastal afforestation experience of bangladesh. in macintosh dj, mahindapala r, markopoulos m. (eds.), sharing lessons on mangrove restoration (pp. 81-93), proceedings and a call for action from an mff regional colloquium, mamallapuram, india. akram m, akhtar s, javed ih, wahid a, rasul e. 2002. anatomical attributes of different wheat (triticum aestivum) accessions/varieties to nacl salinity. international journal of agriculture and biology 4: 166–168. alongi dm. 2002. present state and future of the world’s mangrove forests. environment conservation 29: 331–349. aloni r. 2004. the induction of vascular tissue by auxin. in davies pj. (ed.), plant hormones: biosynthesis, signal transduction, action (pp. 471492), dordrecht: kluwer. arend m, fromm j. 2007. season change in the drought response of wood cell development in poplar. tree physiology 27: 985–992. aziz i, khan ma. 2001. effect of seawater on the growth, ion content and water potential of rhizophora mucronata lam. journal of plant research 114: 369–373. ball mc. 1988. ecophysiology of mangroves. trees structure and function 2: 129–142. bass da, parce jw, dechatelet lr, szejda p, seeds mc, thomas m. 1983. flow cytometry studies of oxidative product formation by neutrophils: a graded response to membrane stimulation. journal of immunology 130(4): 1910–1917. bosire jo, dahdouh-guebas f, walton m, crona bi, lewis rr, field c, kairo jg, koedam n. 2008. functionality of restored mangroves: a review. aquatic botany 89: 251–259. brown b, yuniati w, ahmad r, soulsby i. 2014a. observations of natural recruitment and human attempts at mangrove rehabilitation after seismic (tsunami and earthquake) events in simulue island and singkil lagoon, acheh, indonesia. in santiago-fandino v, kontar ya, kaneda y. (eds.), post-tsunami hazard reconstruction and restoration (pp. 311-327), springer. brugnoli e, lauteri m. 1991. effects of salinity on stomatal conductance, photosynthetic capacity, and carbon isotope discrimination of salt-tolerant (gossypium hirsutum l.) and salt-sensitive (phaseolus vulgaris l.) c3 non-halophytes. plant physiology 95: 628-635. choat b, ball mc, luly jg, holtum jam. 2005. hydraulic architecture of deciduous and evergreen dry rainforest tree species from north-eastern australia. trees-structure and function 19: 305–311. cochard h. 2006. cavitation in trees. comptes rendus physique 7: 1018–1026. dahdouh-guebas f, jayatissa lp, di nitto d, bosire j, lo seen, koedam n. 2005b. how effective were mangrove as a defence against the recent tsunami? current biology 15: 443447. dissanayake np, madarasinghe sk, kodikara kas, jayatissa lp, perera ajd, koedam n, dahdouh-guebas f. 2014. preliminary study on the propagule dependency of rhizophora seedlings. journal of department of wildlife conservation 2: 141-151. hacke ug, sperry js, wheeler jk, castro l. 2006. scaling of angiosperm xylem structure with safety and efficiency. tree physiology 26: 689–701. hoppe-speer scl, adams jb, rajkaran a, bailey d. 2011. the response of the red mangrove rhizophora mucronata lam. to salinity and inundation in south africa. aquatic botany 95: 71– 76. hsiao tc. 1973. plant responses to water stress. annual review of plant physiology 24: 519570. n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science 30 vol 9: 13-31, june 2018 lewis rr. 2005. ecological engineering for successful management and restoration of mangrove forests. ecological engineering 24: 403-418. lewis rr. 2009. methods and criteria for successful mangrove forest restoration. in perillo gme, wolanski e, cahoon dr, brinson mm. (eds.), coastal wetlands: an integrated ecosystem approach (pp. 787–80), elsevier press, amsterdam, the netherlands. lewis rr, brown b. 2014. ecological mangrove rehabilitation – a field manual for practitioners. version 3. mangrove action project indonesia, blue forests, canadian international development agency, and oxfam. jayatissa lp, kodikara kas, dissanayaka np, satyanarayana b. 2016. post-tsunami assessment of coastal vegetation, with the view to protect coastal areas from ocean surges in sri lanka. in santiago-fandiño v, tanaka h, spiske m. (eds.) tsunamis and earthquakes in coastal environments, coastal research library. springer international publishing, pp 47–64. jayatissa lp, wickramasinghe waadl, dahdouh-guebas f, huxham m. 2008. interspecific variations in responses of mangrove seedlings to two contrasting salinities. international review of hydrobiology 93: 700-710. kathiresan k. 2011. conservation and management strategies: restoration technologies. in kathiresan k, ajmal khan s. (eds.), coastal biodiversity in mangrove ecosystems: unuinweh-unesco international training course (pp. 609–632). online at: http://ocw.unu.edu/international-network-on-water-environment-and-health/unuinwehcourse-1-mangroves/restoration-technologies.pdf. kodikara kas, mukherjee n, jayatissa lp, dahdouh-guebas f, koedam n. 2017a. have mangrove restoration projects worked? an in-depth study in sri lanka. journal of restoration ecology 25(5): 705–716. kodikara kas, jayatissa lp, huxham m, dahdouh-guebas f, koedam n. 2017b. the effects of salinity on growth and survival of mangrove seedlings changes with age. journal of acta botanica brasilica 32(1): 37-46. larcher w. 2003. physiological plant ecology: ecophysiology and stress physiology of functional groups. 4th edn. springer-verlag, berlin, pp. 513. lee sy, primavera jh, dahdouh-guebas f, mckee k, bosire jo, cannicci s, diele k, fromard f, koedam n, marchand c, mendelssohn i, mukherjee n, record s. 2014. ecological role and services of tropical mangrove ecosystems: a reassessment. global ecology and biogeography 23: 726–743. mukherjee n, dahdouh-guebas f, koedam n, shanker k. 2015. an interdisciplinary framework to evaluate bioshield plantations: insights from peninsular india. acta oecologica 63: 91–100. munns r, tester m. 2008. mechanisms of salinity tolerance. annual review of plant biology 59: 651-681. naidoo g. 1985. effects of water logging and salinity on plant–water relations and on the accumulation of solutes in three mangrove species. aquatic botany 22: 133–143. naidoo g. 2006. factors contributing to dwarfing in the mangrove avicennia marina. annals of botany 97: 1095–1101. national resources conservation service (nrcs). 2010. soil quality kit: guides for educators, united sates department of agriculture (usda), u.s.a. http://www.nrcs.usda.gov/wps/portal/nrcs/site/national/home/ (accessed in 25 may 2018). parker wc, pallardy sg. 1985. stem vascular anatomy and leaf area in seedlings of six black walnut (juglans nigra) families. canadian journal of botany 63(7): 12661270. https://doi.org/10.1139/b85-175 polidoro ba, carpenter ke, collins l, duke nc, ellison am, ellison jc, farnsworth ej, fernando es, kathiresan k, koedam n, livingstone sr, miyagi t, moore ge, nam vn, ong je, primavera jh, salmo sg, sanciangco jc, sukardjo s, wang y, yong jwh. 2010. http://www.vub.ac.be/apna/staff/fdg/pub/jayatissaetal_2008_internatrevhydrobiol.pdf http://www.vub.ac.be/apna/staff/fdg/pub/jayatissaetal_2008_internatrevhydrobiol.pdf http://ocw.unu.edu/international-network-on-water-environment-and-health/unu-inwehcourse-1-mangroves/restoration-technologies.pdf http://ocw.unu.edu/international-network-on-water-environment-and-health/unu-inwehcourse-1-mangroves/restoration-technologies.pdf http://www.nrcs.usda.gov/wps/portal/nrcs/site/national/home/ https://doi.org/10.1139/b85-175 n.p. dissanayaka et al. responses of xylem in rhizophora seedlings ruhuna journal of science vol 9: 13-31, june 2018 31 the loss of species: mangrove extinction risk and geographic areas of global concern. plos one 5: e10095. primavera jh, esteban jma. 2008. a review of mangrove rehabilitation in the philippines: successes, failures and future prospects. wetland ecology and management 16: 173–253. richards dr, friess da. 2016. rates and drivers of mangrove deforestation in southeast asia, 2000–2012. proceedings of the national academy of sciences 113: 344–349. robert emr, koedam n, beeckman h, schmitz n. 2009. a safe hydraulic architecture as wood anatomical explanation for the difference in distribution of the mangroves avicennia and rhizophora. functional ecology 23: 649–657. samson ms, rollon rn. 2008. growth performance of planted red mangroves in the philippines: revisiting forest management strategies. ambio 37: 234-240. satyanarayana b, van der stocken t, rans g, kodikara kas, ronsmans g, jayatissa lp, mohd-lokman h, koedam n, dahdouh-guebas f. 2017. island-wide coastal vulnerability assessment of sri lanka reveals that sand dunes, planted trees and natural vegetation may play a role as potential barriers against ocean surges. global ecology and conservation 12: 144-157. http://dx.doi.org/10.1016/j.gecco.2017.10.001 schmitz n, verheyden a, beeckman h, kairo jg, koedam n. 2006. influence of a salinity gradient on vessel characters of the mangrove species rhizophora mucronata. annals of botany 98: 1321–1330. shi sh, huang yl, zeng k, tan fx, he hh, huang jz, fu yx. 2005. molecular phylogenetic analysis of mangroves: independent evolutionary origins of vivipary and salt secretion. molecular phylogenetic and evolution 34: 159–166. spalding md, kainuma m, collins l. 2010. world atlas of mangroves. london, u.k. earthscan. tomlinson pb. 1986. the botany of mangroves. london, cambridge university press. pp. 419. tomlinson pb. 1994. the botany of mangroves. cambridge university press, cambridge. werner a, stelzer r. 1990. physiological responses of the mangrove rhizophora mangle grown in the absence and presence of nacl. plant, cell & environment 13: 243–255. http://dx.doi.org/10.1016/j.gecco.2017.10.001 ruhuna journal of science vol 12 (2): 115-127, december 2021 eissn: 2536-8400 © faculty of science http://doi.org/10.4038/rjs.v12i2.106 university of ruhuna © faculty of science, university of ruhuna sri lanka 115 air pollution affects lichen species richness, species density, relative growth form abundance and their secondary metabolite production: a case study in kandy district, sri lanka w.g.d.i. gunawardana*, e.s.m. edirisinghe, c.l. abayasekara and a.d.s.n.p. athukorala department of botany, faculty of science, university of peradeniya, peradeniya, sri lanka *correspondence: 91.duleeka@gmail.com; orcid: https://orcid.org/0000-0001-9960-2893 received: 8th august 2021, revised: 19th november 2021, accepted: 15th december 2021 abstract lichens are symbiotic associations between fungi and algae and/or cyanobacteria, consisting of three forms, viz: fruticose, foliose and crustose. air pollution affects lichen diversity, percent cover, and density. the current study mainly compared the lichen species richness, density and relative abundance of each growth form among three selected sites with different degrees of pollution, viz: a site located in the kandy city (s1), a site located 11 km away from the city (s2) and a site within a forest patch located ~12 km away from the kandy city (s3). a random sampling method was used to collect lichens within a two km distance in the three selected sites. percent cover and density were determined using a quadrat ladder. acetone extracts of lichens were subjected to thin-layer chromatography (tlc), and secondary metabolites were identified by visualizing under uv (254 and 365 nm) and by rf values. a total of 24 lichen species were collected (s1=8, s2=12, and s3=4). percentage richness of crustose and foliose lichens was higher in s2 compared to s1, while leparia sp. and lecanora sp. were common in s1 and s2. atranorin, salazinic acid, and zeorin were detected as common compounds from lichens in s1 and s2 sites, exhibiting photoprotecting and antioxidant properties. fumarprotocetraric acid, which tolerates harsh environmental conditions, and physodalic acid, which is produced in response to pollution stress were detected from lichens from s1. norstictic acid was identified in lichens from s2. the results show a difference between the lichen community and secondary metabolites among the three sites. keywords: air pollution, lichen density, lichen species richness, secondary metabolites, tlc analysis. 1 introduction lichens are symbiotic associations between fungi (mycobiont) and algae and/or cyanobacteria (photobiont), which form a mutual biological union (nash 2008). lichens have different growth forms as crustose, foliose and fruticose (nash 2008). size of the lichen thallus/density, species diversity and the growth form composition https://rjs.ruh.ac.lk/index.php/rjs/index https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v12i2.106 https://creativecommons.org/licenses/by-nc/4.0/ mailto:91.duleeka@gmail.com https://orcid.org/0000-0001-9960-2893 w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 116 (crustose, foliose, fruticose) are affected by many natural factors and man-made factors. air pollution as a result of rapid urbanization has been found to affect the lichen diversity, percent cover, and density, and become one of the limiting factors for establishing successful lichen communities (cislaghi and nimis 1997, nash 2008). lichens are used as biomonitors and bio accumulators especially in europe (henderson-sellers and seaward 1979, nimis and purvis 2002). in addition to morphological and compositional changes that result from habitat changes, chemical profiles can also be affected. lichens are known to be sources of unique secondary metabolites. secondary metabolites are produced by the fungal partner and secreted onto the hyphae as amorphous forms or crystals (dharmadhikari and chettiar 2010). atranorin, parietin, usnic acid, fungal melanins accumulate in the cortex and physodic acid, physodalic acid, and protocetraric acid in the medullary layer. secondary metabolites are produced as a result of limiting one or more nutrients and when growth slows down (moore 1998). the production of secondary compounds may not serve specific functions, but they may have a selective benefit with multiple inadvertent ecological functions (deduke et al. 2012). lichen secondary metabolites are divided into many classes including amino acid derivatives, aliphatic acids, dibenzofurans, depsides, depsidones, depsones, sugar alcohols, macrolytic lactones, monocyclic aromatic compounds, chromones, terpenoids, steroids, carotenoids, diphenyl ethers, quinines, and xanthones (huneck 1999). over 800 secondary metabolites of lichens have been investigated which exhibit a wide variety of biological actions including antibiotic, antimicrobial, antiviral, anti-inflammatory, analgesic, antipyretic, anti-proliferative and cytotoxic effects (boustie and grube 2005, basile et al. 2015). secondary metabolites are suggested to be mainly produced in response to extreme environmental conditions (solhaug and gauslaa 1996, nybakken et al. 2004, mcevoy et al. 2006, boustie et al. 2011). sensitivity of lichen secondary metabolites to heavy metal accumulation has been observed which is suggested to play a general role in metal homeostasis and pollution tolerance. białonska and dayan (2005) reported a decrease in the levels of atranorin, physodic acid and hydroxyphysodic acid in thalli of hypogymnia physodes transplanted to the vicinity of a chemical plant producing chromium, phosphorous and sulfur compounds. however, a significant increase in the level of physodalic acid was observed suggesting that this compound is produced in response to the pollution stress. in evernia mesomorpha and ramalina menziesii, usnic acid and divaricatic acid significantly increased the intracellular uptake of cu2+ but reduced the intracellular uptake of mn2+ (hauck et al. 2009). these compounds were suggested to facilitate the survival of those two particular lichen species. pawlik-skowronska and backor (2011) investigated the effect of zn/pb-pollution on secondary metabolites production in hypocenomyce scalaris, cladonia furcata and lepraria spp. and suggested that zn and pb ions tolerate lichen species that contain higher amounts of lecanoric, fumarprotocetraric, stictic, constictic acids, and atranorin. nakajima et al. (2015) reported that the relative concentrations of stictic and norstictic acids in lichen samples collected at cu w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 117 polluted sites were significantly lower compared to those at control sites. the coordinated production of usnic acid with salazinic acid was investigated by valencia-islas et al. 2007 and amo de paz et al. 2010 and show that usnic acid and salazinic acid share similar effects due to air pollution and antioxidant behaviour. however, studies to elucidate the impact of habitat disturbances on lichen metabolite production coupled with lichen diversity are intermittent globally and are rare in the sri lankan context, though about 700 lichen species have been recorded in sri lanka (weerakoon 2015). therefore, the goal of the current study was to investigate the effect of environmental changes resulting from anthropogenic disturbances on lichen diversity and its biochemistry. the objectives of the current study were to compare lichen species richness and density and to assess the number and type of secondary metabolites produced by each species collected from three sites comprising of different levels of air pollution. 2 material and methods 2.1 selection of study sites kandy city has been declared as one of the most polluted cities in sri lanka with no2, so2, o3 concentrations exceeding the permitted sri lankan standard levels for ambient air quality and subjected to transboundary pollution in certain months of the year (abeyratne and illaperuma 2006). elangasinghe and shanthini (2008) reported a positive correlation between atmospheric particular matter less than 10 μm in size (pm10) concentration in kandy in relation to traffic intensity. site 1 (s1) was around the kandy lake, which is situated in the middle of the city of kandy (7°17'25.4"n 80°38'23.7"e) where heavy traffic is observed throughout the day and night, and with a recorded value of 200 μg/m3 of atmospheric particular matter less than 10 μm in size (pm10) (elangasinghe and shanthini 2008). university of peradeniya, which is located on the opposite basin of the hantana mountain range and at the edge of upper and lower hantana forest reserves was selected as the second site (s2) (7°15'17.6"n 80°35'55.5"e). since the traffic flow is less along the road that runs across the university, the atmospheric particular matter less than 10 μm in size (pm10) was recorded as 25 μg/m3 (elangasinghe and shanthini 2008). gannoruwa forest reserve, which is spread over the mountain facing the upper and lower hantana forest reserve was selected as the third site (s3) (7°17'16.0"n 80°35'48.0"e). it is isolated and located about 5 km away from the less busy highway. based on the location of the three sites, disturbances to the lichen communities inhabiting these sites were assumed to be varied, especially with respect to air quality (figure 1). w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 118 fig 1. three sampling sites (s1, s2, s3) and location of hantana mountain range in the kandy district, sri lanka 2.2 comparison of relative abundance of growth form and lichen species richness corticolous lichen species richness in the three selected sites was determined using a quadrat ladder (each having 10×10 cm2 contiguous quadrats). substrate tree species was recorded, and the quadrat ladder was placed on the tree trunk 1.5 m above the ground level. four aspects viz: n, e, s, w of the trunk were examined and the presence of lichens according to their thallus forms; crustose, foliose and fruticose within the quadrat was recorded. lichen richness was determined by recording lichen species on 10 randomly selected tree species as the above-mentioned procedure and the relative abundance of each thallus forms were calculated using the following formulae. relative abundacne of growth form = no. of lichen of a paricular growth form total no. of lichen belonging to all growth forms x 100 percent coverage of each lichen species and lichen forms were visually estimated by placing the quadrat ladder using the method described above. density of each lichen species and growth forms within the quadrat ladder was obtained using the following formula. density = no. of total individuals of a particular lichen species/form size of the quadrat w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 119 2.3 comparison of chemical profiles of lichen species sample collection three forms of corticolous lichen samples were collected from the selected sites by using random sampling method. site1 was a "~2 km stretch along the road running around the kandy lake in the city of kandy (s1) (figure 1). site 2 was at the university of peradeniya along a ~2 km stretch of the road running through the university (s2), and site-3 was gannoruwa forest reserve along the foot path ~2 km (s3). samples were collected from ten trees from each site and four 10 x10 quadrats per tree. therefore, the total number of quadrats is 40 per site. for each lichen sample, three replicate samples were collected for identification and chemical analyses. subsequently, the samples were sorted out according to thallus form, air dried and stored at room temperature until processing. samples were identified with standard taxonomic keys using morphological and chemical characteristics (weerakoon 2015). thin layer chromatography (tlc) the extract was prepared under sterile conditions according to the method described by culberson (1972). secondary metabolites in all samples were extracted using acetone and identified according to spot characteristics as follows (orange et al. 2001). one millimeter of each acetone extract was spotted on tlc plates with a capillary tube. tlc plates were allowed to dry for 30 minutes to one hour. each plate was run in the developing chamber with the solvent system a; toluene/ethyl acetate/formic acid (139:83:8 v/v/v). tlc plates were observed under a uvilluminator (254 nm and 365 nm) and spot characteristics were recorded. sulfuric acid (10%) was sprayed on tlc plates and plates were oven-dried at 80oc for 10 minutes to develop the characteristics further. secondary metabolites were determined by comparison with known characteristics (orange et al. 2001), by using a standard for rf comparison. a mixture of acetone extract of usnic acid (rf 6) from usnea sp. and norstictic acid (rf 4) from ramalina sp. was used as the standard reference (supplementary file 1). 3 results 3.1 comparison of relative abundance of growth forms and lichen species abundance based on a few previous studies on lichen diversity and air pollution (cislaghi and nimis 1997, nash 2008), higher species richness of lichens and an abundance of fruticose lichens were expected in the gannoruwa forest reserve (s3), while species richness and fruticose and foliose lichen abundance were expected to be lowest w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 120 around the kandy lake (s1). expected patterns were observed in polluted (s1) and semi polluted (s2) habitats in terms of lichen richness and density. as shown in table 1 and 2, the total number of species was higher in the s2 site and the least number of species was found in s3. crustose lichens were found in all sites and foliose lichens were found mainly in s2 and few in s1. fruticose lichens were absent in all sites. relative abundance of crustose and foliose lichens was higher in s2 compared to s1. table 1: percentage diversity and distribution of different forms of lichen thalli in the three sites site total number of species thallus form (relative abundance %) crustose foliose fruticose (s1) 8 37.5 62.5 ab (s2) 12 41.66 58.33 ab (s3) 4 100 ab ab ab = absent table 2: list of lichen species identified from each site, their growth form, and identified secondary metabolites in each lichen species using thin-layer chromatography. site growth form species secondary metabolites s1 crustose graphis sp. 2 physodic acid, fumarprotocetraric acid, acetylportentol crustose lecanora sp. physodic, unidentified compound crustose lepraria sp. physodalic, zeorin foliose leptogium sp. no compound foliose lobothallia sp. norstic acid, zeorin foliose parmotrema sp. atranorin, salazinic, zeorin foliose physcia sp. atranorin, zeorin foliose pyxine sp. lichexanthone s2 crustose cryptothecia sp. chiodectonic acid crustose graphis sp. 2 leanoric acid, stictic acid, norstictic acid crustose lecanora sp. constictic acid, salazinic acid, stictic acid crustose lepraria sp. atranorin, constictic acid, stictic acid, zeorin foliose heterodermia sp. zeorin foliose leptogium sp. no compound foliose parmotrema sp. 1 salazinic acid foliose parmotrema sp. 2 atranorin, salazinic, zeorin foliose parmotrema sp. 3 atranorin, lecanoric acid foliose physcia sp. 1 salazinic acid, zeorin foliose physcia sp. 2 atranorin, zeorin s3 crustose megalospora sp. porina sp. 1 porina sp. 2 trypethelium sp. atranorin note: s1: ~2 km stretch along the road running around the kandy lake in the city of kandy; s2: university of peradeniya along a ~2 km stretch of the road running through the university; s3: gannoruwa forest reserve along the foot path ~2 km. w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 121 3.2 percentage cover and average density of lichen comparison of average percent cover and average density of lichens found in the three sites are shown in figure 2. the highest percent cover (63%) and the least average density (13.42) values were observed in s3. the reason for this observation is the large healthy patches of crustose lichens covering the tree trunks. moderate percent cover (40.12%) and moderate average density (22.35) values were observed in s2 and the least percent cover (7.07%) and the highest average density (28.47) was observed in s1. fig 2. average percent cover and average density of lichens collected from the three sites. s1: kandy city, s2: university of peradeniya, s3: gannoruwa forest reserve. 3.3 composition of secondary metabolites in lichen species collected from the three sites lichen species in s2 and s1, showed higher number of secondary metabolites than lichens found in s3 (table 2). only atranorin was detected from the crustose lichen species (megalospora sp., porina sp. 1, porina sp. 2 and trypethelium sp.) from the s3. atranorin, salazinic acid, and zeorin were detected as common compounds to s2 and s1, which has photoprotecting and antioxidant properties. in s2, norstictic acid was identified, which is produced under low light levels. in s1, fumarprotocetraric acid (suggested as to tolerate harsh environmental conditions) and physodalic acid (suggested as produced in response to pollution stress) were identified (table 2). lichen species leparia sp. and lecanora sp. were observed in both s2 and s1 sites, but their secondary metabolites profiles were different as shown in table 2. 4 discussion the current study compared the species richness, density, relative abundance of different growth forms and secondary metabolite profiles of the lichen species collected from three different sites where air pollution levels are recorded and 0 10 20 30 40 50 60 70 80 s1 s2 s3 site average % coverage average d ensity w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 122 assumed to be different based on previous records. the selection of the polluted site was based on the study carried out by abeyratne and illaperuma (2006) and elangasinghe and shanthini (2008). a higher number (density) of crustose lichen species was observed in polluted habitat (s1) compared to the semi-polluted habitat (s2) even though it exhibited a fewer number of crustose lichen species. this observation supports the previously established fact that crustose lichens tolerate air pollution (pyatt 1970, nash 1974). this also confirms that the air pollution levels are high around the kandy lake (s1) in comparison to the university (s2). absence of fruticose lichen species in both habitats also suggests that both habitats are disturbed and polluted. however, the current study observed only four crustose lichens species in the sample site in the gannoruwa forest reserve (s3), and the least number of species were collected from this site during this study in comparison to the other two sites. possible reasons for this observation could be the less amount of sunlight due to the canopy cover and higher atmospheric moisture content creating unique environmental conditions suitable for a limited number of lichen species exerting a competition so that other species cannot survive on the tree barks rather than air pollution. the microenvironmental conditions prevailing in the forest area may augment the successful spread of these four species. therefore, in future studies, it is recommended that the sampling area should be increased, expanding to the other side of the gannoruwa forest. in addition, the nearby natural forest patches in kandy district including udawaththa forest reserve and hanthana sub mountain forest should also be investigated, to see whether similar patterns are observed. even though only four crustose species were present in gannoruwa forest (s3) their percent cover was very high since the tree barks were covered with large patches of healthy lichen thalli. this further indicates that these four crustose species are well established and dominate the habitat. percent cover of the crustose and foliose lichens in the semi polluted area (s2) was higher than that in the polluted area (s1), indicating that extreme environments can affect the thallus size, hence the growth rate of lichens. based on the fact that secondary metabolites are induced under stress conditions (solhaug and gauslaa 1996, boustie et al. 2011, nybakken et al. 2004, mcevoy et al. 2006), our study expected an increase in the number of secondary metabolites produced by lichens under polluted environments. similarly, crustose lichens were expected to produce more compounds than foliose since crustose tolerate harsh conditions (pyatt 1970, nash 1974). production of secondary metabolites could be one possible mechanism where crustose lichens withstand harsh environments. but both in semi polluted (s2) and polluted habitats (s1), crustose and foliose lichen species showed a similar number (three) of secondary metabolites. from the crustose lichens collected from the unpolluted area (s3), atranorin was identified as the only secondary metabolite that exhibits photoprotecting and antioxidant properties. a previous study has recorded the amount of atranorin in the cortex of parmotreama hypotropum positively correlating with the amount of yearly light reaching the thallus (armaleo et al. 2008). even though there is low light accessing the tree trunk due to w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 123 high canopy cover in the gannoruwa forest area (s3), they still produce atranorin to give protection to the light-sensitive algal partner, however, the quantity may be minimal. this might be an indication that the forest is providing favourable conditions. atranorin was detected in a few lichen samples in all three habitats, and extended studies into investigating the concentration of secondary metabolites would give an apparent effect of air pollution on secondary metabolites production. only two lichen species (crustose) were found to be common in both semi polluted (s2) and polluted (s1) areas. these showed a difference in the numbers of secondary metabolites, and their composition (leparia sp. and lecanora sp.). physodalic acid was identified as one secondary metabolite in leparia sp., and physodic acid was identified in lecanora sp., which was collected from the polluted area (s1). previous studies reported that the level of physodic acid decreased while physodalic acid increased in response to heavy metal pollution (białonska and dayan 2005). under polluted conditions, both species produce physodalic acid which suggests an increase in the levels in response to pollution stress. physodic acid is shown to decrease in the levels under chemical pollution (chromium, phosphorous, and sulfur compounds) (białonska and dayan 2005), which was observed in the current study. this indicates that the same species may produce different metabolites under different pollution levels. this suggests that the secondary metabolite composition can be affected by different pollution levels, and lichens growing under low air quality can be a source of unique metabolites that have significant applications. therefore, further detailed investigation (identification and quantification) on physodalic acid production in lichen species is also important within this context. apart from the number and the type of metabolites, concentration differences of the same compound can be triggered by environmental differences, which was not determined in the current study. one of the best methods to identify exact compounds and quantify them is high-performance liquid chromatography (hplc). according to pyatt (1970), the crustose and leprosa lichens are more tolerant regarding air pollution than foliose and fruticose forms. the current study provides the same evidence by the presence of leparia sp. in both semi polluted (s2) and polluted (s1) habitats. bajpai et al. (2010) suggested leparia sp. which accumulate a high concentration of almost all the heavy metals can be used to determine the heavy metal pollution in the dry and warm climate of central india that has similar climatic conditions as sri lanka. therefore, leparia sp. may be of value to be used as an indicator species for future studies to determine the ambient air quality in a particular area in sri lanka. atranorin, salazinic acid, and zeorin were common compounds detected in lichens collected from semi polluted (s2) and polluted habitat (s1), which has photoprotecting and antioxidant properties. previous studies (deduka et al. 2012, shukla et al. 2016) reported higher abundance and frequency of everniastrum cirrhattum with increasing altitude and detected the presence of a higher quantity of photoprotecting and antioxidant chemicals, especially salazinic acid. this explains the harsh environmental conditions of both semi polluted (s1) and polluted (s2) sites. w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 124 along the galaha road, the university of peradeniya (s2) mature trees were observed with a large canopy cover compared to trees present around the kandy lake (s1); hence lower level of light is received to the lichen thallus in semi polluted site (s2) when compared to the polluted site (s1). therefore, the concentration of these photoprotecting compounds (atranorin, salazinic acid, zeorin) are expected to be higher in the polluted site. armaleo et al. (2008) detected norstictic acid on the medullary hyphae and found a negative correlation with yearly light levels. the authors interpreted that the higher quantities of the medullary compound under lower light levels may be an adaptative link with the need to produce these hydrophobic compounds when water potential increases within the lichen thallus (from low light levels) to facilitate the efficient carbon dioxide diffusion to the algae. as light levels decrease, the water potential in the lichen thallus increases, and at the same time, the requirement of hydrophobic compounds also increases. in semi polluted site (s2), graphis sp.1 were identified with norstictic acid and thus provided little evidence of its advantage under low light levels. fumarprotocetraric acid that is suggested to be produced to tolerate harsh environmental conditions (culberson et al. 1977) was identified only in graphis sp. 2, which was collected from the polluted site (s1). as future directions, the concentration of so2, no2, and o3 should be determined in the three sites to confirm the pollution level. in addition to the gannoruwa forest reserve, other forests such as udawaththa forest and upper hanthana forests need to be examined to find out whether the same trend of lichen diversity is present there. the current study supports the primarily established fact that lichen species diversity and growth form composition decrease with increasing environmental stress, in this case, air pollution. however, studies on the relationship between air pollution levels, lichen metabolite production, including lichen species richness are not prevalent both globally and in the sri lankan context. based on the available literature, this is the third study that has been carried out to determine the effect of pollution stress on secondary metabolites of lichens globally and the first of such studies in sri lanka. this is one of the few studies to investigate the possible impact of low air quality on lichen community in the kandy city, which has been declared as the second most polluted city in sri lanka by the central environmental authority, sri lanka. one of the major limitations of this study was measuring the levels of gaseous air pollutants in each selected site. however, due to the unavailability of resources to monitor the air quality, we went for an alternative method to get already recorded data from possible places. in addition, after careful observation of the location of three sites with respect to traffic conditions, we made an educated guess that the selected sites must have three distinctive levels of air pollution. after an intensive search, we found some published data to support our guess regarding particulate matter (elangasinghe and shanthini 2008) in two of the areas in our study. furthermore, a follow-up study conducted by our research group detected a higher number of heavy metal types on lichen thalli collected from around kandy lake than those collected from the same study area in the university of peradeniya (esm w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 125 edirisinghe 2020, unpublished data). based on these additional data, we are very confident that the air pollution levels in the three selected sites are significantly different from each other and vary as we assumed. further, using the atmospheric purity index could have been ideal in this situation. however, the study design and the data collected during the study do not support calculating api. the current study is an initiative to investigate the relationship between air pollution and lichen metabolite production will be a platform for extensive research on identifying and quantifying unique secondary metabolites produced under certain environmental conditions and further studies will be helpful to identify the mechanism behind the survival of lichens under stress conditions. 5 conclusions this study mainly compared the lichen species richness, density, and relative abundance of each growth form among three selected sites with different degrees of pollution levels. a higher number (density) of crustose lichen species was observed in polluted habitat (s1) compared to the semi-polluted habitat (s2), even though it exhibits a fewer number of crustose lichen species. not finding fruticose lichen species in both habitats suggests that both habitats are disturbed and polluted. percentage richness of crustose and foliose lichens was higher in semi-polluted habitat (s2) compared to polluted habitat (s1). leparia sp. and lecanora sp. were common to s1 and s2, but their secondary metabolites profile differed in each site. the results show a difference between the lichen community and secondary metabolites among the three sites. the current study is an initiative to investigate the relationship between air pollution and lichen metabolite production will be a platform for extensive research on identifying and quantifying unique secondary metabolites produced under certain environmental conditions and further will be helpful to identify the mechanism behind the survival of lichens under stress conditions. acknowledgements all authors acknowledge financial support from the department of botany, faculty of science, university of peradeniya, sri lanka. comments from two anonymous reviewers are acknowledged. references abeyratne vd, ileperuma oa. 2006. air pollution monitoring in the city of kandy: possible transboundary effects. journal of the national science foundation of sri lanka 34(3): 137-141; doi: org/10.4038/jnsfsr.v34i3.3644 amo de paz g, raggio j, gomez-serranillos mp, palomino om, gonzalez-burgos e, carretero, me, crespo a. 2010. hplc isolation of antioxidant constituents from xanthoparmelia spp. journal of pharmaceutical and biomedical analysis 53: 165–171; doi: org/10.1016/j.jpba.2010.04.013 https://doi.org/10.1016/j.jpba.2010.04.013 w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 126 armaleo d, zhang y, cheung s. 2008. light might regulate divergently depside and depsidone accumulation in the lichen parmotrema hypotropum by affecting thallus temperature and water potential. mycologia 100(4): 565-576; doi: org/10.3852/07-162r bajpai r, upreti dk, dwivedi sk. 2010. passive monitoring of atmospheric heavy metals in a historical city of central india by lepraria lobificans nyl. environmental monitoring and assessment 166(1-4): 477-484; doi: org/10.1007/s10661-009-1016-4 basile a, rigano d, loppi s, di santi a, nebbioso a, sorbo s, conte b, paoli l, ruberto fd, molinari am, altucci l, bontempo p. 2015. antiproliferative, antibacterial and antifungal activity of the lichen xanthoria parietina and its secondary metabolite parietin. international journal of molecular sciences 16(4): 7861-7875. białońska d, dayan fe. 2005. chemistry of the lichen hypogymnia physodes transplanted to an industrial region. journal of chemical ecology 31(12): 2975-2991; doi: org/10.1007/s10886-0058408-x boustie j, grube m. 2005. lichens a promising source of bioactive secondary metabolites. plant genetic resources 3(2): 273; doi: org/10.1079/pgr200572 boustie j, tomasi s, grube m. 2011. bioactive lichen metabolites: alpine habitats as an untapped source. phytochemistry reviews 10(3): 287-307; doi: org/10.1007/s11101-010-9201-1 cislaghi c, nimis pl. 1997. lichens, air pollution and lung cancer. nature 387(6632): 463; doi: org/10.1038/387463a0 culberson cf. 1972. improved conditions and new data for identification of lichen products by standardized thin-layer chromatographic method. journal of chromatography 72(1): 113-125; doi: org/10.1016/0021-9673(72)80013-x culberson cf, culberson wl, johnson a. 1977. thermally induced chemical artifacts in lichens. phytochemistry 16(1): 127-130; doi: org/10.1016/0031-9422(77)83029-x deduke c, timsina b, piercey-normore md. 2012. effect of environmental change on secondary metabolite production in lichen-forming fungi. international perspectives on global environmental change. intech 197-230. dharmadhikari m, jite pk, chettiar s. 2010. antimicrobial activity of extracts of the lichen and its isolated mycobiont. asian journal of experimental biological sciences 54-58. edirisinghe esm. 2020. can lichens be indicators for air pollution monitoring in kandy city, sri lanka. (unpublished dissertation) department of botany, faculty of science, university of peradeniya, sri lanka. elangasinghe ma, shanthini r. 2008. determination of atmospheric pm10 concentration in kandy in relation to traffic intensity. journal of the national science foundation of sri lanka 36(3). hauck m, jürgens sr, willenbruch k, huneck s, leuschner c. 2009. dissociation and metal-binding characteristics of yellow lichen substances suggest a relationship with site preferences of lichens. annals of botany 103(1): 13-22; doi: org/10.1093/aob/mcn202 henderson-sellers a, seaward mrd. 1979. monitoring lichen reinvasion of ameliorating environments. environmental pollution 19(3): 207-213; doi: org/10.1016/0013-9327(79)90042-9 huneck s. 1999. the significance of lichens and their metabolites. naturwissenschaften 86: 559570.20. mcevoy m, nybakken l, solhaug k, gauslaa y. 2006. uv triggers the synthesis of the widely distributed secondary lichen compound usnic acid. mycological progress 5(4): 221-229; doi: org/10.1007/s11557-006-0514-9 moore d. 1998. fungal morphogenesis. cambridge university press, uk. 469 pp. nakajima h, hara k, yamamoto y, itoh k. 2015. effects of cu on the content of chlorophylls and secondary metabolites in the cu-hyperaccumulator lichen stereocaulon japonicum. ecotoxicology and environmental safety 113: 477-482; doi: org/10.1016/j.ecoenv.2014.12.038 nash th. 1974. lichens of the page environs as potential indicators of air pollution. journal of the arizona academy of science 9(3): 97-101. nash iii th. 2008. lichen biology, 2nd edition, cambridge university press, cambridge. https://doi.org/10.3852/07-162r https://doi.org/10.1007/s10661-009-1016-4 https://doi.org/10.1016/0021-9673(72)80013-x https://doi.org/10.1016/0021-9673(72)80013-x https://doi.org/10.1016/0031-9422(77)83029-x https://doi.org/10.1007/s11557-006-0514-9 https://doi.org/10.1007/s11557-006-0514-9 https://doi.org/10.1016/j.ecoenv.2014.12.038 w.g.d.i. gunawardana et al. air pollution affects lichen species in kandy, sri lanka ruhuna journal of science vol 12 (2): 115-127, december 2021 127 nimis pl, purvis wo. 2002. monitoring lichens as indicators of pollution. monitoring with lichens monitoring lichens 7-10; https://doi.org/10.1007/978-94-010-0423-7_2 nybakken l, solhaug k, bilger a, gauslaa w. 2004. the lichens xanthoria elegans and cetraria islandica maintain a high protection against uv-b radiation in arctic habitats. oecologia 140(2): 211-216; doi: org/10.1007/s00442-004-1583-6 orange a, james pw, white fj. 2001. microchemical methods for the identification of lichens, british lichen society, london, uk. pyatt fb. 1970. lichens as indicators of air pollution in a steel producing town in south wales. environmental pollution 1(1): 45-56; doi: org/10.1016/0013-9327(70)90005-4 pawlik-skowrońska b, bačkor m. 2011. zn/pb-tolerant lichens with higher content of secondary metabolites produce less phytochelatins than specimens living in unpolluted habitats. environmental and experimental botany 72(1): 64-70; doi: org/10.1016/j.envexpbot.2010.07.002 shukla v, patel dk, bajpai r, semwal m, upreti dk. 2016. ecological implication of variation in the secondary metabolites in parmelioid lichens with respect to altitude. environmental science and pollution research 23(2): 1391-1397; doi: org/10.1007/s11356-015-5311-z solhaug ka, gauslaa y. 1996. parietin, a photoprotective secondary product of the lichen xanthoria parietina. oecologia 108(3): 412-418; doi: org/10.1007/bf00333715 valencia-islas n, zambrano a, rojas jl. 2007. ozone reactivity and free radical scavenging behavior of phenolic secondary metabolites in lichens exposed to chronic oxidant air pollution from mexico city. journal of chemical ecology 33(8): 1619-1634; doi: org/10.1007/s10886-007-9330-1 weerakoon g. 2015. fascinating lichens of sri lanka, colombo, sri lanka, ceylon tea services, plc, 188pp. https://doi.org/10.1016/j.envexpbot.2010.07.002 https://doi.org/10.1007/s11356-015-5311-z ruhuna journal of science vol 13 (1): 41-51, june 2022 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v13i1.114 university of ruhuna faculty of science, university of ruhuna sri lanka 41 characterization and risk evaluation of water samples collected from boreholes situated around a dumpsite in obalende, lagos, nigeria tajudeen o. yahaya*1, yunusa abdulganiyu2, titilola f. salisu3, abdulmalik abdulazeez1, abdulrazaq izuafa1, sofiat a. sanni1, abdulmumini i. ahmadu2 1department of biological sciences, federal university birnin kebbi, pmb 1157, kebbi state, nigeria 2department of geology, federal university birnin kebbi, nigeria 3department of zoology and environmental biology, olabisi onabanjo university ago-iwoye, ogun state, nigeria *correspondence: yahaya.tajudeen@fubk.edu.ng, orcid: https://orcid.org/0000-0002-5252-6536 received: 08th september 2021, revised: 03rd march 2022, accepted: 17th june 2022 abstract: dumpsites are used worldwide for waste disposal because they are costeffective and have the capacity to contain enormous amounts of waste. however, concerns are rife about the impact of dumpsites on the quality of nearby groundwater. the present study assessed the quality of borehole water near a dumpsite in obalende, lagos, nigeria. heavy metal, physico-chemical, and microbiological tests were performed on the samples of the water using standard techniques, and the results were compared to the who permissible limits. the average daily oral ingestion (adoi), average daily dermal ingestion (addi), and hazard quotient (hq) of the heavy metals were also estimated. the heavy metal analysis revealed non-permissible levels of zinc, iron, lead, and manganese, while nickel, cadmium, and silicon were within the permissible limits. physico-chemical analysis showed that turbidity, total suspended solids, total dissolved solids, nitrate, and phosphate were within the permissible limits, but not the ph, electrical conductivity, chloride ion, sulphate and dissolved oxygen. the microbiological examination indicated that the water had high levels of bacteria and coliform counts. the hq of zn, fe, and pb, mainly through dermal exposure was above the recommended limits (>1). overall, the results suggest that the water may predispose consumers in the area to zn, fe, pb, and mn toxicities as well as microbial infections. consequently, consumers are advised to treat the water before consuming it. keywords: average daily ingestion, bacteria, dumpsite, lead, nitrate 1 introduction increasing population expansion and industrialization in lagos, nigeria is synonymous with high waste generation (akande 2018). these wastes include domestic, https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v13i1.114 https://creativecommons.org/licenses/by-nc/4.0/ mailto:yahaya.tajudeen@fubk.edu.ng https://orcid.org/0000-0002-5252-6536 t. o. yahaya et al. characterization of borehole water near a dumpsite in nigeria ruhuna journal of science vol 13 (1): 41-51, june 2022 42 agricultural, and industrial wastes and may be classified as liquid, solid, or gaseous (adebayo and obiekezie 2018). most of the wastes are disposed of in open dumpsites across the state by private sectors, local governments, and the lagos state waste management authority. some wastes are also disposed of indiscriminately within the metropolis and water bodies. the dumpsites are poorly managed and concerns are rife about the possible impact of these dumpsites on the quality of nearby groundwater (aboyeji and eigbokan 2016, oyedele and oyedele 2017). the major environmental contaminants in dumpsites are toxic elements and microorganisms. organic wastes in dumpsites produce microorganisms that degrade them, such as bacillus, escherichia coli, klebsiella, proteus, pseudomonas, staphylococcus and streptococcus spp., aspergillus, fusarium, mucor, penicillium, and saccharomyces spp. (williams and hakam 2016, hamid et al. 2019). the decaying wastes produce weak acidic chemicals, which combine with liquids in the waste to form leachate and landfill gas (hamid et al. 2019). additionally, inorganic wastes such as electronics and food containers contain toxic elements, particularly heavy metals, including lead (pb), mercury (hg), arsenic (as), cadmium (cd), and nickel (ni) (popoola et al. 2019). over time, the leachate, along with toxic elements and microorganisms, leaches into the soil and groundwater, compromising drinking water quality (vodyanitskii 2016). heavy metals generate free radicals in animals and plants and cause oxidative damage (rehman et al. 2018). microorganisms enter the cells and disrupt the immune function or elicit toxins (alberts et al. 2002). groundwater is the commonest source of drinking water in most parts of lagos (african groundwater atlas 2019). this is due to insufficient or inefficient pipe-borne water, comparable to what is obtainable in other regions of nigeria and developing nations (yahaya et al. 2020a). furthermore, when properly managed, groundwater is economical, safe, consistent in quality and quantity, and ample for humans (umar et al. 2017). thus, it becomes imperative to keep regular monitoring of the quality of groundwater around dumpsites to prevent health hazards. literature searches show that there is a dearth of information on the quality of groundwater around a dumpsite in obalende, lagos, nigeria. obalende is highly cosmopolitan and one of the most populated areas in the city. this study, therefore, characterized the quality and evaluated the levels and risk of heavy metals in water collected from boreholes situated around a dumpsite in obalende, lagos, nigeria. 2 material and methods 2.1 description of the study area this study was carried out in obalende, lagos, nigeria (figure 1). lagos is the capital of lagos state, at latitudes of 6°37ʹn and 6°70ʹn and longitudes of 2°70ʹe and 4°35ʹe. lagos covers an area of approximately 3,577 km2, of which land constitutes 2,798 km2 t. o. yahaya et al. characterization of borehole water near a dumpsite in nigeria ruhuna journal of science vol 13 (1): 41-51, june 2022 43 and water covers 779 km2 (wang et al. 2018). lagos is bordered by the republic of benin on the west; ogun state on the east and north; and the atlantic ocean on the south. the state is characterized by tropical vegetation, many water bodies, and high rainfall (yahaya et al. 2019a). obalende is among the most densely populated areas in lagos, but, unfortunately, there are no government-approved dumpsites in the area. so, residents dump all sorts of waste indiscriminately, mainly along the mcgregor canal, under the bridge, in obalende. this has resulted in heaps of filthy and stinking waste in the area, necessitating an assessment of the effects of these wastes on the nearby drinking water sources. fig 1: water sampling locations in the obalende area, lagos, nigeria 2.2 sample collection water samples were collected randomly from ten boreholes situated within dwelling places at about 100 m from the dumpsite in february 2021. the samples were collected in polyethylene terephthalate plastic bottles that had been prewashed with a detergent solution, sterilized with 10% nitric acid for 24 h, and then rinsed with distilled water (yahaya et al. 2020a). the samples were covered tightly and refrigerated at -10 °c in the laboratory. t. o. yahaya et al. characterization of borehole water near a dumpsite in nigeria ruhuna journal of science vol 13 (1): 41-51, june 2022 44 2.3 physicochemical and heavy metal analysis the physicochemical properties of the water were characterized based on the guidelines for measuring water quality as described by apha (2012). time-sensitive properties, such as ph, temperature, and electrical conductivity, were measured on-site with a digital ph meter, a mercury-in-glass thermometer, and a conductivity meter, respectively. other properties such as total suspended solids (tss), total dissolved solids (tds), turbidity, dissolved oxygen (do), chloride, sulphate, nitrate, and phosphate were measured in the laboratory as described by yahaya et al. (2020b). heavy metal analysis was carried out as described by yahaya et al. (2019b). one milliliter of each sample was transferred to a pre-washed 100-ml beaker containing an analytical grade of 25 ml of aqua regia mixture (70% hno3 and hcl in a ratio of 3:1, respectively) and 5 ml of 30% h2o2. the mixture was digested in a digestion vessel at 80 °c until a homogenous solution was obtained. afterwards, the solution was cooled, filtered through a whatman no. 42 filter paper into a 50-ml volumetric flask, and diluted to the mark with deionized water. the filtrate was subjected to atomic absorption spectroscopy using a unicam spectrophotometer (model 969) to determine the concentrations of zinc (zn), iron (fe), sodium (na), manganese (mn), lead (pb), cadmium (cd), nickel (ni), and silicon (si). 2.4 microbial analysis the total bacterial counts were estimated using the membrane filtration technique as described by brock (1984). to this end, 100 ml of each water sample was filtered through a sterile cellulose filter (0.2 µm pore size), and the filter was inoculated into a nutrient agar plate and incubated at 35 °c for 24 h. the total number of bacterial colonies formed on the plate was estimated using a colony counter. the membrane filtration technique was also used to estimate the coliform count. however, the two-step enrichment method was used for microbial growth. the filters containing the bacteria were inoculated into an absorbent pad saturated with lauryl tryptose broth and incubated at 35 °c for 2 h. the filters were thereafter transferred to an absorbent pad saturated with m-endo media and incubated at 35 °c for 22 h. the sheen colonies were observed and estimated by a colony counter. 2.5 risk assessment the risks of heavy metals in the water samples were estimated using equations 1, 2, and 3 (usepa, 2003 and 2004). adoi = cx × ir × ef × ed bwt⁄ × at ⋯ ⋯ ⋯ ⋯ [1] t. o. yahaya et al. characterization of borehole water near a dumpsite in nigeria ruhuna journal of science vol 13 (1): 41-51, june 2022 45 in equation 1 above, adoi represents the average daily oral ingestion of a heavy metal per kilogram of body weight, cx is the concentration of heavy metals in water, ir stands for the ingestion rate per unit time, ef indicates the exposure frequency, ed is the exposure duration (average life expectancy of a resident nigerian), bwt means body weight, and at is the average time (ed x ef). the standard values for these parameters were adopted from yahaya et al. (2020a). add1 = cx × sa × pc × et × ed × et bwt⁄ × 𝐴𝑡 ⋯ ⋯ ⋯ ⋯ [2] in equation 2 above, addi is the average daily dermal ingestion of heavy metals, cx represents the concentration of heavy metals in water, sa denotes the total skin surface area, pc indicates the chemical-specific dermal permeability constant (cm/h), et is the exposure time (h/day), ef stands for the exposure frequency (days/years), ed reveals the exposure duration (years), bwt is the body weight, and at is the average time (ed x ef). the standard values for these parameters were adopted from yahaya et al. (2020a). 𝐻𝑄= exposure rfd ⋯ ⋯ ⋯ ⋯ [3] in equation 3 above, hq represents the hazard quotient via oral or dermal ingestion (no units) and rfd stands for oral/dermal reference dose (mg/l/day). oral/dermal reference doses for the selected heavy metals were adopted from yahaya et al. (2020a). 2.6 data analysis the levels of heavy metals and microorganisms in the water samples were presented as mean ± standard deviation (sd) using excel software. the adoi, addi, and hq of the heavy metals were also calculated using excel. 3 results & discussion 3.1 physico-chemical properties of the water samples table 1 shows the physico-chemical properties of the water samples. the ph, electrical conductivity, tds, chloride ion, sulphate, and do levels were all within the world health organization's (who) allowable drinking water limits. however, the turbidity, tss, nitrate, and phosphate were above the permissible limits. this finding suggests that waste from the dumpsite contaminated the water, making it unsafe to drink. the high tss showed that the water contained high inorganic and organic materials, resulting in high turbidity. high turbidity is associated with endemic gastrointestinal illness (mann et al. 2007). high levels of phosphate can cause digestive problems (kumar and puri 2012). abnormal concentrations of nitrate can cause blue-eye t. o. yahaya et al. characterization of borehole water near a dumpsite in nigeria ruhuna journal of science vol 13 (1): 41-51, june 2022 46 syndrome in children and pregnant women (sawyerr et al. 2017). sources of nitrate and phosphate in the water include agrochemicals, human and animal wastes, sewage leaks, detergents in industrial effluents, and run-off from fertilized farmlands (adesuyi et al. 2015). the results of the current study are consistent with those of odukoya and abimbola (2010) and osinbajo et al. (2016), who reported abnormal levels of some water quality parameters in groundwater surrounding dumpsites in lagos. however, the result contradicts majolagbe et al. (2011) and kayode et al. (2018), who found no abnormal levels of nitrate and phosphate in groundwater around some dumpsites in lagos. table 1: mean physico-chemical properties of the water samples obtained from boreholes around a dumpsite in obalende, lagos, nigeria. parameters mean concentrations unit a recommended values for drinking water ph 6.46±1.163 unit 5.5-9.0 turbidity 58.0±0.100 ntu ≤ 5 electrical conductivity 176.6±0.208 µs/cm ≤ 1500 total dissolved solid 657.0±3.46 mg/l ≤ 500 total suspended solid 359.0±1.00 mg/l ≤ 100 chloride 85.2±0.153 mg/l ≤ 250 nitrate 103.1±0.058 mg/l ≤ 50 phosphate 509.0±1.00 mg/l ≤ 0.1 sulphate 25.46±0.031 mg/l ≤ 750 dissolved oxygen 5.68±0.025 mg/l ≥1.0 a who, 2017 3.2 levels of heavy metals in the water samples the levels of zn, fe, na, mn, pb, cd, ni, and si in the water samples are shown in table 2. zn, fe, mn, and pb were detected above the permissible limits, but na, cd, ni, and si were within the permissible limits. these results again prove that the water might have been compromised and so not suitable for drinking. hemochromatosis and tissue damage may result from an excess of fe (arko et al. 2019). high levels of pb may cause high blood pressure, vitamin d and calcium metabolism imbalances, neurological disorders, and multi-organ damage (popoola et al. 2019). excess zn may cause a range of symptoms, including nausea, diarrhea, and headaches (helen and othman 2014). mn toxicity is associated with multi-organ damage and dopaminergic dysfunction (o'neal and zheng 2015). the results obtained under the current study are in line with those of aboyeji and eigbokhan (2016) and oyeku and eludoyin (2010), who detected abnormal concentrations of heavy metals in groundwater around olososu dumpsite in lagos, nigeria. however, longe and balogun (2010) found no significant impact of a dumpsite on groundwater in lagos with regards to heavy metal concentrations. t. o. yahaya et al. characterization of borehole water near a dumpsite in nigeria ruhuna journal of science vol 13 (1): 41-51, june 2022 47 the risk assessment of the heavy metals further shows that daily intake of water may pose some risks. in particular, the addi of zn was beyond the recommended limit (table 2). thus, residents are prone to toxic effects of zn on the skin, such as skin lesions, decreased wound healing, and acrodermatitis (plum et al. 2010). in addition, the hq of oral ingestion of fe and dermal ingestion of zn, fe, and pb were greater than 1. this suggests that residents that live within the average life expectancy of nigerians (55 years) are strongly at risk of fe, zn, and pb toxicity. the risk becomes more significant with increasing age. table 2: levels, estimated average daily ingestion and hazard quotient of heavy metals in water samples obtained from boreholes around a dumpsite in obalende, lagos, nigeria. heavy metals levels (mg/l) a b recommended values for drinking water exposure hazard quotient oral (mg/day) dermal (mg/day) c rdi oral dermal zn 6.056±0.0017 5.0 0.186 9.079 8 0.002 30.263 fe 30.45±0.0020 0.3 0.937 7.607 10 1.329 10.867 na 1.46±0.0020 30-60 0.92 30-60 mn 0.08±0.0018 0.05 0.0006 0.005 1.8 0.043 pb 0.097±0.0010 0.01 0.003 0.097 0.21 0.002 27.714 cd bdl 0.003 0.06 ni bdl 0.02 0.500 si bdl na a values were expressed as mean ± sd, bdl: below detection levels, na: not available, b who: world health organization, 2017; c rdi: recommended daily intake (yahaya et al. 2020a) the possible sources of fe in the water include steel and iron scraps, and sewage (garba and abubakar 2018). zn might have been introduced through metal processing, anti-oxidants, detergent/dispersant, vehicle brakes, and tire wear (jeong 2022). pb could have been introduced by oil spillage from mechanical workshops, welding, panel beatings, pb-bearing glass, pottery glazes, batteries, old lead-based paints, lead pipes, and sewage sludge (tongesay et al. 2018). possible sources of mn in the water include iron and steel scrap, traffic emissions, glass, dry batteries, and chemicals (garba and abubakar 2018). 3.3 levels of microorganisms in the water samples table 3 reveals the levels of bacteria, coliform, and fungi in the water samples. the total bacteria and coliform were detected at levels above the who permissible limits, while fungi were not detected. these results further suggest that the water may not be suitable for drinking. waterborne bacteria may cause diseases such as cholera, diarrhoea, typhoid fever, and dysentery (philip et al. 2017). excess iron concentration t. o. yahaya et al. characterization of borehole water near a dumpsite in nigeria ruhuna journal of science vol 13 (1): 41-51, june 2022 48 might have increased the turbidity of the water and promoted bacterial growth (sawyerr et al. 2017). table 3: levels of microorganisms (mean ± sd) in the water samples obtained from boreholes around a dumpsite in obalende, lagos, nigeria. microorganism mean level (cfu/ml) permissible limit (who 2008) total bacterial 1011 ± 50.14 ≤100 cfu/ml total coliform 400 ± 20.00 0 cfu/ml total fungi/yeast nd 0 cfu/ml nd: not detected; cfu/ml: coliform forming unit per milliliter, who: world health organization. high levels of other nutrients in the water, such as nitrate and phosphate, might have also induced bacterial growth (singh 2013). the detection of coliforms in the water samples indicated that the water was contaminated by organic matter, particularly faecal matter (adelekan and ogunde 2012). certain strains of coliforms such as escherichia coli 0157:h7 can cause urinary tract infections, bacteremia, meningitis, diarrhea, and acute renal failure (gruber et al. 2014, sawyerr et al. 2017). the result of the present study is consistent with those of adeyemi et al. (2007) and odukoya et al. (2013), who detected high microbial populations in groundwater obtained near dumpsites in lagos, nigeria. 5 conclusions the results have demonstrated that the borehole water is turbid and contains nonpermissible levels of tss, nitrate, phosphate, fe, zn, mn, pb, and microbial populations (bacteria and coliforms). the hq of zn, fe, and pb, mainly through dermal exposure, was above the threshold of 1. this indicates that daily ingestion of the water may predispose consumers to health risks, particularly those related to zn, fe, and pb toxicities. from the findings of this study, we recommend that borehole water in the locations be treated before consumption, and boreholes should not be located within 100 m of the diameter of the dumpsite. residents should be informed about the dangers of drinking contaminated water. the government should devise a strategy to prevent people from dumping wastes in or along the canal. similar studies like the current study should be carried out periodically in the locations. acknowledgements two anonymous reviewers are acknowledged for comments on the initial manuscript. t. o. yahaya et al. characterization of borehole water near a dumpsite in nigeria ruhuna journal of science vol 13 (1): 41-51, june 2022 49 references aboyeji os, eigbokhan sf. 2016. evaluations of groundwater contamination by leachates around olusosun open dumpsite in lagos metropolis, southwest nigeria. journal of environmental management 183: 333–341. https://doi.org/10.1016/j.jenvman.2016.09.002. adebayo fo, obiekezie so. 2018. microorganisms in waste management. research journal science and technology 10 (1): 28-39. doi:10.5958/2349-2988.2018.00005.0. adelekan ba, ogunde oa. 2012. quality of water from dug wells and the lagoon in lagos nigeria and associated health risks. scientific research and essays 7 (11): 1195-211. adesuyi aa, nnodu vc, njoku kc, jolaoso a. 2015. nitrate and phosphate pollution in surface water of nwaja creek, port harcourt, niger delta, nigeria. international journal of geology, agriculture and environmental sciences 3 (5): 14-19 adeyemi o, oloyede ob, oladiji at. 2007. physicochemical and microbial characteristics of leachatecontaminated groundwater. asian journal of biochemistry 2: 343-348. doi: 10.3923/ajb.2007.343.348 africa groundwater atlas. 2019. case study: use and perceptions of groundwater in an urban area lagos, nigeria. british geological survey. (accessed july 26, 2021). http://earthwise.bgs.ac.uk/index.php/case_study_use_perceptions_groundwater_lagos. akande d. 2018. the multifaceted waste management issues in nigeria: lagos state as a case study. available at ssrn. http://dx.doi.org/10.2139/ssrn.3235345. alberts b, johnson a, lewis j, raff, m, roberts k, walter p. 2002. cell biology of infection. in: molecular biology of the cell. 4th edition. new york: garland science. https://www.ncbi.nlm.nih.gov/books/nbk26833/. american public health association, apha. 2012. standard methods for the examination of water and wastewater, 22nd edition, american public health association, washington, dc. available from https://www.wef.org/resources/publications/books/standardmethods/ (accessed dec 24, 2021). arko we, hodgson ioa, nyame fk. 2019. assessment of drinking water quality in the dangbe west district of the greateraccra region, ghana. african journal of environmental science and technology 13 (5): 181-190. brock td. 1984. membrane filtration. a user's guide and reference manual. xii + 381. journal of basic microbiology 24 (4): 236-238. https://doi.org/10.1002/jobm.19840240405. garba st, abubakar ma. 2018. source and distribution of the heavy metals: pb, cd, cu, zn, fe, cr, and mn in soils of bauchi metropolis, nigeria. american journal of engineering research 7 (2): 13-23. gruber js, ercumen a, colford jm. 2014. coliform bacteria as indicators of diarrheal risk in household drinking water: systematic review and meta-analysis. plos one 9 (9): e107429. hamid b, jehangir a, baba za, fatima s. 2019. isolation and characterization of cold active bacterial species from municipal solid waste landfill site. research journal of environmental sciences 13: 1-9. doi: 10.3923/rjes.2019.1.9. helen le, othman oc. 2014. levels of selected heavy metals in soil, tomatoes and selected vegetables from lushoto district-tanzania. international journal of environmental monitoring and analysis 2 (6):313-319. jeong h. 2022. toxic metal concentrations and cu–zn–pb isotopic compositions in tires. journal of analytical science and technology 13: 2. https://doi.org/10.1186/s40543021-00312-3. kayode ot, okagbue hi, achuka ja. 2018. water quality assessment for groundwater around a municipal waste dumpsite. data in brief 17: 579-587. https://doi.org/10.1016/j.dib.2018.01.072. kumar m and puri a. 2012. a review of permissible limits of drinking water. indian journal of occupational and environmental medicine 16 (1): 40–44. https://doi.org/10.4103/0019-5278.99696. longe eo, balogun mr. 2010. groundwater quality assessment near a municipal landfill, lagos, nigeria. research journal of applied sciences, engineering and technology 2 (1): 39-44 majolagbe ao, kasali aa, ghaniyu lo. 2011. quality assessment of groundwater in the vicinity of dumpsites in ifo and lagos, south-western nigeria. advances in applied science research 2 (1): 289-298. https://doi.org/10.1016/j.jenvman.2016.09.002 https://dx.doi.org/10.3923/ajb.2007.343.348 http://earthwise.bgs.ac.uk/index.php/case_study_use_perceptions_groundwater_lagos https://dx.doi.org/10.2139/ssrn.3235345 https://www.ncbi.nlm.nih.gov/books/nbk26833/ https://www.wef.org/resources/publications/books/standardmethods/ https://doi.org/10.1002/jobm.19840240405 https://dx.doi.org/10.3923/rjes.2019.1.9 https://doi.org/10.1186/s40543-%20021-00312-3 https://doi.org/10.1016/j.dib.2018.01.072 https://doi.org/10.4103/0019-5278.99696 t. o. yahaya et al. characterization of borehole water near a dumpsite in nigeria ruhuna journal of science vol 13 (1): 41-51, june 2022 50 mann t, janet t, erika w, ann-marie l, barbra s, jason c. 2007. medicare’s search for effective obesity treatments. american psychologist. 62 (3): 220–233 https://doi.org/10.1037/0003066x.62.3.220 obia ae. 2016. emerging nigerian megacities and sustainable development: case study of lagos and abuja. journal of sustainability and development 9 (2): 27-42. https://doi.org/10.5539/jsd.v9n2p27. odukoya am, abimbola af. 2010. contamination assessment of surface and groundwater within and around two dumpsites. international journal of environmental science and technology 7 (2): 367376. odukoya am, oresanya o, abimbola af. 2013. biogeochemical and engineering characteristics of soils and groundwater around a dumpsite. earth science resource journal 17 (1): 53 – 60. o'neal sl, zheng w. 2015. manganese toxicity upon overexposure: a decade in review. current environmental health reports 2 (3): 315–328. https://doi.org/10.1007/s40572-015-0056-x. osinbajo o, adeyi aa, mojolagbe ao. 2016. characterisation of groundwater quality around soluos dumpsite in lagos, nigeria. international journal of water 11(1):1-12. https://doi.org/10.1504/ijw.2017.081110. oyedele oa, oyedele ao. 2017. impacts of waste dumps on the health of neighbours: a case study of olusosun waste dump, ojota, lagos state, nigeria, journal of civil, construction and environmental engineering 2: (1)27-33. https://doi.org/10.11648/j.jccee.20170201.15. oyeku ot, eludoyin ao. 2010. heavy metal contamination of groundwater resources in a nigerian urban settlement. african journal of environmental science and technology 4 (4): 201-214. philip j, landrigan rf, nereus jr, acosta o, adeyi r, arnold nb. 2017. the lancet commission on pollution and health. lancet communications 391 (10119): 462-512. https://doi.org/10.1016/s0140 6736(17)32345-0. plum lm, rink l, haase h. 2010. the essential toxin: impact of zinc on human health. international journal of environmental research and public health 7(4):1342–1365. https://doi.org/10.3390/ijerph7041342. popoola oe, popoola ao, purchase d. 2019. levels of awareness and concentrations of heavy metals in the blood of electronic waste scavengers in nigeria. journal of health & pollution 9 (21):190311. https://doi.org/10.5696/2156-9614-9.21.190311. rehman k, fatima f, waheed i, akash msh. 2018. prevalence of exposure of heavy metals and their impact on health consequences. journal of cell and biochemistry 119 (1): 157-84. sawyerr ho, adeolu at, afolabi as, salami oo, badmos bk. 2017. impact of dumpsites on the quality of soil and groundwater in satellite towns of the federal capital territory, abuja, nigeria. journal of health and pollution 7 (14):15–22. https://doi.org/10.5696/2156-9614-7.14.15. singh as. 2013. nitrate and phosphate contamination in water and possible remedial measures. in book: environmental problems and plant edition: ist chapter: chapter publisher: springer verlag gmbh heidelberg, germany editors: dwivedi, n. pp. 44-56. tongesayi t, kugara j, tongesayi s. 2018. waste dumpsites and public health: a case for lead exposure in zimbabwe and potential global implications. environmental geochemistry and health 40: 375– 381. https://doi.org/10.1007/s10653-017-9917-6 umar ab, ladan b, gado aa. 2017. groundwater evaluation study using electrical resistivity measurements in bunza area of kebbi state, nigeria. international journal of environment and bioenergy 12 (2): 100-114. united states environmental protection agency, usepa. 2004. risk assessment guidance for superfund (rags). vol. 1, human health evaluation manual [internet] washington, d.c. supplemental guidance for dermal risk assessment. available from https://www.epa.gov/risk/risk-assessmentguidance-superfund-rags-part-e. united states environmental protection agency, usepa. 2003. total dissolved solids in drinking water: background document for development of who guidelines for drinking-water quality washington available from http://www.who.int/water_sanitation_health/dwq/chemicals/tds.pdf. https://doi.org/10.1007/s40572-015-0056-x https://doi.org/10.1504/ijw.2017.081110 https://doi.org/10.3390/ijerph7041342 https://doi.org/10.5696/2156-9614-9.21.190311 https://doi.org/10.5696/2156-9614-7.14.15 https://www.epa.gov/risk/risk-assessment-guidance-superfund-rags-part-e https://www.epa.gov/risk/risk-assessment-guidance-superfund-rags-part-e http://www.who.int/water_sanitation_health/dwq/chemicals/tds.pdf t. o. yahaya et al. characterization of borehole water near a dumpsite in nigeria ruhuna journal of science vol 13 (1): 41-51, june 2022 51 vodyanitskii yn. 2016. biochemical processes in soil and groundwater contaminated by leachates from municipal landfills (mini review). annals of agrarian science 14 (3): 249-256. https://doi.org/10.1016/j.aasci.2016.07.009. wang j, maduako in. 2018. spatio-temporal urban growth dynamics of lagos metropolitan region of nigeria based on hybrid methods for lulc modeling andp rediction. european journal of remote sensing 51 (1): 251-65. world health organization, who. 2008. guidelines for drinking waterquality. third edition incorporating the first and second addenda volume1: recommendations. world health organization, geneva, switzerland. available from https://www.who.int/water_sanitation_health/dwq/fulltext.pdf (accessed dec 24, 2021). world health organization, who. 2017. guidelines for drinking-water quality. fourth edition incorporating the first addendum, pp. 1–631. available from https://www.who.int/publicationsdetail-redirect/9789241549950 (accessed dec 24, 2021). williams jo, hakam k. 2016. microorganisms associated with dump sites in port harcourt metropolis, nigeria. journal of ecology and the natural environment 8(2): 9-12. https://doi.org/10.5897/jene2015.0522. yahaya t, ogundipe o, abdulhazeez a, usman b, danjuman j. 2019a. bioaccumulation and health risk assessment of heavy metals in three vegetables consumed in lagos, south-west nigeria. tropical journal of natural product research 3 (11):332-338. https://doi.org/10.26538/tjnpr/v4i1.3. yahaya t, doherty vf, akinola os, shamsudeen a. 2019b. heavy metal profiles and microbial counts of selected sachet water brands in birnin kebbi metropolis, nigeria. ife journal of science, 21 (1): 229-234. https://doi.org/10.4314/ijs.v21i1.20. yahaya t, esther o, itunuoluwa f, abdulmalik a, yusuf y. 2020a. the concentration and health risk evaluation of heavy metals and microorganisms in groundwater in lagos, southwest nigeria. journal of advances in environmental health research 8 (3): 234 -242. https://doi.org/10.22102/jaehr.2020.245629.1183. yahaya t, oladele o, sifau m, audu g, baala j, shamsudeen a. 2020b. characterization and cytogenotoxicity of birnin kebbi abattoir wastewater. uniport journal of engineering and scientific research 5: 63-70. https://doi.org/10.1016/j.aasci.2016.07.009 https://www.who.int/water_sanitation_health/dwq/fulltext.pdf https://www.who.int/publications-detail-redirect/9789241549950 https://www.who.int/publications-detail-redirect/9789241549950 https://doi.org/10.5897/jene2015.0522 https://doi.org/10.26538/tjnpr/v4i1.3 https://doi.org/10.4314/ijs.v21i1.20 https://dx.doi.org/10.22102/jaehr.2020.245629.1183  © 2009 faculty of science university of ruhuna ruhuna journal of science vol. 4, september 2009, pp 65-74 http://www.ruh.ac.lk/rjs/rjs.html issn 1800-279x abstract. factors that affect on the distribution of mollusc, faunus species along the lunuwila ela, galle was determined using ten environmental parameters. four sites including 16 replicates were used and data were collected for six months. among ten environmental parameters investigated, six of them are significantly different among sites. according to the results of regression analysis, none of the physiochemical parameters is responsible for the distribution of this species along this stream. however, it indicates, contents of the substrate make a favorable contribution on their distribution pattern which highest abundance of faunas species recorded from the substrate with fine sand, silt and clay. keywords: mollusca, faunus sp., lunuwila ela: galle, distribution, environmental parameters, substrate type introduction the phylum mollusca is a large assemblage of animals having diverse shapes, sizes, habits and occupy in different habitats (subba rao, 1993). based on their habitat preference, molluscs can be classified into aquatic and land communities. the biomass of the mollusca is very important for ecosystems. in marine environments, molluscs may form the dominant group, especially in their larval stage. in freshwater environments biomass of molluscs is significant in some environments: for example, in temperate countries, lake bottoms may be covered with the small bivalves, which accounts for 80% of the biomass of benthic invertebrates (seddon, 1986). as molluscs are primary consumers in aquatic ecosystems (brown, 1991), they have been used as environmental indicators. due to their economical and ecological importance, as well as sedentary life, molluscs have been assumed as an important organism in monitoring contaminants in different ecosystems (feldstein et al. 2003). they are abundant, sedentary and easy to collect, which makes them ideal organisms in biomonitoring (bresler et al. 2003) programs. bivalve and gastropod molluscs are among the most useful organisms for environmental monitoring (boening, 1999). some molluscs are important (to the ecosystem) as parasite vectors. freshwater molluscs, especially gastropods are important from the medical and veterinary public health point of view. about 100 species of freshwater gastropods are reported acting as intermediate hosts for the diagnostic trematode parasites and among prosobranchs, members of the family pilidae and thiaridae were recorded as harbourers of larval trematodes (subba rao, 1993). 65 investigation of the factors that influence on the distribution of mollusc, faunus sp. (mollusca: gastropoda: thiaridae) along the lunuwila ela, galle. udayantha hmv* and munasinghe dhn*1 1department of zoology, university of ruhuna, matara correspondence: dhnm@zoo.ruh.ac.lk http://www.ruh.ac.lk/rjs/rjs.html udayantha and munasinghe ruhuna journal of science,4,pp55-74 (2009) investigation of the factors that.... however, molluscs are potentially at risk because of the impacts of human activities. three major ways that fresh/brackish water mollusks have been impacted by human activities are introduction of exotic species, habitat loss or alteration and pollution of water. sometimes, introduced (alien) species became invasive, causing biodiversity loss and inflicting major economic and/or ecological damage. the alteration of waterfront for recreation, housing, or commercial development can decrease the diversity of habitats available for molluscs. industrialization including increasing traffic on the access roads leads to the increase of dust, which accumulates in the bottom of the aquatic habitats. especially, accumulation of silt makes a major impact on the distribution of molluscs. changing of other physicochemical parameters also directly or indirectly affect on the distribution of aquatic mollusks groups in the water bodies (lardicci et al. 1997). lunuwila ela is situated within the municipal council area of galle. it starts from walahanduwa area, which is approximately 16 km away from galle town and ends in the wakgalmodera estuary. the lunuwila ela consists of a mixture of saline and freshwater and this stream surrounds approximately 2/3 of the well-known holcim cement factory. the dust particles released from the factory may accumulate within this water body and leads to change the water quality and substrate type of it. it is prominent that siltation is one of the main ways that pollute this water body which makes a major impact on the distribution of the fauna in this stream. different molluscan groups are available in this stream, which shows an irregular distributional pattern (viraj and munasinghe unpublished data). among them gastropod, faunus species received a considerable attention due to their patchy distribution. therefore, the objective of this study was to investigate the environmental parameters that affect on the distribution of mollusc, faunus sp. along lunuwila ela, galle. methodology the experiment was conducted from january to june 2007 and data were collected once a month. four sites were selected along the lunuwila ela from down to the up-stream direction, and each covered approximately 15m in length. distance between two sites was approximately 250m. each site included four replicates and they were selected from both sides of the stream. the distance between each two replicates was approximately 10m. each replicate was covered 1 m 2 area. sites were labeled as 1, 2, 3, and 4 and replicates were labeled as 1a, 1 b, 1c, 1d etc. after placing quadrates all visible faunas shells were counted. mud and sand particles were removed (upper 5 cm layer) and rocks and vegetation were thoroughly inspected for mollusks. sampling was conducted over a period of 6 months. physical and chemical parameters were determined according to the established standard methods (hadrian, 1985). temperature was measured at the site and water samples were taken from each replicate for the analysis of water quality parameters. dissolve oxygen (do), biological oxygen demand (bod), chemical oxygen demand (cod), alkalinity, suspended solids (ss), total dissolved solids (tds) and salinity were measured at the laboratory. finally, a sample of the substrate from each replicate was mixed with water and allowed to separate into layers in a measuring cylinder and contents were recorded. 66 ruhuna journal of science,4,pp65-74 (2009) udayantha and munasinghe investigation of the factors that.... there was no significant variation determined within temporal data that measured for different parameters. therefore, for a particular replicate, obtained temporal data for each parameter were added up to get a mean value. finally, a single mean value for each parameter was calculated for each site using replicate data. using multiple comparison option of anova, variations of parameters among sites were compared. a regression analysis was carried out to find out the relationship among water quality parameters and abundance of individuals of the genus faunus using spss version 10 statistical packages. objectives results of the anova and regression analysis were given in table 1 and 2. according to the results, cod, bod, do, salinity, temperature, hardness and the abundance of the genus faunus were significantly different among four sites. there is no considerable regression between physicochemical parameters and the abundance of faunus sp was resulted for four sites. shells for faunas sp were not recorded from the site 4. number of shells for faunus sp. that recorded for each replicate is given in the table 3. even within one site, contents of the substrate were highly variable among replicates. substrate with coarse sand, fine sand, silts and clay is the common type for four sites. the highest abundance of faunus sp can be seen in the site 02 c which the substrate consisted with fine sand, silt and clay. figure 1: substrate of the faunus species table 1: mean values, standard deviation and significance levels for different parameters of the four sites. 67 udayantha and munasinghe ruhuna journal of science,4,pp55-74 (2009) investigation of the factors that.... paramete r mean value± sd sig. value site 1 site 2 site 3 site 4 cod/mg l-1 0.0254 ± 0.0030 0.0317 ± 0.0063 0.0254 ± 0.0046 0.0303 ± 0.0061 0.000* bod/mg l-1 3.1668 ± 1.0660 2.0631 ± 0.9603 1.6424 ± 1.3079 1.5102 ± 1.1448 0.000* do/mgl1 4.5143 ± 1.1672 9.2928 ± 1.9744 6.1769 ± 1.9240 6.2256 ± 2.9162 0.022* sal/ppt 25.6353 ± 10.0611 19.8812 ± 9.8910 17.3149 ± 10.6489 14.4282 ±11.4830 0.010* alk/m moll-1 0.0102 ± 0.0091 0.0079 ± 0.0068 0.0069 ± 0.0057 0.0055 ± 0.0053 0.195 temp/ 0c 31.0 ± 2.97 33.0 ± 3.44 33.8 ± 2.53 32.0 ± 2.96 0.018* hard 0.0389± 0.0153 0.0298 ± 0.0166 0.0217 ± 0.0178 0.0272 ± 0.0162 0.001* ss/mgml1 0.0350 ±0.0191 0.0429 ± 0.0260 0.0455 ± 0.0351 0.0385 ± 0.0270 0.632 tds/mg ml-1 0.0214 ± 0.0116 0.0172 ± 0.0139 .00133 ±0.0120 0.0130 ± 0.0123 0.133 abundan ce of faunus sp. 25.45 ± 6.15 147.15 ± 95.07 5.70 ± 4.46 0.00 ± 0.00 0.000* (significant level <0.05) (* indicate the parameters, which are significantly different among four sites.) 68 ruhuna journal of science,4,pp65-74 (2009) udayantha and munasinghe investigation of the factors that.... table 2: results of the regression analysis between physicochemical parameters and the abundance of faunus species parameter r2 value for site 01 r2 value for site 02 r2 value for site 03 r2value for site 04 cod / mgl-1 0.039 0.002 0.063 bod / mgl-1 0.003 0.166 0.053 do / mgl-1 0.019 0.167 0.007 sal / ppt 0.022 0.000 0.109 alk / m moll-1 0.035 0.005 0.024 temp / 0c 0.039 0.242 0.012 hard 0.046 0.004 0.147 ss / mg ml-1 0.070 0.009 0.077 tds / mg ml1 0.006 0.016 0.143 table 3: contents of substrate and the recorded number of faunus sp for each site site replicates content of substrate recorded number 01 a silt, clay 165 b fine sand, silt, clay 416 c fine sand, silt, clay 28 d coarse sand, fine sand, silt, clay 34 69 udayantha and munasinghe ruhuna journal of science,4,pp55-74 (2009) investigation of the factors that.... 02 a coarse sand, fine sand, silt, clay 90 b stone, gravel, coarse sand, fine sand, silt, clay 306 c fine sand, silt, clay 2055 d gravel, coarse sand, fine sand, silt, clay 69 03 a coarse sand, fine sand, silt, clay 25 b coarse sand, fine sand, silt, clay 22 c coarse sand, fine sand, silt, clay 34 d coarse sand, fine sand, silt, clay 19 04 a coarse sand, fine sand, silt, clay 0 b coarse sand, fine sand, silt, clay 0 c coarse sand, fine sand, silt, clay 0 d gravel, coarse sand, fine sand, silt, clay 0 discussion the distribution and dynamics of molluscan communities living in estuary ecosystems are strongly influenced by fluctuations of the physicochemical factors that induced by the mixing of freshwater and sea inflows. according to the results, few parameters varied along the stream. among ten parameters evaluated, five of them were (bod, salinity, alkalinity, hardness, and tds) gradually increased along the stream from head to the mouth. other four parameters (cod, do, ss, temperature) did not show any considerable variation. it could be assumed, that due to the mixing of saline and freshwater, the ionic strength of the water becomes increased towards to the mouth thus increased the salinity, alkalinity and hardness. towards to the mouth, the flow rate gradually decreases which promote the accumulation of organic matter and increase the total amount of suspended particles as well as the bod values. according to the current study, six parameters (cod, bod, do, salinity, temperature, hardness and faunus sp) showed significant difference among sites. however, according to the results of multiple comparisons of anova (turkey hsd), the 70 ruhuna journal of science,4,pp65-74 (2009) udayantha and munasinghe investigation of the factors that.... degrees of variation among sites were varied for measured parameters appendix 1). according to the regression analysis a considerable relationship between distribution of faunus sp. and the physicochemical parameters could not be determined (table 2). the substrate type was varied among the four sites (table 3). the common substrate type consisted with coarse sand, find sand, silt and clay. the highest number of individuals of genus faunus was recorded from the site 2 and none were recorded from the site 4. among studied parameters, six of them (cod, bod, do, salinity, temperature and hardness) were significantly different between site two and others (appendix 1). however this experimental area, especially site 1-3 are more vulnerable to accumulation of very fine particles that are released by the cement factory. slow flow rate of site 2 also caused high accumulation of silt materials on the substrate. the substrate that showed highest abundance of individuals (replicate 2c) consisted with fine sand, silt and clay. these observations conclude that substrate builds up with fine materials leads to accumulation of individuals in high numbers. therefore, it can be assumed substrate is one of the major factors that influence on the distribution of faunus sp along this stream. further it can be concluded that this species could be use as an indicator for the presence of substrate type, which consist with fine particles. acknowledgements we thank n.m.b.g. nishshanka, i.d.k.l. fernando and c.h. priyantha for giving help in field studies. this research was conducted as a partial fulfillment of b.sc. special degree program, which was supported by the department of zoology, university of ruhuna. references boening dw. 1999. an evaluation of bivalves as bio-monitors of heavy metals pollution in marine waters. environmental monitoring and assessment 55, 459-470. bresler v. 2003. marine mollusc in environmental monitoring. ii. experimental exposure to selected pollutants. helgoland marine research 57, 206-211. brown d s. 1991. freshwater snails of africa and their medical importance. taylor & francis, ltd., london. 487 pp. feldstein t. 2003. marine mollusc in environmental monitoring. iii. trace metals and organic pollutants in animal tissue and sediments. helgoland marine research 57, 212-219. hadrian, p. s., malcolm c. m. b., lindsay, g. r & michael, j. p. (1985). chemical and biological methods of water analysis for aqua culturists. sterling printed, great britain. lardicci c, rossi f, & castillo a. 1997. analysis of macrozoobenthic community structure after severe dystrophic crises in a mediterranean coastal lagoon. seddon m. 1886. molluscan biodiversity and the impact of large dams national museum and galleries of wales, uk co-chair iucn mollusc specialist group 71 udayantha and munasinghe ruhuna journal of science,4,pp55-74 (2009) investigation of the factors that.... subba rao nv. 1993. freshwater molluscs of india. in: roa k.s. (ed.). recent advances in freshwater biology. new delhi. anmol publication. volume 2. pp. 187-202. appendix appendix 01: multiple comparisons (turkey hsd) of physicochemical parameters among four sites. appendix 01 (i): multiple comparisons chemical oxygen demand (cod) levels among sites site 1 2 3 4 1 ** ** ** 2 ** --3 ** --4 ** --(*indicates the cod which significantly different among four sites) appendix 01(ii): multiple comparisons of biological oxygen demand (bod) levels among sites site 1 2 3 4 1 ** ** ** 2 ** --3 ** --4 ** --(*indicates the bod which significantly different among four sites) appendix 01 (iii): multiple comparisons dissolve oxygen (do) levels among sites 72 ruhuna journal of science,4,pp65-74 (2009) udayantha and munasinghe investigation of the factors that.... site 1 2 3 4 1 ** --2 ** --3 ---4 ---(*indicates the do which significantly different among four sites) appendix 01 (iv): multiple comparisons of salinity (sal) levels among sites site 1 2 3 4 1 --** 2 ---3 ---4 ** --(*indicates the sal which significantly different among four sites) appendix 01 (v): multiple comparisons of temperature (temp) among sites site 1 2 3 4 1 ** --2 ** --3 ---73 udayantha and munasinghe ruhuna journal of science,4,pp55-74 (2009) investigation of the factors that.... 4 ---(*indicates the temp which significantly different among four sites) appendix 08 (vi): multiple comparisons of hardness (hard) among sites site 1 2 3 4 1 -** ** 2 ---3 ** --4 ** --(*indicates the hard which significantly different among four sites) appendix 08 (vii): multiple comparaisons of abundance of faunus among sites site 1 2 3 4 1 ** --2 ** ** ** 3 -** -4 -** -(*indicates the abundance of faunus sp which significantly different among four sites) 74 vol 13 (2): 217-230, december 2022 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v13i2.127 university of ruhuna sri lanka faculty of science, university of ruhuna sri lanka 217 zooplankton community structure of a tropical lake in a northcentral state, nigeria elvis joefaith ebesi1 yakubu manbe mohammed1*, kate isioma iloba2, kabir mohammed adamu1, and solomon bake adama3 1department of biological sciences, ibrahim badamasi babangida university lapai, niger state, nigeria 2department of animal and environmental biology, delta state university, abraka, delta state, nigeria 3department of animal biology, federal university of technology minna, niger state, nigeria *correspondence: yakubmohammedmanbe@yahoo.com; orcid https://orcid.org/0000-0002-0890-7083 received: 23rd september 2021, revised: 8th february 2022, accepted: 4th september 2022 abstract. one of the biological components that play a vital role in the ecological integrity of an aquatic ecosystem is zooplankton. these organisms are good indicators of stressors, especially anthropogenic stressors in their environment. thus, there is a need to study the diversity in a diversified freshwater lake (dangana lake) in order to provide baseline information and knowledge on the status of the lake for proper and adequate water preservation and conservation. in addition to the zooplankton structure determination, the study also examined the physicochemical parameters of the lake. three sampling sites were identified with distinct characteristics where water and zooplankton samples were collected for physicochemical analyses and biotic identification using standard protocols. sampling was conducted for a duration of eight months. the monitored physicochemical parameters, i.e., temperature, dissolved oxygen (do), biochemical oxygen demand (bod5) and ph revealed no significant difference (p>0.05) among sampling sites. however, significance (p<0.05) amongst sampling sites was recorded in electrical conductivity, nitrate and phosphate. seasonally, there was a significant difference (p<0.05) among sampling months in all the monitored parameters. a total of ten species of zooplankton were recorded with four species each of rotifera and copepoda and two species of cladocera. the order of abundance of zooplankton population was rotifera (56.39%), copepoda (29.18%) and cladocera (14.42%). zooplankton population was significantly (p<0.05) higher during the dry season than in the rainy season. the canonical correspondence analysis (cca) revealed a strong correlation between the zooplankton and physicochemical parameters. ipso-facto, the lake exhibited spatiotemporal changes in zooplankton composition and physicochemical parameters during the sampling period. keywords: anthropogenic activities, dangana lake, physicochemical parameters, zooplankton https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v13i2.127 https://creativecommons.org/licenses/by-nc/4.0/ mailto:kabrmoh@yahoo.com https://orcid.org/0000-0002-0890-7083 elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 218 1 introduction freshwater ecosystems provide distinctive environments that improve ecological status and the survival of many biotic components (mohammed et al. 2021, oparaku et al. 2022). systematic monitoring of freshwater resources (rivers, streams, lakes, springs, and ponds) is required to ensure their socioeconomic function (mohammed et al. 2020). according to arimoro et al. (2015) there is global rise in freshwater environment research as it relates to the conservation of its inhabitants. thus, the contamination of freshwater ecosystems by various anthropogenic activities is becoming an important concern owing to the damage it poses to productivity, environmental sustainability, and social and economic development in africa (arimoro and keke 2016). the activities in the freshwater bodies such as laundry, agricultural operations and hydropower generation introduces pollutants that result in changes observed in the aquatic habitat (adamu et al., 2020; adamu et al., 2018; kun et al. 2015). freshwater bodies such as lakes and reservoirs are valuable resources that possess tremendous economic value as they provide many benefits to nearby settlements, such as means of flood control, recreation, aquatic life support, domestic water supply, irrigation and industrial water supply (maishanu et al., 2022; kdhe 2011). monitoring freshwater bodies such as lake helps to create information on species richness in the ecosystem, as well as information on the health status of the water body being studied (adamu et al., 2020; ajuzie 2012). zooplankton such as rotifers, copepods, and cladocerans are environmental stressor indicators and quick responders (pawlowski et al. 2016, xiong et al. 2019). they contribute to aquatic ecosystem biodiversity. according to xiong et al. (2020), zooplankton are widely accepted and indispensable bioindicators in the ecological conservation and management of aquatic ecosystems. despite their ecological importance, zooplankton richness and diversity are threatened (oparaku et al. 2022). this is exacerbated when variables impacting zooplankton populations remain unknown (mimouni et al. 2018), particularly in areas where zooplankton are regarded as species of low ecological significance (noble and hassall 2015). zooplankton communities are used as bioindicators of environmental changes because they have short life cycles and can respond quickly to shifts in physicochemical variables (kuczyńska-kippen et al. 2020, yuezhao et al. 2021, li and chen 2022). zooplankton also play a key role in the energy transfer between primary producers and other top trophic levels, and responds differently to environmental degradation (chiba et al. 2018, lomartire et al. 2021, makwinja et al. 2021). studies have revealed that anthropogenic activities and climate change are presently threatening the stability of most freshwater ecosystems (oparaku et al. 2022), resulting in a decrease in zooplankton diversity (alahuhta et al. 2019). therefore, understanding the impact of environmental drivers on zooplankton communities is critical in assessing the ecological integrity of a tropical lake elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 219 impacted by various anthropogenic activities. thereof, this study was aimed at investigating the composition, distribution pattern, abundance and community structure of the zooplankton fauna of dangana lake lapai, niger state, nigeria. 2 materials and methods 2.1 description of study area and sampling sites this study was carried out at dangana lake (fig. 1), lapai, lapai local government of niger state, nigeria. dangana lake is located within longitude 6°36129.61e and latitude 9°02112.02n with the elevation of 159 m above sea level. fig. 1: geographical location of dangana lake lapai, niger state, nigeria the area lies in the savannah region of central nigeria with mixed vegetation, the common ones of which include malaina (gmeilana arborea), locust beans (parkia biglobosa), neem (azadirachta indica) and other few native trees and grasses. the climate presents two distinct seasons, a rainy season between may to september, and a dry season (october-april) completely devoid of rain. the characteristics of the three sampling sites are shown in table 1. elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 220 table 1: sampling site characteristics of dangana lake during the sampling period. 2.2 sampling technique determination and analysis of physicochemical parameters water samples for physicochemical parameters were collected monthly between march and october, 2019 within the seasons: dry season (march to may) and rainy season (june to october) from the three selected sampling sites in the lake. the water temperature, dissolved oxygen (do), conductivity and ph were measured in-situ using a multipurpose meter (hanna model 1910), whereas, bod5, nitrate and phosphate were determined by the methods described in apha (2012). zooplankton sample collection and analysis samples of zooplankton were collected at the same time as sampling water for physiochemical analyses using a modified hand trawling approach. at each station, a 50 μm mesh size plankton net was horizontally trawled for zooplankton sampling for twenty minutes at four distinct points within the sampling sites as described in yagit (2006). using labeled plastic containers, samples were stored in a 4% formalin solution and sent to the laboratory for examination and identification. in the laboratory, 1 ml of each of the three samples taken per station was analysed. the plastics holding zooplankton samples were thoroughly shaken to achieve an even distribution of the zooplankton. thereafter, a drop of the 1 ml mixed samples was deposited into a glass slide coated with a coverslip for identification and enumeration using a compound microscope (mag: 10×10) and validated (mag: 10×40) by manually counting. this was done twice for each of the three plastics holding zooplankton samples, one for each plastic container (tash 1971). the specimens were identified by comparing them to plankton identification charts (needhem and needhem 1975, shiel 1995, fernando 2002, botes 2003, perry 2003, witty 2004). the zooplankton individuals recorded per station for the eight-month sampling period were added to give the total number of zooplankton individuals collected in the entire study period. sampling site description site a the entry point of water entry into the lake, where the lake receives its water sources from nearby wetlands and has few human activities taking place. site b this site is dominated with high anthropogenic activities. various human activities in this site include laundry, car wash and other domestic activities. site c this site is free of human activities; it has an opening which allows water to move out of the lake when there is excess water in the lake due to flood or excessive rainfall elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 221 2.3 statistical analyses descriptive statistics (mean, range, and standard deviation) for each physicochemical variable per station were calculated. the statistical differences in zooplankton metrics between stations and seasons were calculated using two-way analysis of variance (anova). the impact of stations and months on physicochemical characteristics were also calculated. tukey posthoc test was used to compare means. species abundance, species richness, taxonomic dominance, evenness index, simpson dominance index, shannon weiner index, and margalef's index of the zooplankton were calculated. canonical correspondence analysis, the link between physicochemical characteristics assessed and zooplankton species composition and abundance were determined. the monte-carlo permutation test with 1000 permutations was used to establish the cca significance. all statistical analyses were conducted using the paleontological statistical package (past) software, version 4.0. 3 results 3.1 physicochemical parameters the physicochemical parameters of sampling sites of dangana lake, lapai niger state for a period of eight months (march-october in 2019) is presented in table 2. there was no significant difference (p>0.05) in water temperature, do, bod5 and ph among all the sampling sites. however, electrical conductivity, nitrate and phosphate differ significantly (p<0.05) among sampling sites. table 2. mean summary of physicochemical parameters of dangana lake, lapai niger state nigeria. parameters stations probabilities site a site b site c stations months water temperature (°c) 26.41 ± 0.38a 26.65 ± 0.74a 23.50 ± 0.90a 0.72 5.81e-06 do (mg/l) 4.02 ± 0.10a 3.98 ± 0.15a 4.13 ± 0.20a 0.54 0.00 bod5 (mg/l) 3.00 ± 0.14a 3.08 ± 0.18a 3.12 ± 0.19a 0.37 3.40e-07 ph 6.88 ± 0.17a 7.05 ± 0.12a 6.98 ± 0.11a 0.13 1.29e-06 conductivity (µs/cm) 86.43 ± 8.57b 91.42 ± 9.04a 94.80 ± 8.54a 0.02 1.97e-08 nitrate (mg/l) 0.68 ± 0.08b 0.76 ± 0.08a 0.67 ± 0.09a 0.00 8.25e-12 phosphate (mg/l) 0.90 ± 0.10a 1.04 ± 0.12a 0.85 ± 0.10b 7.38e-05 3.51e-11 note: values are mean ± standard error; each parameter with the different superscript letters across sampling site showed significant differences (p<0.05, tukey post hoc test), while parameters with same superscript letters showed no significant differences (p>0.05). elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 222 there was a significant difference (p<0.05) among sampling months in all the physicochemical parameters measured. the water temperature recorded the highest mean value of 26.65 ± 0.74°c in sampling site b and the least mean water temperature value of 23.5 ± 0.90°c was recorded in site c. site b recorded the least mean value of 3.98 ± 0.15 mg/l in dissolved oxygen concentration whereas site c recorded the highest mean value of 4.13 ± 0.20 mg/l. mean bod was 3.00 ± 0.14 mg/l in site a and 3.12 ± 0.19 mg/l in site c. the mean ph mean value was lowest in site a (6.88 ± 0.17) and highest (7.05 ± 0.12) in site b. other monitored physicochemical parameters are presented in table 2. 3.2 zooplankton structural assemblage a total of ten species of zooplankton in three divisions were identified in dangana lake, lapai niger state. the groups encountered were copepoda (4 species), rotifera (4 species) and cladocera (2 species) (table 3). the percentage abundance of rotifera, copepoda and cladocera were 56.39, 29.18 and 14.42 respectively. table 3: zooplankton composition, distribution and percentage abundance in dangana lake, lapai during the sampling periods division species code site a site b site c % copepoda diaptomus gracilis dia 6 6 16 cyclops sp cyc 6 8 6 microcyclops varicans mic 10 7 6 mesocyclops sp zoc 7 8 3 total 29 29 31 29.18 rotifera keratella sp ker 29 20 31 lecane sp lec 5 21 9 notholca sp not 10 4 9 brachionus variabalis bra 11 6 17 total 55 51 66 56.39 cladocera daphnia sp dap 8 7 5 ceriodaphnia sp. cer 12 3 9 total 20 10 14 14.42 zooplankton distribution in all sampling sites recorded higher abundance in the dry season months (march to may) compared to the rainy season months (june to october) as presented in figure 2. elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 223 fig 2: monthly distribution of zooplankton in dangana lake 3.3 ecological indices of zooplankton in dangana lake lapai, niger state nigeria the diversity indices of zooplankton in each of the sampling sites are presented in table 4. there were 10 identified zooplankton in each of the sites with 104, 90, 111 individuals recorded in sites a, b and c respectively. shannon and evenness index followed a similar trend with site a>b>c. margalef’s index was highest in site b (2.00) and 1.988 and 1.91 for site a and c respectively. table 4: zooplankton diversity indices value of dangana lake lapai niger state, nigeria. indices site a site b site c taxa_s 10 10 10 individuals 104 90 111 dominance_d 0.1402 0.1437 0.1506 shannon_h 2.147 2.114 2.087 evenness_e^h/s 0.8561 0.8283 0.8057 margalef 1.938 2.00 1.911 3.4 association of zooplankton distribution and physicochemical parameters the total zooplankton compositions were observed to show an association with the mean physicochemical parameters measured of the lake as shown in fig 3. cca ordination axis 1 account for 71.73% of species variation with eigen value of 0.0905 while axis 2 showed 28.27% specie variation with eigen value of 0.0356 as elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 224 presented in table 5. the most significant association between physicochemical parameters was observed for biochemical oxygen demand, conductivity, ph, nitrate, phosphate and temperature. these parameters were strongly positively correlated with diaptomus gracilis, lecane sp., brachionus variabalis, keratella sp., cyclops sp., daphnia sp., microcyclops varicans, and mesocyclops sp. however, notholca sp. and ceriodaphnia sp. correlated negatively with the measured physicochemical parameters. fig 3: canonical correlation analysis ordination plot of physicochemical parameters and zooplankton species of dangana lake lapai, niger state table 5: weighted intra-set correlation of eigen value of cca axis. axis1 axis2 % 71.73 28.27 eigen value 0.090504 0.03566 do -0.74766 0.636934 bod5 0.129254 0.995598 ph 0.672748 0.763448 conductivity 0.050073 0.999897 nitrate 0.999159 -0.00527 phosphate 0.980827 -0.15969 elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 225 4 discussion 4.1 physicochemical parameters the monthly trend in the concentrations of do, bod5, ph, nitrate and phosphate in the lake shows a similar trend to that earlier reported in inland waters in nigeria (akindele et al. 2013, yusuf 2020, mohammed et al. 2021). all the monitored physicochemical parameters show spatial and temporal changes between sampling months, and this could be a result of some natural factors and different anthropogenic activities taking place around the lake (mustapha 2009). the water temperature of the lake falls within the normal temperature range of 20 to 30ºc required by aquatic organisms for metabolic activities (mustapha 2009). moderate water temperature was observed throughout the sampling period due to high rainfall and minimum amount of sunlight. the do and bod5 concentration observed in this study were high in the months which falls within the rainy season than in the dry season, this could be due to an increase in the volume of water in the rainy season as well as an increase in the flow of organic matters into the lake though surface runoff (davies et al. 2009, raji et al. 2015). water ph value obtained from dangana lake slightly falls within the recommended ph range of 6.5-8.5 for productivity in natural waters (raji et al. 2015). high ph and moderate temperature, on the other hand, have been reported to boost primary production in an aquatic environment, which supports zooplankton growth and survival (mustapha 2009). the high conductivity values observed during the rainy season could be attributed to a high amount of dissolved and suspended solid materials present in the water which results in an increase in dissolved cations such as calcium, magnesium and sulphate concentration in the sampling sites (arimoro and keke 2016). nitrates and phosphate are among the limiting factors in the aquatic environment (arimoro et al. 2015). the high mean phosphate value and nitrate value measured in the lake also improve primary production. the availability of nutrients impacts the shape and number of zooplankton (oparaku et al. 2022). an increase in phosphate and nitrate concentrations in aquatic environments has been related to anthropogenic activity; as a result, eutrophication is unavoidable (andong et al. 2019). it has shown that nutrients have a bigger influence on most freshwater biota than temperature (birk et al. 2020), but another study discovered that the effect of temperature on aquatic invertebrates was stronger under nutrient constraint (pomati et al. 2019). 4.2 zooplankton structural assemblage and diversity most of the zooplankton groups observed in this study appears to be the normal inhabitants of lakes, streams, ponds and artificial impoundment in the tropics elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 226 (mustapha 2009, arimoro and oganah 2010, tanko et al. 2016, iloba et al. 2019). zooplankton composition and abundance in any aquatic ecosystem are important in water quality monitoring. zooplankton populations are threatened due to different human activities such as domestic waste, agricultural waste such as runoff manure and fertilizers from nearby farms (anago et al. 2013). the differences in zooplankton distributions and assemblage in this study may be due to sampling space, time and the dynamic nature of aquatic systems (iloba 2019). changes in zooplankton populations in dangana lake could be attributed to different species having their nutritional requirement of different temperature, ph, do, bod5 and nutrients (chia et al. 2012). high nutrient level, high concentration of nitrate, and phosphate favor the growth of baccillariophyta which forms the major diet of zooplankton (chia et al. 2012). zooplankton are an important component in the dynamics of aquatic environment and its productivity (iloba 2019). plankton abundance and distribution are affected by seasonal changes of both physical and chemical parameters as well as phytoplankton in aquatic environments (norris and laws 2017, adamu et al. 2021). zooplankton were dominated by the rotifers in this study, and this could be attributed to their reproduction pattern which is parthenogenetic and short development rate under favorable aquatic condition (akin-ariola 2003, tanko et al. 2016). the low abundance of copepoda, and cladocera groups could be attributed to low abundance of aquatic macrophytes, which may increase the rate of predation by fishes in the lake (arimoro and oganah 2010, edegbene et al. 2022). the zooplankton diversity and abundance reported in this study is similar to the abundance and diversity pattern that is reported by akindele and olutona (2014) in the headstream water of aiba reservoir, iwo, nigeria. in this study, the monthly dynamics of the abundance and composition of zooplankton correlates with the season as higher populations were recorded during the dry season. the low abundance of zooplankton population observed in the rainy season could best be attributed to an influx of materials into the water body which may lead to further growth dilution of nutrients required for essential growth in the lake and surrounding environment (yusuf 2020). the taxonomic richness of zooplankton ecological parameters in this research region was rather high in number (9 zooplankton types) than that reported by abdul et al. (2016). the taxonomic distribution was also relatively even, in contrast to the eleven zooplankton taxa distribution in certain nigerian riverine systems (abdul et al. 2016). the high connection reported in the cca model between zooplankton and measured physicochemical parameters may imply that zooplankton taxonomic assemblages in the dangana lake were influenced by the majority of the prevailing environmental conditions. earlier research found that environmental factors had varying effects on zooplankton populations in nigerian rivers (arimoro and oganah 2009, abdul et al. 2016). however, when compared to other studies in nigeria, this study had lower taxonomic richness and evenness of zooplankton (arimoro and oganah 2009, abdul et al. 2016, edegbene et al. 2022). elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 227 5 conclusion the lake exhibits spatio-temporal changes in zooplankton composition and the monitored physicochemical parameters during the sampling period. anthropogenic activities throughout the sample locations altered the water quality and zooplankton assemblages and their organization. furthermore, the three taxonomic groupings of zooplankton in this study were cladocera, copepoda, and rotifera. zooplankton communities respond differently to environmental conditions. acknowledgments the authors are grateful to dangana community for their support and cooperation during the research work. anonymous reviewers are acknowledged for their comments. references abdul wo, adekoya eo, ademolu ko, omoniyi it, odulate do, akindokun te, olajide ae 2016. the effects of environmental parameters on zooplankton assemblages in tropical coastal estuary, southwest nigeria. egyptian journal of aquatic research 42: 281–28 7. adamu km, mohammed ym, mohammed h, ebesi ej. jimoh yo. 2021. spatio-temporal assessment of phytoplankton and physicochemical parameters of dangana lake, lapai, niger state nigeria. journal of applied life science international 24(12): 39-48. doi: 10.9734/jalsi/2021/v24i1230277 adamu km, muhammad h, ahmad su, ahmad mm, mohammed ym 2020. diversity of bacteria and fungi associated with freshwater fishes from mijawal river, nasarawa, nigeria. journal of applied sciences and environmental management 24(6): 1085-1092. doi:10.4314/jasem.v24i6.22 adamu km, adebola mo, iloba ik, jonathan jd, abdullahi as, onyemata ek, ikomi br 2018. are fungi isolated from water and fish samples in lapai-agaie dam, nigeria pathogenic to human health? production, agriculture and technology 14(2): 58-67. ajuzie cc. 2012. aspect of biodiversity studies in a small rural tropical reservoir (lamingo reservoir) in jos, nigeria. world rural observation 4(1): 23-33 akindele eo, adeniyi if, indabawa i. 2013. spatio-temporal assessment and water quality characteristics of lake tiga, kano, nigeria. research journal of environmental and earth sciences 5(2): 67-77. akindele eo, olutona go. 2014. water physicochemistry and zooplankton fauna of aiba reservoir headwater streams iwo, nigeria. journal of ecosystems 10(5): 54-65. akin-oriola ga. 2003. zooplankton association and environmental factors in ogupa and ona rivers nigeria. review of biology of the tropics 51(2): 391-398. alahuhta j, erlos t, kearnea om, soininen j, wang j, heino j. 2019. understanding environmental change through the lens of trait-based, functional, and phylogenetic biodiversity in freshwater ecosystems. environ review 27(2): 263-273. anago ij, esenowo ik, ugwumba aaa. 2013. the physico-chemical and plankton diversity of awba reservoir university of ibadan nigeria. research journal of environment and earth science 5(1): 638-644. andong fa, ezenwaji ne, melefa td, hinmikaiye ff, nnadi ov, oluwafemi o. 2019. assessment of the physical and chemical properties of lake oguta (nigeria) in relation to the water quality standard established by the nigerian federal ministry of water resources. advance oceanography and limnology 10:74–78 elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 228 apha (american public health association). standard methods for the examination of water and wastewater. wef and awwa. 20th edition; 1998: usa. arimoro fo, keke un. 2016. the intensity of human-induced impact on the distribution and diversity of macroinvertebrates and water quality of gbako river, north central nigeria. energy ecology and environment, 2: 143–154. arimoro fo, odume no, uhunoma si, edegbene ao. 2015. anthropogenic impact on water chemistry and benthic macroinvertebrate associated changes in a southern nigeria stream. environment monitoring assessments, 187: 1-14. arimoro fo, oganah ao. 2009. zooplankton community responses in a perturbed tropical stream in the niger delta, nigeria. the open environmental & biological monitoring journal, 3(1), 1–11. https:// doi. org/ 10. 2174/18750 40001 00301 0001 arimoro fo, oganah ao. 2010. zooplankton community response in a perturbed tropical stream in the niger delta nigeria. the open environment and biological monitoring journal, 3: 1-11. birk s, chapman d, carvalho l, spears bm, andersen he, argillier c, auer s, baattrup-pedersen a, banin l, beklio_glu m, et al. 2020. impacts of multiple stressors on freshwater biota across spatial scales and ecosystems. nature ecology and evolution 4(8):1060–1068. botes l. 2003. phytoplankton identification catalogue, saldanha bay south africa, globallast monograph series no7. imo london pp 10-86 chia am, adelanwa ma, ladan z, lortsuun dn, adanyi se, stephen bj. 2012. interaction of ipomoeaaquatica and ultricularia reflex a with phytoplankton densities in a small water bodies in northern nigeria. oceanology and hydrobiology studies 41(2): 39-47. chiba s, batten s, martin cs, ivory s, miloslavich p, weatherdon lv. 2018. zooplankton monitoring to contribute towards addressing global biodiversity conservation challenges. journal of plankton research 40(5): 509–518. https:// doi. org/ 10. 1093/ plankt/ fby030 davies oa, abowei jfn, tawari cc. 2009. phytoplankton community of elechi creek, niger delta, nigeria a nutrient polluted tropical creek. american journal of applied sciences 6(6): 11431152. edegbene ao, abdullahi y, akamagwuna fc, ogidiaka e, osimen ec, omovoh bo. 2022. are zooplankton useful indicators of ecological quality in afrotropical ephemeral stream impacted by human activities? environmental monitoring assessment 194: 399 https://doi.org/10.1007/s10661022-10061-4 fernando ch. 2002. a guide to tropical freshwater zooplankton; identification, ecology and impact on fisheries, backhuys publishers, leiden, the netherlands. iloba ki. 2019. water properties and zooplankton diversity of aghalokpe wetland in delta state, nigeria. science world journal 14(1): 164-170. iloba ki, akawo n, anani c. 2019. sand dredging impact on macrobenthic invertebrates of a hallowed river in delta state of nigeria. science world journal 14(1): 171-176 kdhe (kansas department of health and environment) 2011. water quality standard white papers. chlorophyll-a criteria for public water supply lakes or reservoirs. pp 1-12. retrieved from https://www.kdhe.ks.gov/documentcenter/view/13305/chlorophyll-a-criteria-for-public-watersupply-lakes-or-reservoirs-january-10-2011-pdf kuczyńska-kippen n, špoljar m, zhang c, pronin m. 2020. zooplankton functional traits as a tool to assess latitudinal variation in the northern-southern temperate european regions during spring and autumn seasons. ecological indicators 128: 106629. https:// doi. org/ 10. 1016/j. ecoli nd. 2020. 106629 kun l, chunguang h, jie z, zhenxing z, hongyong x, zhongqiang w, haijun y, lianxi s, 2015. long-term changes in water quality and macroinvertebrates communities of a subtropical river in south china. water 7: 63-80 li y, chen f. 2020. are zooplankton useful indicators of water quality in subtropical lakes with high human impacts? ecological indicators 113: 106167. https://doi.org/10.1016/j.ecolind.2020.106167. https://doi.org/10.1007/s10661-022-10061-4 https://doi.org/10.1007/s10661-022-10061-4 https://doi.org/10.1016/j.ecolind.2020.106167 elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 229 lomartire s, marques jc, goncalves amm. 2021. the key role of zooplankton in ecosystem services: a perspective of interaction between zooplankton and fish recruitment. ecological indicators 129: 107867. https:// doi. org/ 10. 1016/j. ecoli nd. 2021. 107867 maishanu f, adamu km, mohammed ym 2022. anthropogenic impact on some water quality characteristics of wupa river, federal capital territory, abuja, nigeria. issues in biological sciences and pharmaceutical research 10(3): 30-38. makwinja r, mengistou s, kaunda e, alamirew t. 2021. spatial distribution of zooplankton in response to ecological dynamics in tropical shallow lake: insight from lake malombe, malawi. journal of freshwater ecology 36(1): 127–147. https:// doi. org/ 10. 1080/ 02705 060. 2021.19430 19 mimouni ea, pinel-alloul b, beisner be, legendre p. 2018. summer assessment of zooplankton biodiversity and environmental control in urban waterbodies on the island of montreal ecosphere 9(7): e02277–19. mohammed ym, arimoro fo, ayanwale av, adama bs, keke un, auta yi, umar bl. 2020. seasonal changes in the abundance of benthic macroinvertebrates & physico-chemical condition of moussa stream bida, nigeria. tropical freshwater biology 29(1): 57-70. mohammed ym, arimoro fo, ayanwale av. adamu km, keke un, abubakar md, achebe ac. 2021. the current state of water quality and benthic invertebrate fauna in chikke stream (northcentral nigeria). ukranian journal of ecology 11(3): 26-34.doi: 10.15421/2021_136 mustapha mk. 2009. zooplankton assemblage of oyun reservoir, offa, nigeria. review biol trop 57(4): 1027–1047. needhem jg, needhem pr. 1975. a guide to the study of freshwater biology, san francisco, itodeday publishers. noble a, hassall c. 2015. poor ecological quality of urban ponds in northern england: causes and consequences. urban ecosystem 18(2):649–662. norris b, laws ea. 2017. nutrients and phytoplankton in a shallow hypereutic urban lake, prospects for restoration. water 9(431): 2-11. oparaku nf, andong fa, nnachi ia, okwuonu es, ezeukwu jc, ndefo jc. 2022. the effect of physicochemical parameters on the abundance of zooplankton of river adada, enugu, nigeria. journal of freshwater ecology 37(1): 33-56. doi: 10.1080/02705060.2021.2011793 pawlowski j, lejzerowicz f, apotheloz-perret-gentil l, visco j, esling p. 2016. protist metabarcoding and environmental biomonitoring: time for change. european journal protistology 55(pt a):12–25. perry r, 2003. a guide to the marine plankton of southern california, ucla ocean globe [online], 3rd edition. available at: http://www.msc.ucla.edu/oceanglobe [accessed on: 02/03/2019]. pomati f, shurin jb, andersen kh, tellenbach c, barton ad. 2019. interacting temperature, nutrients and zooplankton grazing control phytoplankton size-abundance relationships in eight swiss lakes. frontier in microbiology 10:3155–3172. raji mio, ibrahim yke, tytler ba, ehinmidu jo. 2015. physicochemical characteristics of water samples collected from river sokoto, northwestern nigeria. atmospheric and climate sciences 5: 194-199. shiel rj. 1995. a guide to identification of rotifers, cladocerans and copepods from australian inland waters. identification guide series no. 3. co-operative research centre for freshwater ecology, albury, nsw, 144 pp tanko d, chia ma, nuhu g. 2016. zooplankton community structure and dynamics during the transition from dry to rainy season: a case study of kufena inselberg (rock) pool zaria nigeria. the zoologist 14: 7-13 tash jc. 1971. the zooplankton of fresh and brackish water of the cape thompson area northern alaska. hydrobiologia, 38(1): 93-121. witty lm 2004. practical guide to identifying freshwater crustacean zooplankton. 2nd edition, cooperative freshwater ecology unit, sudbury, 50 p. elvis joefaith ebesi et al. zooplankton community in dangana lake, nigeria ruhuna journal of science vol 13 (2): 217-230, december 2022 230 xiong w, huang x, chen y, fu r, du x, chen x, zhan a. 2020. zooplankton biodiversity monitoring in polluted freshwater ecosystems: a technical review. environmental science and ecotechnology 1:1–11. xiong w, ni p, chen y, gao y, li s, zhan a. 2019. biological consequences of environmental pollution in running water ecosystems: a case study in zooplankton. environment and pollution 252: 1483–1490. yagit s. 2006. analysis of zooplankton community by the shannon-weaver index in kesikkopru dam lake turkey. tarim bilimeleri dergisi 12(2): 41-46. yuezhao l, chen h, song l, wu j, sun w, teng y. 2021. effects on microbiomes and resistomes and the source-specific ecological risks of heavy metals in the sediments of an urban river. journal of hazardous materials 409, 124472. https:// doi. org/ 10. 1016/j. jhazm at.2020. 124472 yusuf zh. 2020. phytoplankton as bioindicators of water quality in nasarawa reservoir, katsina state nigeria. acta limnologica brasiliensia 32(4): 17-29. vol 13 (2): 110-128, december 2022 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v13i2.119 university of ruhuna sri lanka faculty of science, university of ruhuna sri lanka 110 geochemical relations among elements in stream sediment samples from siojan prospecting area, iran using geostatistical methods aref shirazi1, adel shirazy1,*, seyed omid hosseini2 , ardeshir hezarkhani1, and amin beiranvand pour3,4 1faculty of mining engineering, amirkabir university of technology, tehran, iran 2 department of mine engineering, imam khomeini international university, qazvin, iran 3 institute of oceanography and environment (inos), universiti malaysia terengganu (umt), kuala nerus 21030, terengganu, malaysia 4 geoscience and digital earth centre (insteg), research institute for sustainable environment, universiti teknologi malaysia, johor bahru 81310, malaysia *correspondence: adel.shirazy@gmail.com; orcid https://orcid.org/0000-0001-7756-3205 received: 8th september 2021, revised: 3rd october 2022, accepted: 2nd december 2022 abstract stream sediment samples play an important role in identifying potential areas of metallic and non-metallic mineralization in mineral exploration studies. the relationship of geochemical elements with each other shows how the elements are distributed in the area. also, by identifying related elements, sampling and targeted chemical analysis can be used in the next stages of exploration. the purpose of this study is to investigate the elements related to the copper element in the siojan prospecting area, which is located in south-khorasan province and 30 km northwest of birjand city of iran. in siojan area, 120 stream sediment samples of a 60 square kilometer area were collected to detect geochemical anomalies and were consequently analyzed by inductively coupled plasma mass spectrometry (icp-ms) for 45 elements. preliminary geological studies showed that the studied area has copper mineralization potential, and therefore, copper was selected as the target element in this study. copper trace elements were identified in the area and the results were used to identify copper mineralized anomalies. for the elemental analysis data, methods of principal component analysis (pca), factor analysis (fa), hierarchical cluster analysis (hca) and k-means clustering were performed to identify the relevant elements and relationships among them. statistical analysis of the concentration of geochemical elements in the region revealed that copper and cobalt elements were identified as two elements of the same family in terms of geochemical genetics. the average value for copper and cobalt elements in the analyzed samples was 27.2 ppm and 15.5 ppm, respectively. finally, the relationship between copper and cobalt elements was modeled as an equation using the k-means clustering algorithm. keywords: anomaly separation, clustering, copper trace elements, pca, factor analysis 1 introduction geochemical investigations in mineral exploration are carried out to identify anomalous elements of mineralization (metallic or non-metallic) in the area (roonwal https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v13i2.119 https://creativecommons.org/licenses/by-nc/4.0/ aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 111 2018). in the reconnaissance stage, based on geological and geochemical investigations (sampling), the location of the anomalies of the target element is determined. the importance of this task is to separate the potential areas and barren areas from each other (gourley 2019, talapatra 2020). geochemical relationships between the elements from the stream sediment samples are analyzed by various methods. the most important and practical methods that have been used by researchers in different areas are multifractal analysis, principal component analysis (pca), clustering, artificial neural network (ann) and machine learning algorithms (yin et al. 2021, shirazi 2022, shirazi et al. 2022). each of the mentioned methods has advantages and disadvantages, which are chosen according to the study purpose (garcia et al. 2020, mohammadi et al. 2018). in the present study, the aim is to identify the elements related to the copper element in the exploratory area of siojan. for this purpose, principal component analysis (pca), factor analysis (fa), hierarchical cluster analysis (hca) and k-means clustering algorithm methods have been chosen. in this way, elemental analysis is performed among all analyzed data from stream sediment samples, and elements related to copper element are identified. the elements detected are called copper trace elements in the area. the siojan prospecting area is located in south-khorasan province (east iran) and 30 km northwest of birjand city. in terms of structural geology, this area is located in the structural zone of sistan (tirrul et al. 1983). the studied area has been confirmed in terms of copper mineralization, so that extensive copper oxide and copper sulfide mineralization have been observed in it. considering that the topic of the current research in this area has not been investigated comprehensively so far, the main objectives of the study are: (1) descriptive statistics analysis; (2) determining the geochemical threshold limit of copper element in stream sediment samples of the area; (3) investigating of geochemical relationships among elements; (4) separation of mineral elements and rock-forming elements in the area. 2 methodology 2.1 study area and geological setting the study region is located in 32° 58’ 23’’ n latitude and 58° 56’ 7’’ e longitude 30 km northwest of birjand city in south khorasan of iran (figure 1). siojan area geologically and structurally located in the philish zone of eastern iran. exposed rock units in this area include phyllis-like sediments (paleocene-eocene), limestones, aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 112 sedimentary-volcanic facies including marl, tuff and ignimbrite (eocene) and volcanic rocks (eoceneoligocene) including agglomerate, dacite, andesite, and ignimbrite. fig 1. location and geological map of the area alteration zones in this region are seen in relation to eocene-oligocene volcanic and around the location of faults and fractures. alteration zones include argillic alteration in the center and west of the area, siliceous in the center and parts of the south of the area, and carbonation (malekian et al. 2022). in general, the main structural control faults of the region can be divided into two general categories, which include, 1) northwest-southeast faults that these faults have a greater role in determining the structural status of the region than the second category faults and they often have a non-slip function, and 2) faults with north, northeast-south, southwest, which are more of normal slip slope. in addition to metal mineralization in the study area, small and scattered masses of gypsum can be seen in the south of the study area. fig 2. the location of stream sediment samples aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 113 2.2 sampling geochemical sampling in siojan area was prepared through stream sediment sampling method. the initial design of the sampling points was mainly based on determining the center of stream’s gravity. for this purpose, a map of the drainage systems (shirazi et al. 2022b) of the study area was generated using topographic maps and aerial images (figure 2). stream sediment samples with a particle size of 80 mesh were collected from the study area. the area was covered with 120 stream sediment samples in the form of icp ms 44 elements analysis (figure 2) (shirazi et al. 2018a, b, c). 2.3 technical flow the most important discussion in the analysis of geochemical data is to determine the background limit for each element in the study area and to separate the background from the anomalies of the relevant element (carranza 2008). another important issue in geochemical exploration is the simultaneous study of the elements under study (alahgholi et al. 2018, shirazy et al. 2021 a, b, zhang et al. 2019). one of the most powerful methods in this field is factor analysis as well as principal component analysis. these methods have two advantages (mohammadi et al. 2018): 1) reducing the dimensions of data, and 2) expressing the existing relationship between different elements. especially with the large number of elements studied and the large number of samples, the role of factor analysis becomes more apparent, so that it is much easier to understand the variability of the data . factor analysis is based on pca method. this is a technical method for finding a linear combination of the same initial variables to form a new coordinate system. these linear compounds are called principal component analysis and have the following properties (carranza 2009), 1) most of the variability can be explained by a limited number of new variables, 2) new variables, which are the product of the linear combination of the initial variables, do not show correlation between themselves. before using this method, it is necessary to pay attention to two points :if the initial variables are not correlated (or have a small correlation coefficient) there is no reason to use this method, because they do not give acceptable results .factor analysis is performed when the number of initial variables is sufficient . invoice analysis is performed in four steps : 1) calculation of correlation coefficients, 2) extraction of factors, which includes determining the number and method of calculating factors, 3) applying special transformations on factors, in order to better interpret the relationships between data, 4) calculating the score of each factor for each sample . clustering is the placement of data in groups where the members of each group are similar to each other on a specific parameter. the similarity between the data within each cluster is maximum and the similarity between the data within different clusters is minimal (doodran et al. 2020; shirazy et al. 2022, yang et al. 2019) . the structure of these clusters or groups can be consistent with the nature of the data or the hidden aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 114 structure that lies within the data. clustering is finding a structure in a set of data that is not categorized. the main reason for using clustering methods is to discover new structures that exist naturally in the data without any prior knowledge of the structure of classes or categories (nazerian et al. 2021, shirazy et al. 2018 d, e, yu et al. 2019). the results of geochemical studies can be useful in future studies in the field of mineral exploration (beygi et al. 2021, shirazi et al. 2018b). for example, in remote sensing surveys and processing of satellite images, they are used in combination to verify the mapping performed (shirazi et al. 2018 b, shirazy et al. 2021 c,d, shirazy et al. 2020 a,b). it is also used in exploratory geophysical studies to integrate information layers and determine the drilling location of boreholes (khayer et al. 2020; shirazy et al. 2021). 3 results and discussion 3.1 descriptive statistical investigation due to the polymetallicity of the region, the correlation of elements (cu, co, ag, sb, mo, zn) (table 1) were identified for analysis. because our data are not normal, spearman method was used to determine the correlation. table 1: elements and their correlation with spearman method. co cu mo sb zn ag co 1 .586 .076 .369 .204 .097 cu .586 1 .141 .385 .419 .057 mo .076 .141 1 .448 .131 .199 sb .369 .385 .448 1 .255 .195 zn .204 .419 .131 .255 1 .080 ag .097 .057 .199 .195 .080 1 the statistical results on the data are given in table 2 . table 2: basic statistical results. min max sum mean std. deviation variance skewness kurtosis ag 0 1 42 0.35 0.202 0.041 1.298 0.221 2.659 0.438 co 9 23 1860 15.50 2.961 8.770 0.013 0.221 0.080 0.438 cu 15 50 3272 27.27 5.391 29.066 0.450 0.221 1.490 0.438 mo 0 3 122 1.01 0.422 0.178 1.764 0.221 3.789 0.438 sb 0 7 126 1.05 1.013 1.026 3.123 0.221 12.336 0.438 zn 30 68 5417 45.14 8.063 65.015 0.419 0.221 -0.579 0.438 aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 115 3.2 copper threshold limit investigation the most important discussion in the analysis of geochemical data is to determine the background threshold for each element in the region and to separate the background threshold from the anomaly of the relevant element (cheng et al. 1996, khosravi et al. 2022, shirazy et al. 2020 c,d, shirazi et al. 2018 a, c, shirazy et al. 2022 a ,b). using statistical methods, the threshold concentration of geochemical elements is identified. to be used in geochemical modelling and geochemical halo identification. identifying the concentration threshold of the desired elements is used in mineral processing studies to decide on the economics and how the elements are extracted from the mineral (doodran et al. 2020, khakmardan et al. 2020b, khakmardan et al. 2018). anomaly threshold calculator formula: according to the above formula: mad is 3.1 ppm and median is 27.1 ppm as well as after insertion in the formula : anomaly = median + 3 mad our anomalous threshold was estimated to be 33.1 ppm and our anomalous map was created as figure. blue indicates anomaly and background is green and other colors indicate anomalous probabilities (mohammadi et al. 2016). fig 3. anomalous areas with blue and green is background (coordination in utm, zone : 39) (1) sxam 2+= (2) sxae 3+= (3) || meanximeanmad −= (4) || medianximedianmad −= aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 116 3.3 principal component analysis (pca) according to the selected elements, the analysis was performed on the data and the results were announced according to table 3 . table 3. principal component analysis table. variable pc 1 pc 2 pc 3 pc 4 pc 5 pc 6 cu 0.513 -0.371 -0.048 0.04 0.296 0.712 co 0.46 -0.345 -0.247 0.461 0.143 -0.613 ag 0.201 0.521 -0.811 -0.157 0.024 0.072 sb 0.492 0.261 0.236 0.204 -0.763 0.103 mo 0.318 0.603 0.462 0.053 0.553 -0.117 zn 0.376 -0.201 0.1 -0.847 -0.062 -0.296 eigen value 2.3101 1.1922 0.8529 0.8148 0.4707 0.3592 proportion 0.385 0.199 0.142 0.136 0.078 0.6 cumulative 0.385 0.584 0.726 0.862 0.94 1 in the table 3, according to the amount of special values and also the following diagram (figure 4), the first three components that explain 73% of the behavior were selected . fig 4. scree plot in pca the outlier data were announced by principal component analysis method with mahalanobits interval as shown in figure 5. aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 117 fig 5. outlier data detection based on mahalanobis distance. outlier data were replaced with appropriate values. then the corrected data were used for statistical analysis (levin 2011, kim 2013, kwak and kim 2017, alahgholi et al. 2018, aali et al. 2022a, b, shirazy et al. 2022a, b, c, d, shirazi et al. 2022a, b). based on the principal component analysis (pca) method, the scores of different elements in each principal component were evaluated. then the scores of the elements were mapped in each of the first, second and third principal components (pc1, pc2 and pc3). the map obtained from each of the first, second and third principal components are presented in figures 6, 7 and 8, respectively. fig 6. contour map of principal component analysis method, first component (pc1) (coordination in utm, zone: 39) aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 118 fig 7. contour map of principal component analysis method, second component (pc2) (coordination in utm, zone: 39) fig 8. contour map of principal component analysis method, third component (pc3) (coordination in utm, zone: 39) the studied elements were clustered using principal component analysis (pca) method. for this purpose, the loads related to each of the elements in the first and second main components were plotted against each other. elements close to each other were grouped together. based on this clustering, 3 different group classes were obtained. silver and molybdenum elements group, zinc, cobalt and copper elements group and antimony single element group (figure 9). aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 119 fig 9. clustering by pca method using components one and two. 3.4 factor analysis (fa) factor analysis is a multivariate statistical method that establishes a special relationship between a large set of seemingly unrelated variables under a hypothetical model. in factor analysis, a large number of variables are expressed in terms of a small number of dimensions or structures; this structure is called a factor. in methods based on eigenvectors using eigenvalues and eigenvectors, directions with maximum variability are identified. in order to analyze the geochemical data of stream sediments, the rotary type factor analysis method with variomex rotation was used (reimann et al. 2005, shirazy et al. 2018a, b, c, shirazy et al. 2019, shirazi et al. 2021, hedayat et al. 2022, shirazy et al. 2022c). in table 4, by selecting 3 factors and also the method of maximum likelihood, the numbers for each factor in each element are shown in figures 10 to 12. table 4. factor analysis table. variable factor 1 factor 2 factor 3 communality cu 0.63 -0.35 0.398 0.678 co 0.979 0.057 -0.25 0.962 ag 0.106 -0.213 -0.177 0.088 sb 0.405 -0.574 -0.179 0.526 mo 0.104 -0.599 -0.344 0.488 zn 0.239 -0.383 0.338 0.318 variance 1.5979 1.0056 0.4555 3.0589 % var 0.266 0.168 0.076 0.51 aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 120 fig 10. contour map of factor one (coordination in utm, zone: 39) fig 11. contour map of factor two (coordination in utm, zone: 39). fig 12. contour map of factor three (coordination in utm, zone: 39) aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 121 according to the loads in the first two factors, figure 13 can be drawn. fig 13. factor classification according to factor one and two. or in the case of factors one and two, it is shown schematically in figure 14. fig 14. factor classification according to factor one and two schematically. based on the obtained results of factor analysis method, another grouping of the studied elements can be presented. this grouping includes cobalt, copper and zinc groups, silver and molybdenum groups and single element antimony group. 3.5 hierarchical cluster analysis (hca) in data mining and statistics, "hierarchical clustering" is a method that performs the act of categorizing and grouping observations and data in a hierarchical manner. the point that distinguishes this method from other clustering methods is the sequence and topdown (or bottom-up) view that exists in this technique (khakmardan et al. 2020a; khayer et al. 2021; madani et al. 2021; nazerian, catania, et al. 2022; nazerian, aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 122 shirazy, et al. 2022). for clustering, the hierarchical distance clustering method was used and the results were presented according to table 5 and the dendrogram in figure 15. table 5: hierarchical clustering parameters. agglomeration schedule stage cluster combined coefficients stage cluster first appears next stage cluster 1 cluster 2 cluster 1 cluster 2 1 3 5 74.721 0 0 2 2 3 4 136.962 1 0 4 3 1 2 18884.820 0 0 4 4 1 3 57148.272 3 2 5 5 1 6 177736.097 4 0 0 fig 15. final hierarchical clustering dendrogram. geochemical elements can be grouped according to the results of hierarchical analysis. in this grouping, silver, molybdenum and antimony elements are in one group, copper and cobalt elements in one group and zinc element as a single element in a separate group. 3.6 k-means clustering analysis as it was observed, according to the results of statistical analysis, copper and cobalt elements are inherently geochemically related to each other. therefore, the geochemical behavior of these two elements in relation to each other (as an exploratory key in the study region) (shirazy et al. 2019), is important. in order to investigate the geochemical behavior of cu and co elements, k-means clustering method was used khayer et al. 2020, shirazy et al. 2022c). * * * * * h i e r a r c h i c a l c l u s t e r a n a l y s i s * * * * * * dendrogram using average linkage (between groups) rescaled distance cluster combine c a s e 0 5 10 15 20 25 label num +---------+---------+---------+---------+---------+ ag 3 ─┐ mo 5 ─┼───────────────┐ sb 4 ─┘ ├───────────────────────────────┐ cu 1 ─────┬───────────┘ │ co 2 ─────┘ │ zn 6 ─────────────────────────────────────────────────┘ aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 123 classification of 3 classes, with an average value of 0.5237 classification of 4 classes, with an average value of 0.5556 classification of 5 classes, with an average value of 0.5468 classification of 6 classes, with an average value of 0.4677 classification of 7 classes, with an average value of 0.4471 classification of 8 classes, with an average value of 0.4975 fig 16. profile images obtained from the classification of cu and co elements along with the mean values of the utility function. initially, the data set was categorized into different classes using the k-means algorithm. then the amount of utility function for each classification was calculated. as shown in figure 17, the image of the silhouettes is presented along with the mean value of the utility function. based on the results of geochemical data classification of copper and cobalt elements, classification of 4 classes with the highest value of the utility function was selected as the optimal clustering (ahmadi et al. 2022, shirazy and hezarkhani 2018, aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 124 shirazy et al. 2022; shirazy et al. 2021). the graph of the values of the utility function for classifying 3 classes to 20 classes is presented in figure 17. fig 17. the graph of the values of the utility function (silhouette) for classifying 3 classes to 20 classes. fig 18. geochemical behavior of copper (cu) vs. cobalt (co) elements. cu and co concentrations were calculated in the centers of 4-class classification based on k-means algorithm. using regression method, the values of the centers of the clusters were plotted. the diagram shown in figure 18 is actually the geochemical behavior of copper and cobalt elements in samples of stream sediments in the siojan region. as can be seen in the geochemical behavior diagram of the elements copper and cobalt, the concentrations of these two elements are directly related to each other. thus, with increasing cu concentration, co concentration also increases. this increase in value is calculated as a quadratic curve. the third-degree equation of geochemical behavior of copper and cobalt concentrations is: cu = 0.272co3 13.345co2 + 218.47co 1163.9 4 conclusions siojan exploration region is located in south khorasan province of iran with the mineralization potential of copper element. copper mineralization has occurred in this area. therefore, it was used as a target for studies to identify geochemical anomalies. aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 125 120 geochemical samples of stream sediments were collected according to the designed sampling map. after preparing the samples, they were analyzed by icp-ms method. from each sample, 44 elements were obtained as a result of analysis . data were prepared using statistical methods. then, in order to identify the relationships of geochemical elements in the region, the methods of principal component analysis, factor analysis and hierarchical analysis were used. the grouping performed on the geochemical elements by the methods of principal component analysis and factor analysis found similar results. based on these results, three groupings of geochemical elements of the region were presented. cu, co and zn group, ag and mo group and single element sb group. according to the same result, the methods of principal component analysis and factor analysis in the grouping of geochemical elements in the region, in order to increase the accuracy of research and identify elements related to the element, hierarchical analysis was used. based on this method, geochemical elements showed different groupings. this grouping also had similarities with the grouping obtained from the results of other methods. according to this grouping, the elements ag, mo and sb in one group, cu and co elements in one group and zn as a single element in a separate group. due to the fact that the target element in this study was copper, based on the results of statistical studies, the genetic family of this element in siojan region is cobalt. although the element zinc was also included in the two groups, which was part of the copper group family by the method of principal component analysis and factor analysis, but it cannot be mentioned as an intrinsic relationship in the region. another reason is the inherent and permanent correlation between zinc and lead, which was not achieved in statistical analysis . geochemical behavioral studies of cu and co elements based on k-means algorithm show that these two elements have direct behavior with each other. behavioral changes in the concentration of elements relative to each other were calculated as a quadratic equation. using this equation, the concentration of each element can be calculated based on the other element. the calculated equation is cu = 0.272co3 13.345co2 + 218.47co 1163.9. acknowledgements we appreciate the department of mining and metallurgy engineering amirkabir university of technology (tehran polytechnic). the institute of oceanography and environment (inos), universiti malaysia terengganu (umt) and research institute for sustainable environment, universiti teknologi malaysia are also acknowledged for providing facilities during editing, rewriting, re-vising, and re-organizing the manuscript. great appreciation should go to anonymous journal reviewers/ editors for their constructive comments on the primary version of the manuscript. aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 126 references aali aa, shirazi a, shirazy a, khakmardan s, maghsoudi a, hezarkhani a. 2022a. geophysical study to identify iron mineralization anomalies using terrestrial magnetometry in the chak-chak exploration area, iran. türkiye jeoloji bülteni 65(2): 159-170. aali aa, shirazy a, shirazi a., pour ab, hezarkhani a, maghsoudi , hashim m, khakmardan s. 2022b. fusion of remote sensing, magnetometric, and geological data to identify polymetallic mineral potential zones in chakchak region, yazd, iran. remote sensing 14(23): 6018. ahmadi me, mesroghli s, hedayat b, nazerian h, shirazi a, shirazy a. 2022. assessment of the influence of sulfuric acid/hydrogen peroxide mixture on organic sulfur reduction of high sulfur coals and their chemical composition. open journal of geology 12(3): 199-214. alahgholi s, shirazy a, shirazi a. 2018. geostatistical studies and anomalous elements detection, bardaskan area, iran. open journal of geology 8(7): 697-710. beygi s, talovina iv, tadayon m, pour ab. 2021. alteration and structural features mapping in kachomesqal zone, central iran using aster remote sensing data for porphyry copper exploration. international journal of image and data fusion 12(2): 155-175. carranza ejm. 2008. geochemical anomaly and mineral prospectivity mapping in gis. elsevier. carranza ejm. 2009. controls on mineral deposit occurrence inferred from analysis of their spatial pattern and spatial association with geological features. ore geology reviews 35(3-4): 383-400. cheng q, agterberg f, bonham-carter g. 1996. a spatial analysis method for geochemical anomaly separation. journal of geochemical exploration 56(3): 183-195. doodran rj, khakmardan s, shirazi a, shirazy a. 2020. minimalization of ash from iranian gilsonite by froth flotation. journal of minerals and materials characterization and engineering 9(1): 1-13. garcia rjl, da silva júnior jb, abreu im, soares sar, araujo rgo, de souza es, ribeiro hjs, hadlich gm, de souza queiroz af. 2020. application of pca and hca in geochemical parameters to distinguish depositional paleoenvironments from source rocks. journal of south american earth sciences 103: 102734. gourley ac. 2019. key elements of a model mining code: a middle east case study. mineral economics 32(2): 187-204. hedayat b, ahmadi me, nazerian h, shirazi a, shirazy a. 2022. feasibility of simultaneous application of fuzzy neural network and topsis integrated method in potential mapping of lead and zinc mineralization in isfahan-khomein metallogeny zone. open journal of geology 12(3): 215-233. khakmardan s, doodran rj, shirazy a, shirazi a, mozaffari e. 2020a. evaluation of chromite recovery from shaking table tailings by magnetic separation method. open journal of geology 10(12): 1153-1163. khakmardan s, doodran rj, shirazy a, shirazi a, mozaffari e. 2020b. evaluation of chromite recovery from shaking table tailings by magnetic separation method. open journal of geology 10: 11531163. khakmardan s, shirazi a, shirazy a, hosseingholi h. 2018. copper oxide ore leaching ability and cementation behavior, mesgaran deposit in iran. open journal of geology 8(09): 841. khayer k, shirazy a, shirazi a, ansari a, hezarkhani a. 2020. cementation exponent estimate in carbonate reservoirs: a new method. global journal of computer sciences: theory and research 10(2): 66-72. khayer k, shirazy a, shirazi a, ansari a, nazerian h, hezarkhani a. 2021. determination of archie’s tortuosity factor from stoneley waves in carbonate reservoirs. international journal of science and engineering applications (ijsea) 10: 107-110. khosravi v, shirazi a, shirazy a, hezarkhani a, pour ab. 2022. hybrid fuzzy-analytic hierarchy process (ahp) model for porphyry copper prospecting in simorgh area, eastern lut block of iran. mining 2(1): 1-12. kim hy. 2013. statistical notes for clinical researchers: assessing normal distribution (2) using skewness and kurtosis. restorative dentistry & endodontics 38(1): 52. kwak sk, kim jh. 2017. statistical data preparation: management of missing values and outliers. korean journal of anesthesiology 70(4): 407. levin ri. (2011). statistics for management. pearson education india. aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 127 madani n, maleki m, sepidbar f. 2021. application of geostatistical hierarchical clustering for geochemical population identification in bondar hanza copper porphyry deposit. geochemistry 81(4): 125794. shabani a, ziaii m, monfared ms, shirazy a, shirazi a. 2022. multi-dimensional data fusion for mineral prospectivity mapping (mpm) using fuzzy-ahp decision-making method, kodeganbasiran region, east iran. minerals 12(12): 1629. mohammadi nm, hezarkhani a, maghsoudi a. 2018. application of k-means and pca approaches to estimation of gold grade in khooni district (central iran). acta geochimica 37(1): 102-112. mohammadi nm, hezarkhani a, saljooghi bs. 2016. separation of a geochemical anomaly from background by fractal and u-statistic methods, a case study: khooni district, central iran. geochemistry 76(4): 491-499. nazerian h, shirazy a, shirazi a, hezarkhani a. 2022. design of an artificial neural network (bpnn) to predict the content of silicon oxide (sio2) based on the values of the rock main oxides: glass factory feed case study. international journal of science and engineering applications (ijsea) 2(11): 41-44. malekian m, momayez m, bellett p, mancino s, carrea f. 2022. stability investigation of open-pit slopes during blasting activities. international journal of engineering research & technology. nazerian h, shirazy a, shirazi a, hezarkhani a. 2022. design of an artificial neural network (bpnn) to predict the content of silicon oxide (sio2) based on the values of the rock main oxides: glass factory feed case study. international journal of science and engineering applications (ijsea) 2: 41-44. reimann c, filzmoser p, garrett rg. 2005. background and threshold: critical comparison of methods of determination. science of the total environment 346(1-3): 1-16. roonwal gs. 2018. mineral exploration: practical application. springer. shirazi a. 2022. integration of intelligent analytical methods for geochemical modeling and identification of exploration keys for the sahlabad [applied research, amirkabir university of technology (tehran polytechnic)]. tehran, iran. shirazy a, shirazi a, hezarkhani a. 2018a. predicting gold grade in tarq 1: 100000 geochemical map using the behavior of gold, arsenic and antimony by k-means method. journal of mineral resources engineering 2(4): 11-23 shirazi a, hezarkhani a, beiranvand pour a, shirazy a, hashim m. 2022a. neuro-fuzzy-ahp (nfahp) technique for copper exploration using advanced spaceborne thermal emission and reflection radiometer (aster) and geological datasets in the sahlabad mining area, east iran. remote sensing 14(21): 5562. shirazi a, hezarkhani a, pour ab. 2022b. fusion of lineament factor (lf) map analysis and multifractal technique for massive sulfide copper exploration: the sahlabad area, east iran. minerals 12(5) 549. shirazi a, hezarkhani a, shirazy a. 2018a. exploration geochemistry data-application for cu anomaly separation based on classical and modern statistical methods in south khorasan, iran. international journal of science and engineering applications 7(4): 39-44. shirazi a, hezarkhani a, shirazy a. 2018b. remote sensing studies for mapping of iron oxide regions, south of kerman, iran. international journal of science and engineering applications 7(4): 45-51. shirazi a, shirazy a. 2020. introducing geotourism attractions in toroud village, semnan province, iran. international journal of science and engineering applications 9(16): 79-86. shirazi a, shirazi a, hezarkhani a. 2022c. advanced integrated methods in mineral exploration. lap lambert academic publishing. shirazi a, shirazy a, karami j. 2018c. remote sensing to identify copper alterations and promising regions, sarbishe, south khorasan, iran. international journal of geology and earth sciences 4(2): 36-52. shirazi a, shirazy a, nazerian h, khakmardan s. 2021. geochemical and behavioral modeling of phosphorus and sulfur as deleterious elements of iron ore to be used in geometallurgical studies, sheytoor iron ore, iran. open journal of geology 11(11): 596-620. aref shirazi et al. stream sediment analysis from siojan prospecting area, iran ruhuna journal of science vol 13 (2): 110-128, december 2022 128 shirazi a, shirazy a, saki s, hezarkhani a. 2018d. geostatistics studies and geochemical modeling based on core data, sheytoor iron deposit, iran. journal of geological resource and engineering 6: 124133. shirazi a, shirazy a, saki s, hezarkhani a. 2018e. introducing a software for innovative neuro-fuzzy clustering method named nfcmr. global journal of computer sciences: theory and research 8(2): 62-69. shirazy a, hezarkhani a, shirazi a, khakmardan s, rooki r. 2022a. k-means clustering and general regression neural network methods for copper mineralization probability in chahar-farsakh, iran. türkiye jeoloji bülteni 65(1): 79-92. shirazy a, hezarkhani a, shirazi a, timkin t, voroshilov vg. 2022b. geophysical explorations by resistivity and induced polarization methods for the copper deposit, south khorasan, iran. bulletin of the tomsk polytechnic university geo assets engineering 333(3): 99-110. shirazy a, hezarkhani a, timkin t, shirazi a. 2021a. investigation of magneto-/radio-metric behavior in order to identify an estimator model using k-means clustering and artificial neural network (ann)(iron ore deposit, yazd, iran). minerals 11(12): 1304. shirazy a, shirazi a, ansari a, hezarkhani a. 2020a. cementation exponent estimate in carbonate reservoirs: a new method. global journal of computer sciences: theory and research 10(2): 66-72. shirazy a, shirazi a, ferdossi mh, ziaii m. 2019. geochemical and geostatistical studies for estimating gold grade in tarq prospect area by k-means clustering method. open journal of geology 9(6): 306326. shirazy a, shirazi a, heidarlaki s, ziaii m. 2018b. exploratory remote sensing studies to determine the mineralization zones around the zarshuran gold mine. international journal of science and engineering applications 7(9): 274-279. shirazy a, shirazi a, hezarkhani a. 2018c. predicting gold grade in tarq 1: 100000 geochemical map using the behavior of gold, arsenic and antimony by k-means method. journal of mineral resources engineering 2(4): 11-23. shirazy a, shirazi a, hezarkhani a. 2020b. behavioral analysis of geochemical elements in mineral exploration:-methodology and case study. lap lambert academic publishing. shirazy a, shirazi a, nazerian h. 2021b. application of remote sensing in earth sciences–a review. international journal of science and engineering applications 10(5): 45-51. shirazy a, shirazi a, nazerian h, hezarkhani a. 2021c. investigation of geochemical sections in exploratory boreholes of mesgaran copper deposit in iran. international journal for research in applied science and engineering technology 9(8): 2364-2368. shirazy a, shirazi a, nazerian h, khayer k, hezarkhani a. 2021d. geophysical study: estimation of deposit depth using gravimetric data and euler method (jalalabad iron mine, kerman province of iran). open journal of geology 11(8): 340-355. shirazy a, ziaii m, hezarkhani a. 2020c. geochemical behavior investigation based on k-means and artificial neural network prediction for copper, in kivi region, ardabil province, iran. iranian journal of mining engineering 14(45): 96-112. shirazy a, ziaii m, hezarkhani a, timkin t. 2020d. geostatistical and remote sensing studies to identify high metallogenic potential regions in the kivi area of iran. minerals 10(10): 869. talapatra ak. (2020). geochemical exploration and modelling of concealed mineral deposits. springer. r. tirrul, i. bell, r. griffis, v. camp. 1983. the sistan suture zone of eastern iran. geological society of america bulletin 94(1): 134-150. yang j, grunsky e, cheng q. 2019. a novel hierarchical clustering analysis method based on kullback– leibler divergence and application on dalaimiao geochemical exploration data. computers & geosciences 123: 10-19. yin b, zuo r, xiong y, li y, yang w. 2021. knowledge discovery of geochemical patterns from a datadriven perspective. journal of geochemical exploration 231: 106872. zhang s, xiao k, carranza ejm, yang f, zhao z. 2019. integration of auto-encoder network with densitybased spatial clustering for geochemical anomaly detection for mineral exploration. computers & geosciences 130: 43-56. ruhuna journal of science vol 13 (2): 167-180, december 2022 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v13i2.123 university of ruhuna faculty of science, university of ruhuna sri lanka 167 process optimization for the development of easy-to-cook jackfruit bulbs chalani akmeemana*, indira wickramasinghe, piyumi chathurangi wanniarachchi and thisum vithanage department of food science and technology, faculty of applied sciences, university of sri jayewardenepura, nugegoda, sri lanka *correspondence: chathurangichalani@sci.sjp.ac.lk, orcid: https://orcid.org/0000-0002-1832-085x received: 9th july 2022, revised: 15th november 2022, accepted: 29th december 2022 abstract overproduction of jackfruit (artocarpus heterophyllus) in harvesting seasons and its short post-harvest shelf life make it an underutilized fruit species. the present study was conducted to analyze the feasibility of preparing processed jackfruit products to enhance utilization and reduce wastage. jackfruit bulbs were cut into 4.5 cm ± 0.5 cm in length and 0.5 cm ± 0.3 cm in width pieces. the twofactor factorial technique was used as the experimental design when developing the products. as factors, the freezing operation and dipping treatment were considered. hot water extracts of a. heterophyllus leaves and trachyspermum ammi leaves were used as two dipping treatments and sensory evaluation tests were carried out to determine which dipping treatment is better. t. ammi dipping treatment had better sensory acceptance, and hence continued as the selected dipping method. the effect of freezing (-18 °c for 12 h), pre-cooking time, and rehydration on cooking time was also investigated. frozen and non-dipped jackfruit bulb pieces had the highest sensory acceptance, hence its proximate composition, dpph, total phenolic acid, and colourimetric values were determined. moisture, ash, fat, crude protein and carbohydrate content of the frozen non-dipped jackfruit bulbs were 8.08 ± 0.28 %, 0.88 ± 0.01 %, 0.31 ± 0.03 %, 3.93 ± 0.06 %, 81.09 ± 0.54 % respectively. the colourimetric values (l*, a*, and b*) of the frozen non-dipped jackfruit bulbs were 79.40 ± 0.61, 1.87 ± 0.35, 25.43 ± 1.33, 25.50 ± 1.35 respectively. the results revealed that the freezing operation significantly (p<0.05) reduce the cooking time of jackfruit bulbs. based on the sensory evaluation results, jackfruit bulbs that were frozen without dipping the t. ammi leaf extract were identified as having the best sensory acceptability. the final processed product possesses a good nutritional profile concerning the ash, fat, crude protein, carbohydrate, mineral content, dpph and total phenolic acid content. keywords: easy to cook food, jackfruit bulb, proximate composition. https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v13i2.120 https://creativecommons.org/licenses/by-nc/4.0/ https://orcid.org/0000-0002-1832-085x http://orcid.org/0000-0003-1805-4923 htpp://orcid.org/0000-0003-3364-6433 https://orcid.org/0000-0002-5495-6724 https://orcid.org/0000-0002-1832-085x https://orcid.org/0000-0002-1832-085x c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 168 1 introduction jackfruit, artocarpus heterophyllus belongs to the family moraceae and is mainly grown in tropical and sub-tropical regions (vazhacharickal et al. 2015). a mature jackfruit tree often yields between 10 200 fruits every year. normally it takes around three to seven months for the development of fruit for edible purposes. a cross-section of jackfruit when studied from outside to inside, it consists of an outer peel, rags, latex, bulb, core, seeds and arils. jackfruit bulbs as well as seeds are the edible portions of the fruit (akmeemana et al. 2022), accounting for 25% to 30% of the total weight (baliga et al. 2011). there are two main types of jackfruit: one is firm, less sweet, and crispy, while the other is extremely soft, sweet-tasting, small, and fibrous (swami et al. 2012). jackfruit is a healthier source of calories because it does not contain any cholesterol or saturated fats (goswami and chacrabati 2016). every part of the jackfruit tree can be utilized in a wide range of ways because of that there is no waste generated (ranasinghe et al. 2019). every year, a huge post-harvest loss of jackfruit is reported due to a lack of knowledge on post-harvest technology, improperly organized marketing structures, having a smaller number of processing plants, poor demand in the local market, difficulty in initial processing steps, and rapid quality degradation of edible portion after removing from jackfruit peel (vazhacharickal et al. 2015). global acceptance of convenient instant food is increasing due to changing lifestyles, the need for convenience, an increase in the number of working professionals, and a growing inclination toward western culture. low-caloric and highly nutritious instant food mixes are among the most preferred by consumers globally (ajisha et al. 2018). the present study aimed to investigate the feasibility of using jackfruit bulbs to create an easy-to-cook novel food product, thereby effectively utilizing jackfruit bulbs thus reducing post-harvest losses. furthermore, the objectives of the present study were (i) to examine the impact of a. heterophyllus and trachyspermum ammi dipping treatments on sensory quality (ii) to determine the impact of pre-cook time, rehydration, and freezing on cooking time, and (iii) to analyze the proximate composition and antioxidant activity for the frozen-non-dipped jackfruit bulb treatment, which has the highest level of sensory acceptability. 2 material and methods 2.1 raw materials and chemical ingredients jackfruit bulbs, artocarpus heterophyllus and trachyspermum ammi leaves were harvested from the piliyandala area (6°48′31.8"n 79°57′10.1"e) in sri lanka. chemical reagents chloroform, methanol, concentrated sulfuric acid, boric acid, c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 169 hydrochloric acid, kjeldahl catalyst tablets, kjeldahl indicator, sodium hydroxide, phenol, standard d glucose, phenolphthalein indicator were of the analytical grade. 2.2 collecting of raw materials and initial processing jackfruit a. heterophyllus and t. ammi leaves harvested from the piliyandala area in sri lanka were washed well using clean, running water. jackfruit was cut and jackfruit bulbs were separated. bulbs were stored in an airtight container and taken to the food science and technology laboratory at the university of sri jayewardenepura, sri lanka. jackfruit bulbs were cut into pieces (4.5 cm ± 0.5 cm length, 0.5 cm ± 0.3 cm width in size) and were stored at 4 °c temperature for further use. 2.3 experimental design a two-factor factorial design was used to develop four sample formulations statically (table 1). the variables freezing (-18 °c for 12 h) and dipping (t. ammi hot water extract) were used as factors. table 1. formulation of jackfruit bulbs for the two-factor factorial design. sample no. description 01 non-frozen, non-dipped sample (control sample) 02 frozen, non-dipped sample 03 non-frozen dipped sample 04 frozen, dipped sample 2.4 selection of the best dipping treatment preparation of hot water extracts hot water extracts of t. ammi leaves and a. heterophyllus leaves were prepared according to fernando et al. 1991. in a shaking water bath, 200 g of each leaf material was placed in a beaker with 1000 ml of distilled and shaken for 3 hours at 100 °c until the final volume was reduced to 200 ml. dipping treatment of jackfruit pieces to formulate three solutions with different concentrations, a. heterophyllus hot water extracts were diluted with distilled water by dilution factors of 1/10, 2/10, and 4/10. similar series of solutions were prepared for t. ammi leaves extract using the same dilution factors mentioned above. c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 170 in each of the solutions prepared, 100 g of jackfruit bulb pieces were dipped for about 30 mins and were drained well. then the drained jackfruit bulb pieces were dried in a dehydrator at 60 °c for 12 h. optimization of dipping treatment three sensory evaluation tests with 30 semi-trained panellists were used to select the best dipping treatment and concentration. the first sensory test was performed to determine which of the three concentrations of a. heterophyllus leaves solution-dipped jackfruit pieces provided the best sensory experience. the second sensory test was performed to determine the best sensory treatment among the three concentrations of t. ammi leaves solution-dipped jackfruit pieces. the selected concentration of a. heterophyllus leaves and t. ammi leaves dipped jackfruit bulb samples was used for the third sensory test to finalize the dipping treatment for the subsequent product development process. 2.5 analysis of the effect of pre-cooking time, frozen operation, and rehydration for the cooking time to check whether the freezing, rehydration, and precooking processes affect the cooking time of the dehydrated jackfruit pieces, 15 ± 3 g samples were precooked at one-minute intervals for up to 15 minutes. pre-cooked samples were dried in a dehydrator for 12 h at 60 °c. to demonstrate the effect of the freezing operation on the cooking time of jackfruit bulbs, another 15 samples of 15 g ± 3 g of jackfruit pieces were pre-cooked from 1 min to 15 min separately at the one-minute interval and were frozen at -18 °c for 12 h. frozen samples were thawed for 15 mins and dried in a dehydrator for 12 h at 60 °c. each dehydrated sample was divided into two sets. the first group was boiled in 250 ml of distilled water (100 °c) until the desired softness was obtained. the second group was prepared after soaking the jackfruit pieces in distilled water for 10 min and boiling (100 °c) them in 250ml of distilled water until they reached the desired softness. the samples were tested for softness by pressing them between the fingers and thumb (sethi et al. 2014). 2.6 final sensory evaluation using a five-point hedonic scale rating with 30 semi-trained panelists, the final sensory evaluation test was conducted for four samples created using a two-factor factorial design. before serving the samples for sensory analysis, they were rehydrated for 10 minutes in distilled water and boiled for 2 minutes at 100 °c to achieve the desired cooking quality. data analysis was carried out using spss statistical software version 24.0, and the sample with the best sensory acceptability was identified by using the friedman test. c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 171 2.7 proximate composition analysis the aoac (association of analytical communities) official method 2000 was used to determine the moisture, ash, and crude protein content. the fat content was analyzed according to (sanchez-machado et al. 2004) with slight modifications. initially, 2:1 chloroform: methanol solution was used to extract lipids included in the samples. then, 2 g of the sample (already dried, grounded) was weighed into a centrifuge tube. under the fume hood, 14 ml of the initially prepared solution mixture was added into the centrifuge tube and the centrifuge tube was closed. tubes were vortexed for 2 min in a vortex mixer. then the content of the tube was filtered through whatman no. 41 paper. the residue was re-extracted by using 5 ml of the solvent mixture and vortexed for 30 s. then the extract was filtered through whatman no. 41 paper. the filtrates were allowed to dry under the fume hood. when drying was completed, the final residue was measured and taken as the total lipid content of the sample. 2.8 analysis of total carbohydrate content total carbohydrate content was measured by the dubois method (dubois et al. 1956) with slight modifications. initially, a beaker was placed on an analytical balance and approximately 100 mg of grounded jackfruit bulb sample powder was transferred into the beaker. then, 50 ml of 2 m hcl solution was added to the beaker. the mixture was boiled at 100 °c for about 2 h in a water bath. after that beaker was taken out from the hot water bath and neutralized with 50 ml of 2 m naoh. then the mixture was filtered by using whatman filter paper. a test tube was filled with 0.5 ml of filtrate and 0.5 ml of 20% phenol. afterwards, 2.5 ml of conc. h2so4 acid was added rapidly. then the mixtures of solutions were kept at room temperature for 30 min. the absorbance was measured at 490 nm by using a uv mini spectrophotometer. blank solutions were prepared by using distilled water without jackfruit bulb samples. the amount of sugar in the sample was found with the help of the d-glucose reference curve. the same procedure used to prepare the d-glucose standard curve was used to test d-glucose solutions at concentrations of 1, 0.5, 0.25, 0.125, 0.0625, 0.03125, and 0.015625 mg/ml. triplicate solutions were used to minimize the errors in the results. 2.9 analysis of the mineral content (aas method) sample preparation for metal analysis using the atomic absorption spectrometer was done as per guidelines provided in the aoac official method (aoac 2000). 2.10 determination of titratable acidity for jackfruit bulbs the titratable acidity (ta) was determined according to jagadeesh et al. 2007 with some modifications. accordingly, 5 g of sample was measured using an analytical c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 172 balance and transferred to a vortex tube. then 100 ml of distilled water was added to the tube and the tube was vortexed for 5 min. after that, the mixture was filtered through whatman 41 filter paper and 10 ml of the filtered sample was transferred into a conical flask. as an indicator 4 drops of phenolphthalein were added. the mixture was titrated against 0.1 n naoh until the pink colour endpoint was reached, and ta% was calculated as follows: ta% = normality of naoh ∗ volume of naoh ∗ 192.124 3 ∗ weight of the sample 2.11 determination of antioxidant properties of jackfruit bulbs the dried jackfruit bulb sample was mixed with methanol at 1:40 (g/ml). the sample mixture was allowed to stand at room temperature for 6 h in a shaker and filtered through whatman 41 filter paper. the above mixture was used for the phytochemical composition analysis. dpph radical scavenging activity the free radical activity of the jackfruit bulb sample was measured by dpph following the method of (gunathilake and ranaweera 2016) with some modifications. hence, 6.309 mg dpph was dissolved in 100 ml of methanol solution. next, 2 ml of dpph methanol solution and 2 ml of the diluted solution were transferred into a test tube by using a micro pipet. the solutions were mixed for 1 min by using a vortex mixer. afterwards, the mixture was kept in a dark place for 30 min and the absorbance was measured at 517 nm. absorbance was calculated by using the following equitation. % of inhibition = (a control – a sample) 𝐴 𝑐𝑜𝑛𝑡𝑟𝑜𝑙 ∗ 100 where a control equals to absorbance of the pure dpph methanol solution. total phenolic content total phenolic content was measured according to the folin – ciocalteu reagent assay method described by lim et al. 2002 with some modifications. 2.5 ml of folin – ciocalteu reagent was mixed with 25 ml of distilled water. the final solution was diluted up to 10 folds and allowed to stand at room temperature for 5 min. then l ml of methanol extracted sample and 5 ml of folin – ciocalteu reagent was added to a test tube. the solutions were allowed to stand for 10 min at room temperature. then, 4 ml of 7.5% sodium carbonate solution was added and the mixture was mixed thoroughly. the solutions were kept for 30 min in a dark place at room temperature. absorbance was measured at 765 nm. the standard curve of gallic acid was used to compare the absorbance reading. nine different concentrations (10, 20, 30, 40, 50, 60, 70, 80, and 90 mg/l) of gallic acid solutions were made from 1000 mg/l stock solution. the c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 173 standard curve was constructed from regression analysis using the software minitab (r17). tpc was expressed as gallic acid equivalent in mg/l of jackfruit bulb extract. 2.12 color values for easy-to-cook jackfruit bulb pieces l*, a*, and b* values of dehydrated jackfruit bulb samples were measured by using a chronometer (lovibond® lc100). l* reflects brightness ranging from black to white, a* denotes green to red values, and b* represents blue to yellow values. colour values were taken as replicates (n=3) on different areas of the sample. 2.13 data analysis the collected data were analyzed using minitab 17 statistical software with a 95% confidence interval using one-way anova. the friedman test was used to analyze the sensory evaluation data at a 95% confidence interval using spss software. 3 results and discussion in the present study, a new product formulation was developed using a two-factor factorial design. as factors, freezing and dipping variables were used. hot water extracts of a. heterophyllus and t. ammi leaves were used as dipping variables. a sensory test range was carried out to determine the best dipping treatment and concentration. hot water extract from the leaves of a. heterophyllus was occupied because of its antioxidant properties, (loizzo et al. 2010) whereas hot water extract from the leaves of t. ammi was used because it acts as a digestive stimulant by promoting the enzyme reactions that are responsible for bile secretion, digestion, or both. (prakash et al. 2009) the jackfruit bulb pieces were precooked before dehydrating because it takes considerably more time for the developed product to achieve the desired cooking quality without precooking. the cooking time evaluation test was conducted to determine the pre-cooked time. the biochemical analysis was performed on the sample that was perceived as the most acceptable sensory sample. 3.1 sensory evaluation for optimization of dipping treatment of jackfruit bulb figure 1 illustrates the radar charts used to get an insight into the best sensory quality sample, from the two basic dipping treatments using a. heterophyllus and t. ammi leaves extracts applied to jackfruit bulbs. according to figure 1, the sample code-100, i.e., jackfruit bulb pieces dipped in 1/10 dilution factor of a. heterophyllus leaves hot water extract, and the sample code-500, i.e., jackfruit bulb pieces dipped in 2/10 dilution factor of t. ammi leaves hot water extract was selected from the dilution factor c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 174 series based on the results of sensory evaluation tests. the jackfruit bulb pieces dipped in a 2/10 dilution factor of t. ammi leaves hot water extract was the best sensory quality sample in terms of appearance, texture, odour, taste, and overall acceptability, according to the results of the sensory evaluation tests conducted for the above-selected samples (code 100 and code 500). fig 1. optimization of the sensory treatments in jackfruit bulbs. (note: the codes 100, 200, and 300 represent 1/10, 2/10 and 4/10 dilution factors of hot water extract of a. heterophyllus leaves, respectively. similarly, 400, 500, and 600 represent 1/10, 2/10 and 4/10 dilution factors of hot water extract of t. ammi leaves, respectively.) 3.2 evaluation of cooking time fig 2. the effect of cooking time on rehydration, freezing, and pre-cooked time c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 175 according to figure 2, with the optimization of the pre-cooking time of jackfruit bulb pieces, the frozen treatment has reduced the cooking time of the processed jackfruit bulbs. according to the results, 6 min pre-cooking, and 10 min rehydration was selected to reduce the final cooking time of the developed sample. according to kuna a et al. (2019), the frozen operation reduced the cooking time significantly in the red gram dhal. when the pre-cooked time increased, the cooking time was reduced accordingly. according to ghadge et al. (2008), the pre-cooking and freezing operation reduced the cooking time of the instant whole legume. 3.3 final sensory evaluation fig 3. radar chart for the sensory evaluation of the final processed jackfruit bulb samples. (the codes 103, 203, 303 and 403 represent non-frozen and non-dipped samples, frozen and non-dipped samples, non-frozen and dipped samples, and frozen and dipped samples respectively). according to the radar chart (figure 3), the sample containing frozen jackfruit bulbs, without t. ammi leaves hot water extract was the best sensory quality sample concerning the appearance, texture, odour, taste and overall acceptability. 3.4 proximate composition the proximate composition of the best sensory-acceptable sample (the frozen jackfruit sample without t. ammi leaves hot water extract) is given in table 2. moisture content opkala (2010) reported that 5 mins steamed blanched, 0.1% sodium sulfate treated, 55 °c dried jackfruit bulb flour contains 15.19 ± 0.01 % moisture content. similarly, yi et al. (2016) reported the influence of pre-drying treatments on the physicochemical c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 176 and organoleptic properties of explosion puff dried jackfruit chips. according to a study at 65 °c hot air-dried jackfruit bulb pieces contained 6.13 ± 0.12 % moisture content. the results of the current study showed 8.08 ± 0.28 % moisture content at 60 °c and which is less than the moisture content of the dried jackfruit bulbs at 55 °c and greater than the dried jackfruit bulbs at 65 °c. table 2: proximate composition of the jackfruit sample with the best sensory quality. variable value moisture 8.08 ± 0.28 % ash 0.88 ± 0.01% total fat 0.31 ± 0.03% crude protein 3.93 ± 0.06% total carbohydrate 81.09 ± 0.54% [data presented as mean ± s.d. (n = 3) ash content goswami et al. (2011) stated that the biochemical parameters of fresh jackfruit bulbs changed according to the growing area and the ash content of fresh jackfruit bulbs can be ranged from 1.11 to 0.70%. the current study demonstrated 0.88 ± 0.01% ash content which is comparable to goswami et al. (2011). fat content opkala (2010) reported that 5 mins of steamed blanched, 0.1 % sodium sulfate treated, 55 °c dried jackfruit bulb flour contains 0.20 ± 0.04 % fat content. however, according to the variety, growing region (goswami et al. 2011), and maturity stage (maheswari and valsan 2020), the biochemical parameters of jackfruit can vary. ranasinghe et al. (2019) reported that the fat content of young jackfruit can vary between 0.1 and 0.6 % while the fat content of ripened jackfruit can vary between 0.1 and 0.4 %. the present study reported 0.31 ± 0.03 % fat content and it was in the range of the above studies. crude protein content maheswari and valsan (2020) stated that the crude protein content of jackfruit can vary from 1.61 ± 0.02 % to 4.36 ± 0.06 % depending on the maturity stage of the jackfruit bulbs. the current study reported 3.93 ± 0.06 % of crude protein content and it complies with the above range reported in the previous study. 3.5 mineral content analysis maheswari and valsan (2020) further reported 44.67 ± 0.50 % of calcium to be present in raw mature jackfruit bulbs while ranasinghe et al. (2019) reported 2.0 41.0 % of sodium content to be available in raw jackfruit bulbs. the mineral composition of the c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 177 best sensory-acceptable sample with respective to na, ca, mg, k, and p is given in table 3. table 3: mineral content of the jackfruit sample with the best sensory quality. mineral type na ca mg k p concentration (mg/100g) 8.47 ± 0.04 43.88 ± 0.25 61.49 ± 0.47 709.65 ± 0.46 48.76 ± 0.21 [ data presented as mean ± s.d. (n = 3) ] 3.6 titratable acidity titratable acidity of jackfruit sample with the best sensory acceptability, i.e. frozen without dipping in t. ammi leaves extract was 0.29 ± 0.01 % (mean ± sd; n = 3). yi et al. (2016) reported 0.30 ± 0.02 % titratable acidity in dehydrated jackfruit bulbs at 65 °c. yi, wang et al. (2016) reported 0.38 ± 0.04 % titratable acidity for freeze-dried jackfruit bulb pieces. results of the current study were much lower compared to those. according to the (opkala 2010) study, 0.81 ± 0.28 % titratable acidity was reported in jackfruit bulb flour made out of 5 min steamed blanched, 0.1 % sodium sulfate treated and 55 °c dried in a conventional hot-air drier. the variations may be due to the variety, growing condition and maturity stage of the jackfruit. 3.7 antioxidant properties of easy-to-cook jackfruit bulb dpph (1-1-diphenyl 1-2-picrylhydrazyl) radical scavenging activity assay jagtap et al. (2010) reported that radical scavenging activity for the ethanolic extract of raw jackfruit bulb concentration ranged from 1, 2, 3, 4, and 5 mg/ml was in between 55% and 65%. the obtained value in this study (38.43 ± 0.12%) was less than the previously reported values, and it may be due to heating of the bulbs at 100 °c for 6 mins and dehydration at 60 °c. total phenolic content yi, zhou et al. (2016) stated that the total phenolic content of hot air-dried jackfruit bulb was 0.84 ± 0.06 (mg gae / g). according to (yi, zhou et al. 2016) the total phenolic content of the fresh jackfruit bulb was 1.8 ± 0.1 (mg gae / g) and after freezedrying, it was decreased up to 1.6 ± 0.1 mg gae / g. according to the (jagtap et al. 2010) study, the total phenolic content of the fresh jackfruit bulbs was 0.46 ± 0.014 mg gae/ g. the value obtained for phenolic content in the present study (0.38 ± 0.04%) is lower than the above-reported values. this may be due to the pre-cooking of the bulbs for 6 mins at 100 °c, frozen treatment, thawing for 15 mins and drying treatment carried out at 60 °c. c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 178 3.8 colour values of easy-to-cook jackfruit bulb table 4 indicates the colourimetric values of the best sensory-acceptable jackfruit bulb sample. table 4. colour values of best sensory quality sample. l* a* b* c* 79.40 ± 0.61 1.87 ± 0.35 25.43 ± 1.33 25.50 ± 1.35 * data presented as mean ± s.d. (n = 3) the l* value is a measure of the lightness of the product colour. it ranges from 100 to 0. a* and b* values indicate the redness/greenness and yellowness/blueness respectively (swami et al. 2016). 4 conclusions four different formulations were developed by jackfruit bulbs following a two-factor factorial design, dipping or non-dipping and frozen or non-frozen as the variables. among the three sensory tests conducted, the best sensory quality sample concerning appearance, texture, odour, taste and overall acceptability was the 2/10 dilution factor of hot water extract of t. ammi leaves dipped jackfruit bulb pieces sample. the results revealed that the frozen operation, pre-cooked time and rehydration significantly reduce the cooking time of jackfruit bulbs. to reduce the final cooking time of the developed product, 6 minutes of pre-cooking and 10 minutes of rehydration were selected based on the results of the cooking time evaluation. final sensory evaluation indicates frozen non-dipped sample as the most sensory-acceptable sample. furthermore, the final processed product possesses a good nutritional profile concerning the ash, fat, crude protein, carbohydrate, mineral content, dpph and total phenolic acid content. acknowledgements the authors thankfully acknowledge the university of sri jayewardenepura, nugegoda, sri lanka, for providing research facilities to conduct the study. anonymous reviewers are thanked for their comments on the initial manuscript. references ajisha kh, sharon l, aneena r, panjikkaran t. 2018. development and evaluation of raw jackfruit and jackfruit seed based instant payasam mix. indian journal of scientific research 19(1): 10-15. akmeemana c, wickramasinghe i, wanniarachchi pc, vithanage t. 2022. effect of drying and frying pre-treatments on nutrient profile, antioxidant capacity, cooking time, and c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 179 sensory acceptability of easy to cook jackfruit seeds. applied food research 100234 p. doi: 10.1016/j.afres.2022.100234 baliga ms, shivashankara ar, haniadka r, dsouza j, bhat hp. 2011. phytochemistry, nutritional and pharmacological properties of artocarpus heterophyllus lam (jackfruit): a review. food research international 44(7): 1800-1811. doi: 10.1016/j.foodres.2011.02.035. fernando mr, wickramasinghe sn, thabrew mi, ariyananda pl, karunanayake eh. 1991. effect of artocarpus heterophyllus and asteracanthus longifolia on glucose tolerance in normal human subjects and in maturity-onset diabetic patients. journal of ethnopharmacology 31(3): 277-282. doi: 10.1016/0378-8741(91)90012-3 ghadge pn, shewalkar sv, wankhede db. 2008. effect of processing methods on qualities of instant whole legume: pigeon pea (cajanus cajan l.). international commission of agricultural and biosystems engineering (cigr). goswami c, chacrabati r. 2016. jackfruit (artocarpus heterophyllus). in: nutritional composition of fruit cultivars. 317-335. doi:10.1016/b978-0-12-408117-8.00014-3 gunathilake kpp, ranaweera kkds. 2016. antioxidative properties of 34 green leafy vegetables. journal of functional foods 26: 176-186. doi: 10.1016/j.jff.2016.07.015 jagadeesh sl, reddy bs, swamy gsk, gorbal k, hegde l, raghavan gsv. 2007. chemical composition of jackfruit (artocarpus heterophyllus lam.) selections of western ghats of india. food chemistry 102(1): 361-365. doi: 10.1016/j.foodchem.2006.05.027 jagtap ub, panaskar sn, bapat va. 2010. evaluation of antioxidant capacity and phenol content in jackfruit (artocarpus heterophyllus lam.) fruit pulp. plant foods for human nutrition 65(2): 99-104. doi: 10.1007/s11130-010-0155-7 kuna a, sreedhar m, jagan c, rani, d.s bhagyamma m, rani vs. 2019. development and evaluation of quick cooking redgram dhal (cajanus cajan l.). the journal of research pjtsau 3: 33 lim sn, cheung pck, ooi vec, ang po. 2002. evaluation of antioxidative activity of extracts from a brown seaweed, sargassum siliquastrum. journal of agricultural and food chemistry 50(13): 3862-3866. doi: 10.1021/jf020096b loizzo mr, tundis r, chandrika ug, abeysekera am, menichini f and frega ng. 2010. antioxidant and antibacterial activities on foodborne pathogens of artocarpus heterophyllus lam.(moraceae) leaves extracts. journal of food science 75(5): m291-m295 doi: 10.1111/j.1750-3841.2010.01614.xm295. maheswari tu, valsan v. 2020. value addition of jackfruit through production of chips. doi: 10.47587/sa.2020.1201 opkala mo. 2010. development and evaluation of baked products from jackfruit (artocarpus heterophyllus lam) seed kernel and pulp flour. m.sc. dissertation, university of nigeria. prakash o, kumar r, mishra a, gupta r. 2009. artocarpus heterophyllus (jackfruit): an overview. pharmacognosy reviews 3(6): 353-358. ranasinghe rasn, maduwanthi sdt, marapana rauj. (2019). nutritional and health benefits of jackfruit (artocarpus heterophyllus lam.): a review. international journal of food science. doi: 10.1155/2019/4327183. sánchez-machado di, lópez-cervantes j, lopez-hernandez j, paseiro-losada p. 2004. fatty acids, total lipid, protein and ash contents of processed edible seaweeds. food chemistry 85(3): 439-444. doi: 10.1016/j.foodchem.2003.08.001 c. akmeemana et al. process optimization for easy-to-cook jackfruit bulbs ruhuna journal of science vol 13 (2): 167-180, december 2022 180 sethi s, samuel dvk, khan i. 2014. development and quality evaluation of quick cooking dhal—a convenience product. journal of food science and technology 51(3): 595-600 doi.: 10.1007/s13197-011-0534-6 swami sb, thakor nj, haldankar pm, kalse sb. 2012. jackfruit and its many functional components as related to human health a review. comprehensive reviews in food science and food safety 11(6): 565-576. doi. : 10.1111/j.1541-4337.2012.00210.x swami sb, thakor nj, orpe s, kalse sb. 2016. development of ripe jackfruit bulb powder and its quality evaluation. journal of food research and technology 4(1): 22-29 vazhacharickal pj, sajeshkumar nk, mathew jj, kuriakose ac, abraham b, mathew rj, albin an, thomson d, thomas rs, varghese n, jose s. 2015. chemistry and medicinal properties of jackfruit (artocarpus heterophyllus) a review on current status of knowledge. international journal of innovative research and review 3(2): 83-95 yi j, wang p, bi j, liu x, wu x, zhon y. 2016. developing novel combination drying method for jackfruit bulb chips instant controlled pressure drop (dic)-assisted freeze drying. food and bioprocess technology 9(3): 452-462. doi: 10.1007/s11947-015-1643-4 yi j, zhou l, bi j, chen q, liu x, wu x. 2016. influence of pre-drying treatments on physicochemical and organoleptic properties of explosion puff dried jackfruit chips. journal of food science and technology 53(2): 1120-1129. doi: 10.1007/s13197-0152127-2. sv-lncs ruhuna journal of science vol. 5: 16-30, 2014 http://rjs.ruh.ac.lk/ issn: 1800-279x 16  faculty of science, university of ruhuna in vitro regeneration of hypochaeris radicata l. from sodium alginate-encapsulated synthetic seeds jamuna senguttuvan and paulsamy subramaniam* department of botany, kongunadu arts and science college, coimbatore-641 029, tamil nadu, india *correspondence: paulsami@yahoo.com received: 01 may 2014, revised version accepted: 23 october 2014 abstract. synthetic seeds were produced from in vitro derived leaf, root and callus explants of hypochaeris radicata (hairy cat’s-ear) by encapsulating different concentrations of sodium alginate hydrogel (16%) containing ms medium. the texture, conversion frequency and the effect of temperature on shoot emergence were evaluated. among three explants attempted, in vitro derived leaf segments encapsulated beads stored at 25°c observed successful shoot regeneration above 80% in the standardized ms medium supplemented with 2mg/l bap. the regenerated shoots were rooted well (73.54%) on ms medium with 1mg/l naa. the high frequency of plant re-growth was found in four month old leaf segment encapsulated beads. generally, four months stored beads of all explants at 25°c found to have higher regeneration rate in comparison with 2 and 6 months old encapsulated beads. keywords. asteraceae, hypochaeris radicata, in vitro regeneration, synthetic seeds 1 introduction synthetic seeds are artificially encapsulated plant propagules (somatic embryos, shoot buds, cell aggregates or any other tissues) that can be used for sowing as a seed and possess the ability to convert into a plant under in vitro or ex vitro conditions, and also retain this potential after storage (krishna kumar and dennis thomas 2012). plant propagules are encased in protective coating of gelling agents such as alginate, agar, carrageenan, gellan gum (gerlite), sodium pectate, ethylene glycol, dimethyl sulfoxide and carboxyl jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 17 methyl cellulose (harikrishna and ong 2002). the coating protects the explants from mechanical damage during handling and allows germination and conversion to occur without inducing undesirable variations. it’s potential advantages include stability during handling, potential for long term storage without losing viability, ability to transport and plant directly from in vitro to field conditions and low cost production at higher scale (ghosh and sen 1994). synthetic seeds have diverse applications such as: multiplication of non-seed producing plants, ornamental hybrids or polypoid plants, propagation of male or female sterile plants for hybrid seed production, germplasm conservation of recalcitrant species and multiplication of transgenic plants. the concept of artificial seed technology has been applied successfully in commercial settings in cardamom (ganapathy et al. 1994), sugar beet (dennis et al. 1991), sandal wood (bapat and rao 1998), banana (ganapathi et al. 2002), garlic (bekheet 2006), rice (roy and mandal 2008), neem (mithilesh and rakhi 2013), sweet neem (aman and shruti 2013) and also in some orchidaceae members (sharma et al., 1992; anju et al. 2012). hypochaeris radicata commonly called as hairy cat’s-ear, is an edible, perennial herb belongs to the family asteraceae. it is native to south africa and distributed in forest margins of nilgirs, the western ghats, tamil nadu at 2000m above msl. it possesses several medicinal properties like antiinflammatory, anticancer, antioxidant (jamuna et al. 2012, 2014) and antimicrobial (jamuna et al. 2013a; jamuna et al. 2013b). it is being prescribed for the treatment of jaundice, rheumatism, dyspepsia, constipation, hypoglycemia and kidney related problems in traditional medicinal practices of tamil nadu, india (pullaiah 2006). it is used for some medicinal purposes like controlling wound infections in meghalaya (tynsong et al. 2006) and also used as food for ruminants in british columbia (lans et al. 2007). further, h. radicata is reported to have many bioactive compounds of medicinal importance like phytol, acetate, hexadecanoic acid, methyl ester, 9,12,15-octadecatrienoic acid, methyl ester, urs-12-en-3-ol, acetate, (3 beta, 1benzazirene-1-carboxylic acid, 2,2,5a-trimethyl-1a-(3-oxo-1-butenyl) perhydro-methyl ester (jamuna and paulsamy 2013). the leaves are usually blanched, steamed and cooked or used as spices in a range of western and eastern culinary preparations and bevarages (www.ifood.tv/network/catsear). as the seed longevity is poor under natural conditions, the germination and hence the population sizes are affected drastically in the upper reaches of nilgiris (paulsamy et al. 2008). to overcome this problem, the production of synthetic seeds and sowing them during appropriate period is essential. in http://www.ifood.tv/network/catsear jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 18 light of this fact, the present study was aimed at determining the optimum concentration of encapsulation matrix (sodium alginate solution) to optimize the texture, storage period and temperature of the alginate beads for effective bud-sprouting and also to assess the germination efficiency of the seeds in terms of multiple shoot formation, rooting and hardening attributes to ascertain the feasibility of their use as an alternative to true seeds by standardizing the murashige and skoog (ms) medium. 2 materials and methods 2.1 plant materials leaf, root and callus (leaf-derived) explants from in vitro derived plants of h. radicata were used. all these explants were cut into 2 to 3 mm size for further use. 2.2 preparation of encapsulation matrix for encapsulation process, various concentrations of sodium alginate at 1, 2, 3, 4, 5 and 6% (w/v) were prepared in 100ml of ms liquid medium without agar. for complexation (an ion exchange reaction between na + and ca 2+ resulting in the formation of insoluble calcium alginate), 1.016g of calcium chloride (cacl2.2h2o) in 150ml in distilled water was prepared. both solutions were sterilized at 120°c for 15 min at 1.12kg/cm 2 pressure and allowed to cool. 2.3 formation of beads for the formation of beads, the in vitro derived organs viz., leaf, root and leafderived callus explants, each numbering 50 were aseptically transferred into each concentration of sodium alginate solution and incubated at room temperature for 15-20min. later, the micropropagules in alginate solution were picked up by pasteur pipette and dropped into a sterile solution of cacl2.2h2o. the drops, each containing a single micropropagule when left in cacl2.2h2o for 30 min was placed on shaker to form round-shaped firm beads as a result of the ion exchange reaction between on na + and ca 2+ ions jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 19 (louis et al. 1998). the beads were recovered by recycling and decanting cacl2.2h2o solution and washed thrice with autoclaved ms liquid medium. the beads were then transferred to sterile filter paper in petridishes. blot dried beads were stored in culture bottles sealed with parafilm for 2, 4 and 6 months in two different temperature conditions viz., at 4ºc and 25°c separately. 2.4 culture medium and condition the encapsulated explants were cultured on the ms medium (murashige and skoog 1962) consists of ms mineral salts and vitamins supplemented with 3% sucrose and 0.8% agar. the ph of the medium was adjusted to 5.6 to 5.8 before sterilization by autoclaving at 121°c with 1.12kg/cm 2 pressure for 15 min. all cultures were maintained under white fluorescent light having 2000 lux light intensity. the incubation temperature was 25±2°c with 16 hours light and 8 hours dark period in every 24 hours cycle. 2.5 induction of organogenesis the encapsulated beads were implanted on ms medium containing different growth regulators viz., bap (0.5-3.0mg/l) alone and it in individual combination with ga3 (0.5mg/l), kn (0.5mg/l) and iaa (0.5mg/l) for shoot formation. for root induction, the micro-shoots (2-3cm) were excised and sub-cultured onto the ms medium supplemented with various concentrations of growth regulators viz., iaa, iba and naa (0.3, 0.5 and 1.0mg/l respectively). 2.6 hardening and acclimatization well-developed healthy plantlets were removed from the culture flasks and were thoroughly washed in running tap water to remove remnant nutrient medium completely without causing any damage to roots. then the root portion was soaked in 1% (w/v) fungicide, methyl-2 benzimidizole carbamate (bavistin) solution for 10min and transferred to small earthen pots of 15×15cm (h×w) size each filled with various types of sterilized potting mixtures garden soil: sand: vermicompost (1:1:1 by volume), red soil: sand: vermicompost (1:1:1 by volume), vermicompost: soil (1:1 by volume), red soil: sand (1:1 by volume) and decomposed coir waste: perlite: vermicompost (1:1:1 by volume). jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 20 2.7 statistical analysis for encapsulation of synthetic seeds and in vitro regeneration, randomly selected 30 individuals of each explant and for this purpose in triplicate were maintained. the data on conversion frequency (after 15 days of incubation of encapsulated beads), shooting frequency (after 25 days of incubation of encapsulated beads), multiple shoot induction (after 25 days of incubation of encapsulated beads), root induction (after 30 days of incubation of in vitro derived shoots) and number of plantlets survived (after 60 days of hardening) were statistically analyzed using anova, and means were compared by using duncan’s multiple range test (p<0.05) (duncan 1955). 3 results table 1. effect of sodium alginate concentration on texture and conversion frequency of leaf, root and callus explants encapsulated beads of hypochaeris radicata in the standardized ms medium supplemented with 2mg/l bap. explants concentration of sodium alginate (%) texture of beads conversion frequency (%)* in vitro leaf 1 too soft 28.49±0.31 ab 2 soft 41.49±0.31 c 3 soft 86.64±1.04 g 4 firm 71.12±0.85 f 5 hard 47.11±0.02 c 6 very hard 38.71±0.65 bc in vitro root 1 too soft 21.94±0.80 a 2 soft 38.94±0.80 b 3 soft 42.37±0.40 c 4 firm 51.33±0.89 d 5 hard 46.43±0.85 c 6 very hard 31.00±0.29 b in vitro callus 1 too soft 43.24±0.28 c 2 too soft 43.24±0.28 c 3 soft 63.12±0.79 e 4 firm 55.10±0.42 d 5 hard 48.37±0.39 cd 6 very hard 30.30±0.02 b *mean of three replicates of 30 explants ± standard deviation. values within the column followed by same superscript are not significantly different at p<0.05. jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 21 3.1 effect of different concentrations of sodium alginate on conversion frequency the encapsulated beads differed morphologically with respect to texture with different concentrations of sodium alginate (1-6%) (table 1). an encapsulation matrix of 3% sodium alginate with 100mm of cacl2.2h2o was found to be the most suitable for the formation of ideal beads (figs. 1a and 2a). among the three explants attempted, the highest conversion frequency (86.64%) was observed in leaf explants encapsulated beads followed by the callus encapsulated beads with 63.12% conversion frequency. based on these findings, leaf and callus explant encapsulated beads were taken for further sub-culturing experiments. table 2. effect of storage period and temperature on shoot emergence from different explant encapsulated beads of hypochaeris radicata in the standardized ms medium supplemented with 2mg/l bap. beads with explants storage period (months) temperature (°c) shooting frequency (%)* encapsulated leaf 2 4 25 32.23±0.21 c 59.71±0.04 e 4 4 25 61.54±0.33 f 82.52±1.06 g 6 4 25 43.21±0.26 d 67.28±0.71 f encapsulated root 2 4 25 21.34±0.05 b 47.28±0.23 e 4 4 25 10.33±0.52 a 32.94±0.43 c 6 4 25 11.33±0.29 a 29.00±0.05 b encapsulated callus 2 4 25 25.23±0.18 b 39.46±0.48 c 4 4 25 32.66±0.29 c 51.10±0.31 e 6 4 25 18.20±0.22 a 29.31±0.56 b *mean of three replicates of 30 explants ± standard deviation. values within the column followed by same superscript are not significantly different at p<0.05. jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 22 fig. 1. stages of regeneration in encapsulated in vitro leaf segments of hypochaeris radicata. (a) beads encapsulated with leaf segments in 3% sodium alginate, (b) induction of synthetic seeds on ms medium with bap at 2mg/l, (c) ruptured beads showing sprouting of shoots on ms medium containing bap at 2mg/l (d) after 2 weeks of culture (e) well defined multiple shoot emergence on ms medium with bap at 2mg/l (f) root induction on ms medium fortified with naa at 1mg/l (g) after 2 weeks of culture (h) regenerated plantlet of h. radicata after successful acclimatization (i) storage of synthetic seeds in a culture bottle. jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 23 fig. 2. stages of regeneration in in vitro callus segments of h. radicata. (a) synthetic seeds produced by 3% sodium alginate, (b) induction of synthetic seeds on ms medium with bap at 2mg/l, (c) ruptured beads showing sprouting of shoots on ms medium containing bap at 2mg/l, (d) after 2 weeks of culture (e), multiple shoots regenerated from an encapsulated callus beads on ms medium containing bap at 2mg/l, (f) root induction on ms medium fortified with naa at 1mg/l, (g) after 2 weeks of culture, (h) acclimatized transformed plant, 8 weeks after transferred to soil, (i) storage of synthetic seeds in a culture bottle. jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 24 3.2 effect of storage period and temperature on shoot emergence from encapsulated beads table 2 shows the shooting frequency of encapsulated in vitro derived leaf, root and callus segments, after 2, 4 and 6 months of storage period at 4 and 25°c. four months old stored beads at 25°c resulted in high rate of shoot proliferation in all explant segments. longer storage of 6 months period has significantly decreased the regeneration ability of explants. higher shooting frequency was noticed in four months stored leaf explant encapsulated beads (82.52%) followed by callus encapsulated beads (51.10%). the root explants encapsulated beads when stored for two months at 25°c produced shoots with lower conversion frequency (47.28%) in comparison to leaf and callus encapsulated beads. 3.3 effect of various growth regulators on bud sprouting and regeneration of encapsulated beads data on certain shooting attributes of 3% sodium alginate encapsulated beads cultured on ms medium with various combinations and concentrations of some plant growth regulators are given in table 3. the shooting response varied according to growth regulators used in the medium. ms medium supplemented with 2mg/l bap gave higher frequency (88.55%) (fig. 1c) for the conversion of leaf encapsulated beads into multiple shoots with a higher number of shoots per explant (15.59shoots/encapsulated bead). when the medium was supplemented by bap (2.0mg/l) in combination with ga3 (0.5mg/l) the shoot emergence frequency was 78.54% with shoot length of 7.44cm. however, the shoot proliferation frequency from callus encapsulated beads was noted to be lower (63.87%) than leaf encapsulated beads (fig. 2c). 3.4 rooting and acclimatization of plantlets regenerated from encapsulated beads the rooting characters of in vitro derived leaf and callus explants encapsulated synthetic seeds are presented in table 4. the rooting frequency was significantly higher (73.54%) for leaf encapsulated beads in the ms medium fortified with 1mg/l naa. at the same time, the number of roots produced (5.30 roos/shoot) and the average root length (6.37cm) were higher in this medium (fig. 1f). the rooting attributes of in vitro derived callus explants ecncapsulated synthetic seeds were lower (59.40%) (fig. 2f) than jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 25 that of leaf encapsulated beads. the in vitro regenerated plantlets were acclimatized successfully using garden soil, sand and vermicompost (1:1:1 by volume). the higher survivability rates were 70% for leaf encapsulated beads (fig. 1h) and 60% for callus encapsulated beads (fig. 1h) (table 5). table 3. effect of plant growth regulators on shoot multiplication, shoot number and shoot length of in vitro derived leaf, root and callus encapsulated beads of hypochaeris radicata in 3% sodium alginate. growth regulators multiple shoot induction (%)* no. of shoots/explant* shoot length (cm)* bap ga3 kn iaa leaf callus leaf callus leaf callus 0.5 0.0 0.0 0.0 1.0 0.0 0.0 0.0 31.15±0.54 c 24.43±0.22 b 1.23±0.09 b 1.03±0.02 b 2.54±0.15 b 2.05±0.09 b 1.5 0.0 0.0 0.0 64.65±0.83 f 54.53±0.11 f 1.35±0.16 b 1.15±0.11 b 5.33±0.47 de 3.12±0.71 c 2.0 0.0 0.0 0.0 88.55±2.82 h 63.87±1.86 g 15.59±0.51 g 10.35±0.41 e 6.75±0.23 e 4.77±0.38 d 2.5 0.0 0.0 0.0 60.23±0.93 f 47.85±0.46 ef 3.52±0.19 c 2.85±0.34 b 2.70±0.11 b 1.96±0.09 a 3.0 0.0 0.0 0.0 32.00±0.51 c 33.96±0.46 c 1.32±0.55 b 1.02±0.09 b 1.60±0.06 a 1.26±0.76 a 0.5 0.5 0.0 0.0 1.0 0.5 0.0 0.0 1.5 0.5 0.0 0.0 34.54±0.62 c 31.12±0.54c 1.56±0.91 b 1.05±0.34 b 2.76±0.09 b 2.17±0.76 b 2.0 0.5 0.0 0.0 78.54±0.65 g 63.87±0.73g 7.11±0.51 f 5.10±0.75 d 7.44±0.19 c 6.44±0.21 f 2.5 0.5 0.0 0.0 51.10±0.16 e 42.65±0.32e 3.55±0.03 b 0.98±0.19 a 3.25±0.32 e 3.53±0.11 c 3.0 0.5 0.0 0.0 30.43±0.22 bc 20.53±0.09b 1.41±0.64 b 0.75±0.43 a 3.10±0.31 c 1.36±0.24 a 0.5 0.0 0.5 0.0 1.0 0.0 0.5 0.0 1.5 0.0 0.5 0.0 35.22±0.42 c 38.32±0.03 d 1.22±0.14 b 0.96±0.11 a 2.13±0.51 b 1.32±0.32 a 2.0 0.0 0.5 0.0 69.24±0.25 f 58.21±0.75 f 3.89±0.57 e 1.72±0.49 b 2.43±0.32 b 2.64±0.45 b 2.5 0.0 0.5 0.0 47.44±0.27 d 17.00±0.51 b 1.67±0.83 b 1.16±0.27 b 4.07±0.33 d 3.32±0.12 c 3.0 0.0 0.5 0.0 29.31±0.02 b 09.53±0.05 a 0.78±0.24 a 1.01±0.17 b 2.30±0.64 b 2.11±0.09 b 0.5 0.0 0.0 0.5 1.0 0.0 0.0 0.5 1.5 0.0 0.0 0.5 2.0 0.0 0.0 0.5 64.32±0.52 f 56.32±0.16 f 3.20±0.21 d 2.23±0.65 c 2.13±0.17 b 1.54±0.49 a 2.5 0.0 0.0 0.5 38.13±0.63 c 35.05±0.02 cd 2.74±0.53 c 1.52±0.38 b 3.98±0.52 c 3.12±0.32 c 3.0 0.0 0.0 0.5 15.54±0.33 a 10.01±0.02 a 1.02±0.42 b 1.65±0.47 b 2.42±0.26 b 2.00±0.19 b *mean of three replicates of 30 explants ± standard deviation. values within the column followed by same superscript are not significantly different at p<0.05. 4 discussion the encapsulation technique for producing synthetic seeds or artificial seeds has become an important asset in micropropagation. encapsulation of in vitro derived vegetative propagules to develop synthetic seeds has been employed in recent years as a suitable alternative to the use of somatic embryos (nor asmah et al. 2012). jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 26 table 4. effect of different concentrations of growth regulators on root initiation, root number and root length from in vitro derived shoots of hypochaeris radicata. growth regulators (mg/l) shoots rooted (%)* no. of roots/shoot* root length (cm)* iaa iba naa leaf callus leaf callus leaf callus 0.0 0.3 0.0 12.22±0.56 a 10.09±0.18 a 1.12±0.37 a 1.03±0.49 a 0.32±0.02 a 0.44±0.10 a 0.0 0.5 0.0 24.31±0.71 b 18.09±0.07 ab 1.55±0.13 a 1.32±0.17 a 1.64±0.33 b 2.00±0.09 c 0.0 1.0 0.0 35.44±0.29 c 30.32±0.54 c 2.63±0.11 b 1.93±0.29 ab 4.55±0.29 d 2.98±0.29 c 0.0 0.0 0.3 21.10±0.30 b 16.04±0.26 a 1.00±0.12 a 1.22±0.48 a 1.09±0.25 b 1.22±0.55 b 0.0 0.0 0.5 48.32±0.37 d 24.11±0.43 b 1.23±0.43 a 1.74±0.22 ab 2.22±0.75 c 2.76±0.06 c 0.0 0.0 1.0 73.54±0.19 e 59.40±0.33 d 5.30±0.64 c 5.43±0.48 c 6.37±0.33 e 6.09±0.43 d 0.3 0.0 0.0 0.5 0.0 0.0 10.10±0.03 a 25.52±0.04 b 1.10±0.22 a 1.33±0.29 a 1.02±0.32 b 1.43±0.05 b 1.0 0.0 0.0 39.54±0.09 c 31.54±0.32 c 2.66±0.26 b 2.32±0.40 b 2.33±0.11 c 2.15±0.45 c *mean of three replicates of 30 explants ± standard deviation. values within the column followed by same superscript are not significantly different at p<0.05. table 5. effect of various composition of hardening medium on survivability of hypochaeris radicata plantlets. hardening medium composition (v/v) no. of plantlets under hardening no. of plantlets survived* survivability (%)* leaf callus leaf callus garden soil: sand: vermicompost (1:1:1) 30 21±0.12 d 18±0.03 d 70.00±0.34 e 60.00±0.56 d red soil: sand: vermicompost (1:1:1) 30 16±0.04 c 15±0.22 c 53.33±0.18 d 50.00±0.04 c vermicompost: soil (1:1) 30 13±0.23 b 14±0.13 b 43.33±0.61 b 46.66±0.19 b red soil: sand (1:1) 30 09±0.11 a 09±0.43 a 30.00±0.21 a 30.00±0.52 a decomposed coir waste: perlite: vermicompost (1:1:1) 30 08±0.09 c 06±0.57 c 26.66±0.21 c 20.00±0.49 c *mean of three replicates of 30 explants ± standard deviation. values within the column followed by same superscript are not significantly different at p<0.05. sodium alginate (c6h7o6na) is a colorless or light yellow filaments, granules, or powder which forms a viscous colloid in water and used in food thickeners and stabilizers, in medicine, textile printing, paper coating and water-base paint. also it is known as algin; alginic acid sodium salt and sodium polymannuronate. it is a copolymer composed of d-mannuronic acid and lglucuronic acid units and has been extensively studied because of its biocompatibility, biodegradability and its capacity to form hydrogels in the jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 27 presence of divalent cations. the ridge structure and large pore size of these gels which are insoluble in water make them useful for the encapsulation of live cells of plants. polymer concentration, degree of viscosity of the alginate used, cacl2 concentration and curing time are important parameters determining the permeability, resistance and hardness of the resulting beads and the subsequent success of the encapsulation method (block 2003). in the present study, the polymerizing ability of 3% sodium alginate showed to be the best for encapsulating explants. however, sodium alginate concentration below 3% were not suitable for encapsulation because the resulting beads were without a definite shape and were too soft to handle, whilst at higher concentration the beads became hard causing a considerable delay in regeneration. sodium alginate preparation at lower concentration became unsuitable for encapsulation, probably because of a reduction in its gelling after exposure to high temperature during autoclaving (pattnaik et al. 1995). the conversion of encapsulated leaf and callus segments into plantlets after considerable period of storage could be attributed to the inclusion of ms salts in encapsulation matrix which serve as an artificial nutrient to the encapsulated explants for regeneration (ganapathi et al. 2001). the best storage temperature determined was 25°c for all three storage periods: 2, 4 and 6 months for leaf and callus explants encapsulated beads. the morphology and growth of regenerated shoots were not affected at 25°c. the shoots emerged from beads stored at 4ºc exhibited slow growth with necrotic and vitrification symptoms. after 6 months of storage, the per cent frequency of conversion was reduced along with death and decay of the encapsulated explants due to crackings and dehydration of the beads. saradha and paulsamy (2013) also reported similar findings in a study with hildegardia populifolia. the leaf and callus explants encapsulated synthetic seeds produced higher amount of multiple shoots on the ms medium containing 2mg/l bap compare with other plant growth regulators. it may indicate, that bap (cytokinin) is an important growth regulator for shoot initiation, as observed by other workers (gopinath et al. 2014; ummi et al. 2014). the regenerated micro-shoots were excised and implanted on to ms medium containing different concentrations of iaa, iba and naa for rhizogenesis. the best rooting performance was observed in 1mg/l naa for encapsulated beads of both explants. several studies have demonstrated the positive effect of auxins on the root formation and development (udayakumar et al. 2013; abdel jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 28 hamid et al. 2013). the in vitro developed plantlets were transferred to potting media containing garden soil, sand and vermicompost (1:1:1) for acclimatization. after one month they were planted in earthen pots containing normal garden soil and maintained in green house condition. 5 conclusion the results of the study revealed that the leaf encapsulated beads stored for four months at 25°c could be used to propagate hypochaeris radicata successfully. therefore, this procedure provides an effective for propagation of this important plant species. in addition, sowing of beads in high hills of nilgiris in suitable microclimatic sites with the advent of monsoon season (june) will enhance the population of hypochaeris radicata and can reduce the pressure upon the wild habitats where the species being endangered. . acknowledgment the authors are acknowledging dr. m. aruchami research foundation (arf/ra2012/018 dt. 12.02.2012), coimbatore for financial support to carry out the work. references abdel hamid, m.h., rosna mat, t. and sadegh, m. 2013. in vitro induction and proliferation of adventitious roots in pinapple (ananas comosus l.) cultivars of smooth cayenne and morris. aust j crop sci. 7(7): 1038-1045. aman, n. and shruti, s. 2013. production of artificial seeds from nodal region of sweet neem (murraya koenigii). j adv pharmaceu res biosci. 1(2): 71-74. anju, p., suraj, p.v., promila, p. and lucky, k.a. 2012. augmented shelf-life and regeneration competence of activated charcoal (ac) supplemented synthetic seeds in cymbidium pendulum (roxb.). sw. curr bot. 3(5): 30-34. bapat, b.a. and rao, p.s. 1998. sandalwood plantlets from synthetic seeds. plant cell rep. 7: 434-436. bekheet, s.a. 2006. a synthetic seed method through encapsulation of in vitro proliferation bulblets of garlic (allium sativum l.). arab j biotech. 9(3): 415-426. block, w. 2003. water status and thermal analysis of alginate beads used in cryopreservation of plant germplasm. cryobiology 47(1): 59-72. dennis, j.g., purohit, a. and triglano, r.n. 1991. somatic embryogenesis and development of synthetic seed technology. cr rev plant sci. 10(1): 33-61. duncan, d.b. 1955. multiple range and multiple f tests. biometrics 11(1): 1-42. jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 29 ganapathi, t.r., bapat, v.a. and rao, p.s. 1994. in vitro development of encapsulated shoot tips of cardamom. biotechnol tech. 8(4): 239-244. ganapathi, t.r., srinivas, l., suprasanna, p. and bapat, v.a. 2001. regeneration of plants from alginate-encapsulated somatic embryos of banana cv. rasthali (musa spp. aab group). in vitro cell dev biol (plant). 37: 178-181. ganapathi, t.r., suprasanna, p., kulkarni, v.m., bapat, v.a. and rao, p.s. 2002. strategies for in vitro propagation and synthetic seeds in banana. barc newsletter. founders day special issue, pp.168-176. ghosh, b. and sen, s. 1994. plant regeneration from alginate encapsulated somatic embryos of asparagus cooperi baker. plant cell rep. 13(7): 381-385. gopinath, b., gandhi, k. and saravanan, s. 2014. in vitro propagation of an important medicinal plant artemisia annua l. from axillary bud explants. adv appl sci res. 5(1): 254-258. harikrishna, k. and ong, c.h. 2002. synthetic seeds in. basic principles of biotechnology and their application in forestry. krishnapillay b (ed.). asian pacific association of forestry research institution (apafri), malaysia. pp. 87-91. jamuna, s., paulsamy, s. and karthika, k. 2013a. in vitro antibacterial activity of leaf and root extracts of hypochaeris radicata l. (asteraceae) – a medicinal plant species inhabiting the high hills of nilgiris, the western ghats. int j pharmacy pharmaceu sci. 5(1): 175-178. jamuna, s., paulsamy, s. and karthika, k. 2013b. in vitro antifungal activity of leaf and root extracts of the medicinal plant, hypochaeris radicata l. int j pharmacy pharmaceu sci. 5(3): 758-761. jamuna, s., paulsamy, s. and karthika, k. 2014. phytochemical analysis and evaluation of leaf and root parts of the medicinal herb, hypochaeris radicata l. for in vitro antioxidant activities. asian pac j trop biomed. 4(suppl. 1): s359-s367. jamuna, s. and paulsamy, s. 2013. gc-ms analysis for bioactive compounds in the methanolic leaf and root extracts of hypochaeris radicata l. (asteraceae). int j curr res. 5(12): 4070-4074. jamuna, s., paulsamy, s. and karthika, k. 2012. screening of in vitro antioxidant activity of methanolic leaf and root extracts of hypochaeris radicata l. (asteraceae). j appl pharmaceu sci. 2(7): 149-154. krishna kumar, g. and dennis thomas, t. 2012. high frequency somatic embryogenesis and synthetic seed production in clitoria ternatea linn. plant cell tiss org cult. 110: 141151. lans, c., turner, n., khan, t., brauer, g. and boepple, w. 2007. ethnoveterinary medicines used for ruminants in british columbia, canada. j ethnobiol ethnomed. 3: 11. louis, h.p., waldman, a.s., schechinger, l., govindarajoo, g. and nowick, j.s. 1998. the alginate demonstration: polymers, food science, and ionexchange. j chem edu. 75(11): 1430-1431. mithilesh, s. and rakhi, c. 2013. somatic embryogenesis in neem (azadirachta indica a. juss.): current status and biotechnological perspectives. in: junaid aslam, p.s.s. and m.p. sharma (editors). somatic embryogenesis and gene expression. narosa publishing house, new delhi, pp. 35-55. murashige, t. and skoog, f. 1962. a revised medium for rapid growth and bioassays with tobacco cultures. physiol plantarum15(3): 473-497. nor asmah, h., nor hasnida, h., noraliza, a., nashatul zaimah, n.a. and nadiah salmi, n. 2012. in vitro propagation of acacia hybrid through alginate-encapsulated shoots and axillary buds. afr j biotechnol. 11(65): 12814-12817. jamuna and paulsamy in vitro regeneration of h. radicata ruhuna journal of science (december 2014) 30 pattnaik, s.k., sahoo, y. and chand, p.k. 1995. efficient plant retrieval from alginateencapsulated vegetative buds of mature mulberry trees. sci hortic-amsterdam 61(3-4): 277-239. paulsamy, s., vijaya kumar, k.k., murugesan, m., sivashanmugam, m., senthilkumar, p. and suresh, d. 2008. medicinal and other economic plants of shola understories, the nilgiris, western ghats. internationa book distributors, dehradun, uttarakhand, india. pp. 1-112. pullaiah, t. 2006. encyclopedia of world medicinal plants. regency publication, new delhi, pp, 1-525. roy, b. and mandal, a.b. 2008. development of synthetic seeds involving androgenic and proembryos in elite indica rice. indian j biotechnol. 7: 515-519. saradha, m. and paulsamy, s. 2013. development of propagation strategies and evaluation of medicinal properties for the endangered plant species, hildegardia populifolia (roxb.) schott and endl. (sterculiaceae). ph.d., thesis submitted in bharathiar university. sharma, a., tandon, p. and kumar, a. 1992. regeneration of dendrobium wardianum warner. (orchidaceae) from synthetic seeds. indian j exp biol. 30(8): 747-748. tynsong, h., tiwari, b.k. and lynser, m.b. 2006. medicinal plants of meghalaya, india. centre for environmental studies. north-eastern hill university shillong, india. medicinal plant network news. news letter of mappa/icimdd, nmpb & mpcn, pp. 710. udayakumar, r., choi, c.w., kim, k.t., kim, s.c., kasthurirengan, s., mariashibu, t.s., sahaya rayan, j.j. and ganapathi, a. 2013. in vitro plant regeneration from epicotyl explant of withania somnifera (l.) dunal. j med plants res. 7(1): 43-52. ummi nur, a., chong boon, o., tiew sing, y. and li kiaw, l. 2014. in vitro microporpagation of stevia rebaudiana bertoni in malaysia. braz arch biol technol. 57(1): 23-28. ruhuna journal of science vol 13 (1): 70-91, june 2022 eissn: 2536-8400 faculty of science http://doi.org/10.4038/rjs.v13i1.117 university of ruhuna faculty of science, university of ruhuna sri lanka 70 analysis of statistical downscaling model (sdsm) projected future rainfall in northwestern, western and southern provinces of sri lanka mohamed riflan1, miyuru b. gunathilake,2,3, jayanga t. samarasinghe4, isuru m. bandara5, imiya m. chathuranika1 and upaka rathnayake1* 1department of civil engineering, faculty of engineering, sri lanka institute of information technology (sliit), malabe, sri lanka 2hydrology and aquatic environment, division of environment and natural resources, norwegian institute of bioeconomy and research, ås, norway 3water energy and environmental research unit, faculty of technology, university of oulu, finland 4department of earth environmental and resource sciences, university of texas, el paso, tx 79968, usa 5central engineering consultancy bureau, 415, bauddhaloka mawatha, colombo, sri lanka *correspondence: upakasanjeewa@gmail.com, orcid: https://orcid.org/0000-0002-7341-9078 received: 5th october 2021, revised: 15th june 2022, accepted: 28th june 2022 abstract even though an extensive amount of climate change studies have been carried out in different parts of the world, sri lanka is one of the least focused countries in this regard. climate projections are important and encouraged to manage the futuristic adverse impacts. identifying this research gap, future rainfall projections were carried out in three provinces in sri lanka, i.e. northwestern (puttalam and kurunegala), western (katunayake and colombo), and southern province (galle and hambantota). the canadian earth system model (canesm2) under the representation concentration pathways (rcp8.5) was downscaled using the statistical downscaling model (sdsm). non-parametric tests, including mann-kendall (mk) test and the sen’s slope estimator, were used to determine the significance of trends and magnitude of the slope of the historical trends (1990-2019) and future projected trends (20202100). the trends were analyzed for four major seasons in sri lanka, including first inter-monsoon (fim), southwest monsoon (swm), second intermonsoon (sim), and the northeast monsoon (nem). the standard error and model bias at rainfall stations were 0.014-0.034 mm and 1-1.1 respectively, which are acceptable when compared to previous studies. several significant rainfall trends were identified, including positive trends in the mid-future (2041-2070), and negative trends in the far-future (2071-2100). in addition, rainfall indices, including rx5day, r20mm, consecutive dry days (cdd), and consecutive wet days (cwd) were tested in future projected and historical rainfalls. the results of the present study will be useful for policymakers for decision-making processes in water resources management and agriculture. keywords: non-parametric tests, projected rainfall, rainfall indices, rainfall trends, statistical down scaling model (sdsm). https://rjs.ruh.ac.lk/index.php/rjs/index https://creativecommons.org/licenses/by-nc/4.0/ https://orcid.org/0000-0002-7237-5368 m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 71 1 introduction precipitation and evapotranspiration are two of the major drivers of the hydrologic cycle. these components play an indispensable role in terrestrial climatic systems linking water, energy, and carbon cycles (katul et al. 2012, vihma et al. 2016). the climate of the earth has been changed drastically due to the earth’s natural variability and anthropogenic activities. increases in prolonged drought periods and erratic rainfall events are some of the changes experienced in many parts of the world (bates et al. 2008, taylor et al. 2013). hence, studying the changes in the climate is of major importance for research and applications for many professions such as hydrologists, meteorologists, ecologists, hydropower planners, and environmentalists. climate changes naturally have adverse impacts on many sectors affecting the day-to-day life of humans and all other living beings (gunasekara et al. 2007, cui et al. 2018, qin et al. 2020). rainfall plays an important role in the formation of natural ecosystems which thrive living beings. moreover, rainfall projections are important in many sectors including weather forecasting, disaster preparation and management, agriculture, and hydrology. the statistical downscaled model (sdsm) projected an annual increase in the average temperature of 2.83°c in the 2080s for colombo under representation concentration pathways (rcp 8.5) compared to the 1961-1990 period. the annual increase in rainfall is 33% (dorji et al. 2017). from 2020 to 2100 under rcp 2.6, 4.5 and 8.5 scenarios, the future annual precipitation in the hanwella sub-catchment of the kelani river basin will have variations between 2100 to 5130 mm (dissanayake and rajapakse 2017). interestingly, no significant trend in sri lanka’s mean annual rainfall (mar) has been observed during the past century. but a higher variability is evident (jayatillake et al. 2005, chandrapala ndmc, pers. comm.) although atmospheric temperatures are steadily rising in many regions of the world, the direction of trends in rainfall varies with the geographic location (decreasing or increasing). for instance, the amount of rainfall received in the higher latitudes of the globe is increasing, while the rainfall is decreasing in the tropical and sub-tropical regions (trenberth et al. 2007). the economy of sri lanka is mostly based on agriculture, hence, any such decreases will cause detrimental impacts on the agriculture sector of the country. studies report a decreasing trend in mar during 1961-1990 of 144 mm (7%) compared to that during 1931-1960 (chandrapala 1996, jayatillake et al. 2005). therefore, climate change is not always global warming, but it is with many other variations. thus, the intergovernmental panel on climate change (ipcc 2007) defines climate change as “a change in the state of the climate that can be identified (e.g. using statistical tests) by changes in the mean and/or the variability of its properties and that persists for an extended period, typically decades or longer”. nevertheless, the increases in greenhouse gas (ghg) emissions cause intensification of the variations of the hydrologic cycle, which result in changes in water availability and the occurrence of extreme events (simonovic and li 2003, jiang et al. 2007). the germanwatch global climate risk index 2020 report has ranked sri lanka as the sixth most-affected country due to climate change in 2018 (eckstein et al. 2019). m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 72 many of the climate change studies carried out in sri lanka have demonstrated a steady rise in atmospheric temperatures (fernando and chandrapala 1992, chandrapala 1996, fernando 1997, de costa 2008). however, variations of increases and decreases in rainfall levels were observed in different parts of the country by ratnayake and herath (2005), jayatillake et al. (2005), and shantha and jayasundara (2005). in addition, the prevalence of extreme events has increased in the recent past in sri lanka (ratnayake and herath 2005, imbulana et al. 2006, premalal 2009, eriyagama and smakhtin 2010, esham and garforth 2013). furthermore, droughts have been prevalent in the mahaweli river basin, the largest river basin in sri lanka in recent decades (withanachchi et al. 2014). however, it should be noted that el-nino southern oscillation (enso) also causes variations in rainfall in sri lanka (malmgren et al. 2003). hence, besides anthropogenic reasons, enso can also affect the regional climates. sumathipala and punyadeva (1998) and punyawardena and cherry (1999) reported correlations between the southern oscillation (so) phenomenon and the seasonal rainfall of sri lanka. premalal (2013) observed a trend for below normal rainfall for the southwest monsoon season (may-september) during el nino event, while the trend is above normal during the second inter-monsoon (octobernovember). climate change significantly impacts agriculture. agriculture is the main employment sector, of which nearly 30% of the country’s population is engaged in sri lanka. this usually creates 7.8% of the gross domestic product (gdp) of the country (ministry of agriculture 2018). however, it should be noted that most of the agriculture in the dry zone is on irrigated water. hence, the changes and variabilities of the rainfall with prolonged drought periods and rainfall deficiencies will affect crop production on a greater scale (ministry of environment sri lanka 2011). on top of that, the hydropower sector, which is the clean energy sector contributes a large proportion of the country’s energy generation. nevertheless, de costa et al. (2012) have pointed out that studies carried out in sri lanka in the context of climate change using readily available global and regional climate models are still lacking. global climate models (gcms) and regional climate models (rcms) have been extensively used for the projection of future climate in different regions of the world (jha et al. 2006, agarwal et al. 2015). rcms are preferred over gcms due to the accuracy in simulating regional climates and topographic features due to the higher resolution (sharannya et al. 2018). however, caveats are still seen in rcms due to internal climate variability, imperfect conceptualization, etc. in model structures of climate models (teutschbein and seibert 2010). since bias and model uncertainties are engaged in downscaling climate models, bias correction should be performed to reduce them. bias correction methods such as linear scaling, quantile mapping, and power transformation are some of the most commonly adopted bias correction methods in literature (chen et al. 2013). in addition to statistical downscaling, dynamic downscaling methods are also available in the research world. however, dynamic downscaling methods require significant computational capacities; thus, scientists prefer linear scaling for statistical downscaling. among them (linear scaling), the m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 73 statistical downscaling model (sdsm) model is a prominent tool. the availability in the public domain is one of the attractive features of the sdsm model. although extensive research has been carried out with downscaling of gcms using the sdsm model for climate change studies in the rest of the world, a few research could be found within sri lanka. according to the best understanding of the authors’ of this paper, dorji et al. (2017), imbulana et al. (2018), bandara and weerakoon (2020), khalid et al. (2017), muhinadeen et al. (2016), de silva (2006), shantha and jayasundara (2005), de silva (2007), herath et al. (2016), nandalal et al. (2012), basnayake and vithanage (2004) and basnayake et al. (2004) have projected future rainfall in different regions of sri lanka based on climate models (i.e. using gcms, rcms and weather research forecasting (wrf) and sdsm). ratnayake et al. (2010), de silva et al. (2012) and de silva et al. (2016) have projected future rainfall in sri lanka by wrf and sdsm models and drove in hydrologic and hydraulic models to forecast flood inundation. in addition to these studies, rcm projections in conjunction with artificial neural networks (anns) have been used for rainfall forecasting (nagahamulla et al. 2011, nagahamulla et al. 2014), future hydropower projection (khaniya et al. 2020), future reservoir inflow prediction (karunanayake et al. 2020), etc. in sri lanka. dharmarathna et al. (2012) have also predicted future paddy production in the kurunegala district using the decision support system for agrotechnology transfer (dssat) model with climate inputs from the sdsm model. in addition, it is noteworthy to state, dantanarayana et al. (2021) have used numerical weather prediction models (nwp) to newscast rainfall in the colombo area. not only the rainfalls but also the future temperatures were also projected using rcm and gcms (basnayake et al. 2004, jayatillake 2004). however, more importantly, only dissanayake and rajapakse (2019) have examined the impact of future climate on streamflow. the northwestern province of sri lanka contributes a significant share for the paddy cultivation of sri lanka. since paddy cultivated through rainfed water, understanding the future water resource availability is of paramount importance. it is of major importance to understand the future rainfall in colombo, the major economic city of the country. in addition, no research has been carried out to analyze and identify the future climatic trends in north-western and southern provinces in sri lanka in detail. therefore, this study for the first time in sri lanka presents the past and future rainfall trends in northwestern, and southern provinces in sri lanka. the rainfall trends were examined in different time scales, including monthly, seasonal, and annual scales. the future rainfall data were projected and downscaled using the publicly available statistical downscaling model (sdsm). canadian earth system model (canesm2), global climate model (gcm) under representation concentration pathways (rcp8.5) were downscaled using the statistical downscaling model (sdsm) at six rainfall gauging stations. the understanding of the future availability of water resources is imperative to crop management, design of water storage, and flood mitigation measures. m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 74 2 materials and methods 2.1 study area sri lanka is surrounded by the indian ocean, and it has a geographical area of 65625 km2 (sarathchandra et al. 2021). it is located in the tropical region of the earth, in between longitudes 79°e and 82°e and latitudes 5°n and 10°n. depending on the rainfall received, the country is mainly divided into four climatic zones namely the wet zone, intermediate zone, dry zone, and semi-arid zone (alahacoon and edirisinghe 2021). according to the climate classification, the main areas like colombo, katunayake, and galle are in the wet zone, while hambantota and puttalam are in the semi-arid zone. however, kurunegala is in the intermediate climatic zone (figure 1). fig 1. selected stations in the current study the monsoon winds bring most of the precipitation to sri lanka. therefore, the rainy seasons are based on the monsoons. the first inter-monsoon (fim) occurs from march m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 75 to april and the second inter-monsoon (sim) occurs from october to november, while december to february constitutes the northeastern monsoon (nem), and the southwestern monsoon (swm) occurs from may to september (thambyahpillay 1954, wickramagamage 2010). therefore, the country has a rich rainfall throughout the year in most of the parts. the mean annual rainfall can be as high as 5500 mm for some places (e.g., watawala and kenilworth) on exposed southwest windward slopes at elevations between 1000 and 1300 m. the mean annual rainfall reaches amounts in between 800 and 1200 mm (malmgren et al. 2003) to some of the areas in the southeast and northwest of the country. in addition, the average yearly temperature for the country ranges from 26°c to 28°c. day and night temperatures may vary by 4°c to 7°c (de silva 2006). in order to assess all these climatic regions of the country, several areas, including puttalam, kurunegala, katunayake, colombo, galle, and hambantota, were considered for this research. these selected stations for this study are given in figure 1. 2.2 observed rainfall data rainfall data from six meteorological stations located in north-western (kurunegala and puttalam), western (colombo and katunayake), and southern provinces (galle and hambantota) were used in this study. observed daily rainfall data were purchased from the department of meteorology, sri lanka for 30 years (1990 to 2019). the station id, name, coordinates, elevation of these stations are listed in table 1. the obtained rainfall had missing data of less than 3%. hence, the nearest neighbor method was used to fill the missing days of these stations. table 1: stations selected for the study station id station name coordinates (e, n) elevation (m above mean sea level) 43424 puttalam (79.83, 8.03) 2 43441 kurunegala (80.37, 7.47) 116 43450 katunayake (79.88, 7.17) 8 43466 colombo (79.87, 6.90) 7 43495 galle (80.22, 6.03) 12 43497 hambantota (81.13, 6.12) 16 2.3 gcm rainfall data the second-generation canadian earth system model (canems2) was used in this study. the canesm2 couples together an atmosphere-ocean general circulation model, a land-vegetation model, and five terrestrial and oceanic interactive carbon cycle m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 76 (chylek et al. 2011). the canesm2 can be used directly in the statistical downscaling model (sdsm) since it comprises predictor variables on a daily scale. more information on canems2 can be found in https://open.canada.ca/data/en/ dataset/aa7b6823-fd1e-49ff-a6fb-68076a4a477c. therefore, canesm2 rainfalls under the representative concentration pathways 8.5 (rcp8.5) were used in future climate projections. the stations showcased in table 1 were used as the locations for the data points. 2.4 ncep/ncar reanalysis data the ncep/ncar reanalysis 1 project uses a state-of-the-art analysis and forecast system to perform data assimilation using past data from 1948 to date. these data are produced by the national oceanic and atmospheric administration’s (noaas) physical sciences laboratory. national centers for environmental prediction (ncep) predictor variables were used for model calibration and validation. the ncep data were extracted for the period of 30 years from 1990 to 2019 in this study. detailed information on ncep data is available through https://psl.noaa.gov/. 2.5 statistical methods and models mann-kendall test mann-kendal trend test (mann 1945) is a widely used non-parametric test to determine the significance of trends in meteorological variables including rainfall, temperature, humidity, etc., and hydrologic variables such as streamflow (khattak et al. 2011, dashkhuu et al. 2015). no specific assumption is required for data distribution to apply the mk test (kundzewicz and robson 2000). the test has been widely used to detect the strength of trends in many regions of the world (dumitrescu et al. 2015, rathnayake 2019, ruwangika et al. 2020, perera and rathnayake 2019). equations 1 – 4 present the mathematical formulations for the mann-kendall test. s = ∑ ∑ sgn(xj − xi) n j=i+1 n−1 i=1 (1) where xj and xi are time series values in climate data and n is number of data points and the s is the test statistics. the “sgn” function is given as the following equation. sgn(xj − xi) = { +1, > (xj − xi), 0, = (xj − xi), −1, < (xj − xi). (2) the variation of mann-kendall statistics s is given by the equation 3. https://open.canada.ca/data/en/%20dataset/aa7b6823-fd1e-49ff-a6fb-68076a4a477c https://open.canada.ca/data/en/%20dataset/aa7b6823-fd1e-49ff-a6fb-68076a4a477c https://psl.noaa.gov/ m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 77 var(s) = n(n − 1)(2n + 5) − ∑ ti(i)(i − 1)(2i + 5) m i=1 18 (3) where ti is the number of ties up to sample i. the mann-kendall statistics zc is given by the equation 4. zc = { s − 1 √var(s) , s > 0, 0, s = 0, s + 1 √var(s) , s < 0. (4) the alpha value of the mann-kendall test was varied until a trend in the variables was detected. when the calculated p value is smaller than the alpha value, then, the null hypothesis is accepted. sen’s slope estimator the detected climatic trend from the mann-kendall test is usually quantified by the sen’s slope estimator test (sen 1968). the governing equations for sen’s slope estimator are given by equations 5 – 6. the slope ti for all data pairs given as equation 5. ti = xj − xk j − k (5) and the sen’s slope estimator (qi) is given as the equation 6. qi = { t(n+1)/2, n is odd, 1 2 (tn 2 + tn+2 2 ) , n is even. (6) rclimdex computer software rclimdex was produced in the national climate data center (ncda) of noaa by byron gleason in 2001 for the calculation of extreme climatic indices (zhang and yang 2004). rclimdex 1.0 version was used for this study. the 27 core indices, suggested by the commission for climatology (ccl)/climate variability and predictability expert team for climate change detection monitoring and indices (etccdmi) can be calculated using this computer software. these 27 indices include 16 temperature and 11 precipitation related indices. rclimdex requires r1.84 or a later version for the execution of the program (downloadable at http://etccdi.pacificclimate.org/software.shtml). http://etccdi.pacificclimate.org/software.shtml m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 78 2.6 overall methodology the future climate (rainfall) predictions were carried out using the sdsm 4.2 version, which was developed by wilby et al. (2002). this comprises a multiple linear regression model (mlr) and a stochastic weather generation model (swg). the main processes of the sdsm model comprise screening of variables, calibration and validation of the model, and future climate prediction (this model is available to be downloaded from https://sdsm.org.uk/software.html). the stepwise procedure of sdsm is explained in detail in the sdsm manual and hussain et al. (2015). the stepwise development of the sdsm model is described below. first, the quality of the collected data sets was checked. this is to ensure that missing rainfall values are not presented in the input data sets. the data were also run to check the homogeneity using widely used homogeneity test (standard normal homogeneity test (snht), buishand range test, pettitt test, and von neumann ratio tests) and found out that the data are in good quality. then, the most sensible relationship between large-scale climate variables (ncep climatic data) and local scale climatic variables (rain gauge data) was identified. the strength of this relationship was determined through two statistical indicators, the significance level (𝒑) and the partial correlation (𝒓). thus, the strongest relationship can be determined among 26 predictor variables. the strongest variable has 𝒑 values closer to 0 and 𝒓 values closer to 1. next, the model calibration and validation were carried out by comparing the performance of simulated rainfall (from ncep data through the sdsm model) against observed rainfall data. this is usually compared through several statistical indicators including coefficient of determination (𝐑𝟐), and nash-sutcliffe efficiency (𝐍𝐒𝐄). the collected 30 years of data were divided into two parts. the first 20 years (2/3rd, 19902008) of data were used for calibration of the model and the next 10 years (1/3rd, 20092019) of data were used for validation. finally, the canesm climate model was used to project the climate from the years 2020 to 2100 for the northwestern, western, and southern provinces of sri lanka. the future rainfalls from 2020 to 2100 were generated. rcp8.5 was selected for future climate prediction in the canesm model for future rainfall data. rcp 8.5 scenario represents the extreme case of future climate. the developed rainfall data were then processed to identify the future rainfall trends using non-parametric tests, like mannkendall (mk) test and sen’s slope estimator. mk test and sen’s slope method, which have been frequently adopted to examine the significance of the trends and magnitude of the trends present in the hydro-meteorological variables (khattak et al. 2011). the tests were carried out on the monthly, seasonal and annual scales. in addition, the observed and project rainfalls were tested for rainfall indices, including rx5days (5-day maximum precipitation), r20mm (number of very heavy precipitation days), cdd (consecutive dry days), and cwd (consecutive wet days). the rx5days index is a measure of heavy precipitation, while the r20mm index gives a similar meaning to heavy rainfalls. cdd and cwd are two contrasting indices for the consecutive dry and wet days. the strength of these indices was tested based on https://sdsm.org.uk/software.html m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 79 the statistical 𝑝-value. very strong (vs) index can be seen for 0 < 𝑝 ≤ 0.05 and strong (s) index can be observed for 0.05 < 𝑝 ≤ 0.1. it should be noted that the more the 𝑝-value is the weaker the indices. a similar classification for the 𝑝 value can be found in khattak et al. (2011). 3 results and discussion 3.1 development of the statistical downscaling model (sdsm) as it was stated in section 2, the selection of a suitable predictor variable for calibration is the first step in sdsm model development. the selected predictor variables are listed in table 2 with corresponding 𝒑 and 𝒓 values. among the 26 predictor variables in the sdsm model, the “p*p1_fgl (1000 hpa wind speed)” was found to be the dominating predictor, while “p*p8_zgl (850 hpa relative vorticity of true wind)” and “p*p1zhgl (1000 hpa divergence of true wind)” were ranked second and third most strong predictor variables, respectively. table 2: details of selected predictor variables. station predictors 𝒑-value partial 𝒓 puttalam p*p8_zgl.dat 0.0202 0.053 p*p1zhgl.dat 0.2463 -0.026 kurunegala p*p1_fgl.dat 0.0007 -0.060 p*s850gl.dat 0.0326 0.039 katunayake p*p1_fgl.dat 0.0928 -0.031 p*p8_zgl.dat 0.1028 0.030 colombo p*p8_fgl.dat 0.0033 -0.050 p*p5_fgl.dat 0.0221 0.040 galle p*shumgl.dat 0.0083 0.043 p*p1_fgl.dat 0.0090 -0.042 hambantota p*p1_vgl.dat 0.0003 -0.078 p*p1zhgl.dat 0.0130 -0.056 the model calibration and validation processes were carried out and the selected final values of variance inflation and bias correction are listed in table 3. not all stations required a bias correction parameter (indicated by a mean bias higher than 1) for precipitation, which involves a conditional process. similar ranges of results for these indicators were previously obtained by gagnon et al. (2005). therefore, the model has a valid justification. m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 80 table 3: results of calibration procedure. station variance inflation standard error (mm) mean bias puttalam 12 0.023 1 kurunegala 14 0.019 1 katunayake 12 0.020 1.1 colombo 12 0.015 1 galle 12 0.014 1 hambantota 14 0.034 1.1 the mean monthly rainfall for observed and modeled rainfall data between 2009 2019 are shown in figure 2(a-e). (a) for puttalam (b) for kurunegala (c) for katunayake (d) for colombo (e) for galle (f) for hambantota fig 2. mean monthly rainfall for modeled and observed rainfalls. m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 81 similar monthly rainfall patterns can be observed in modeled rainfall against the observed monthly rainfall for all six stations. however, none of the stations showcase a perfect match to the observed rainfalls. it is very usual to have errors in climate projections. nevertheless, the correlation of both data sets (modeled and observed) shows some interesting results (puttalam – 0.93, kurunegala – 0.86, katunayake – 0.86, colombo – 0.77, galle – 0.94, and hambantota – 0.69). puttalam and galle have higher correlations whereas hambantota has the lowest. however, the lowest is 0.69. this is acceptable in most climate studies. therefore, it is evident that the validated sdsm model provides reasonable results when compared to the observed mean monthly rainfall. (a) for puttalam (b) for kurunegala (c) for katunayake (d) for colombo (e) for galle (f) for hambantota fig 3. seasonal variation for modeled and observed rainfalls. figure 3 showcases the seasonal variation of rainfall for modeled and observed rainfalls for the six locations (a-f). the modeled data show the usual patterns of observed rainfalls. therefore, the modeled rainfall can be used to visualize seasonal m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 82 behaviour and patterns. in addition, the southwestern monsoon showcases the highest seasonal rainfall. this is expected as the northwestern, western, and southern provinces in sri lanka usually have higher rainfall during the southwestern monsoon. figure 4 illustrates the annual rainfall variation for future projected rainfall for puttalam under the rcp8.5 emission scenario (only puttalam is shown here). the annual rainfalls were calculated from the projected daily rainfall data. the figure presents the rainfall in three clusters as stated earlier; near future (2020-2040), mid future (2041-2070), and far future (2071-2100) and shown in a straight line for illustration purposes, even though there is no direct connection from this year’s rainfall to the next year’s rainfall. fig 4. annual rainfall variation for projected rainfall. 3.2 trend analysis of historical (1990 to 2019) and future (2020 to 2100) rainfall table 4 presents the results of mk test and sen’s slope for seasonal and annual rainfalls at puttalam, kurunegala, katunayake, colombo, galle, and hambantota stations during fim, swm, sim, and nwm seasons and annual scales at 95% confidence level. even though the increasing trends were present at all stations except katunayake for the fim period, only hambantota shows a significant increasing trend, which is about 1.44 mm. the swm and nem are the dominant rainfall receiving seasons of sri lanka. however, none of the stations shows significant trends for the swm season. however, colombo shows a significant increase during the nem, and it is 4.75 mm per year, which can be a significant contribution to the water resources in the colombo district. however, colombo is in the wet zone and frequently experienced some floods. a significant increase can also be observed in sim for colombo (4.44 mm/year). therefore, the stakeholders may have to concern about these rainfall increments. increasing trends for annual rainfall were identified in colombo in previous studies (malmgren et al. 2003). increasing trends during the nem season in colombo m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 83 demonstrated in this study are in-lined with the finding of perera et al. (2020). the increasing trends in annual rainfall in galle, colombo, and katunayake are similar to the direction of trends observed by amarasinghe (2020). therefore, the results of this study are well-validated. table 4: results of mk and sen’s slope tests for seasonal and annual rainfall. station result fim swm sim nem annual puttalam p value 0.617 0.929 0.830 0.521 0.830 sen’s slope (mm/year) 1.25 -0.15 -1.07 1.95 -1.56 trend status a is is is is is kurunegala p value 0.748 0.521 0.335 0.544 0.669 sen’s slope (mm/year) 0.99 3.02 -4.1 3.54 2.16 trend status is is is is is katunayake p value 0.943 0.972 0.617 0.412 0.830 sen’s slope (mm/year) -0.20 0.17 2.2673 -2.45 2.54 trend status is is is is is colombo p value 0.392 0.617 0.044 0.034 0.354 sen’s slope (mm/year) 3.67 -3.07 4.44* 4.75* 8.72 trend status is is s s is galle p value 0.803 1.000 0.522 0.669 0.034 sen’s slope (mm/year) 1.19 0.38 7.06 1.29 11.55* trend status is is is is s hambantota p value 0.042 0.972 0.972 0.592 0.643 sen’s slope (mm/year) 1.44* -0.153 -0.100 1.285 3.250 trend status s is is is is a trend status significant (s) or insignificant (is) overall, the results obtained for annual trends in this study suggest that at most of the stations in western, north-western and southern provinces the annual rainfall is expected to increase even though most of them are insignificant. trend analysis results for projected future rainfalls are shown in table 5. none of the gauging stations have shown any significant rainfall trends in the near future (20202040). noteworthy, a distinct trend pattern cannot be observed in any temporal or spatial scales. however, positive significant trends were found in the projected rainfall for the mid future (2041-2070). puttalam and galle have shown rainfall increasing trends in the south-western monsoon, where these two regions get the usual significant rainfalls. therefore, it would be interesting to see the behavior and planning strategies of water resources in the 2040s. nevertheless, the trend patterns are turned to negative trends in the far future for the projected rainfalls. kurunegala, karunayake, and galle have shown negative trends from the 2070s. therefore, the water resources in these districts under the rcp8.5 climate scenario would be interesting to understand. thus, pre-planning would be an essential task in the 2070s world. m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 84 table 5: results of mk and sen’s slope test for future projected rainfall. station for 2020-2040 for 2041-2070 for 2071-2100 puttalam all seasons fim swm sim & nem fim swm sim puttalam is is s (1.8 mm) is is is is kurunegala is is is is is is s (-2.2 mm) katunayake is is is is s (-1 mm) is is colombo is is is is is is is galle is is s (3.3 mm) is is is is hambantota is is is is is is is trend status significant (s) or insignificant (is) 3.3 trend analysis of rainfall indices trends patterns in the tested rainfall indices are given in figure 5. the variation of rx5day, r20mm, cdd and cwd indices were presented. figure 4a presents these indices for the observed rainfalls in the 1990-2019 period. the rx5day index is either in l or w for all six stations. therefore, the heavy precipitation events were not observed in the 1990-2019 period. in addition, this observation is confirmed from the r20mm index. it is either l or w for all stations. consecutive dry days and consecutive wet days show a similar trend; however, except for a couple of higher cases (puttalam and galle in cwd and colombo in cdd). similar trend patterns can be seen in figure 4b for the 2020-2040 period for the projected rainfalls. it can be recalled herein that, none of the stations showed any potential trends from the non-parametric tests (see table 5) for a similar time slot. however, puttalam shows some increasing trends in rx5day and r20mm indices for the 2041-2070 time slot. this is in-lined with the observations from the mk test. nevertheless, a similar argument is not observed for galle, which has positive trends in mk tests, while lower or insignificant trends in rx5day and r20mm indices. therefore, a valid relationship from mk test to rainfall indices is not visible. this is well accepted in the research community and time slot 2071-2100 has confirmed it. m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 85 (a) for 1990-2019 (b) for 2020-2040 (c) for 2041-2070 (d) for 2071-2100 fig 5. trend patterns in rainfall indices. m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 86 4 summary and conclusions the current study showcases the future rainfall projections and their trends in three provinces of sri lanka, including northwestern, western, and southern provinces. these three provinces for the first time were tested for their projected rainfall under the rcp scenarios. the rainfall data from 1990 to 2019 were considered the historical period, while the future climate was projected until 2100 under the rcp8.5 scenario. rainfall trends under the non-parametric tests showcased some positive trends in mid future (2041-2070), while some negative trends in the far future (2071-2100). however, the results from the rainfall indices are inconclusive with these trends. nevertheless, based on the results of the study, adaptation options should be performed to mitigate the negative impact due to changes in the climate and increases in extreme events. therefore, this study provides some useful recommendations for providing adaptation strategies to mitigate and counteract negative impacts under future expected climatic conditions. in terms of water resources management, this study promotes rainwater harvesting not only to store rainfall water to use on dry days but also to reduce the peak surface flows on heavy rainy days. in terms of irrigation, the study proposes micro-irrigation measures whereas, changing of planting dates, and increased crop diversity are proposed for the agricultural sector. nevertheless, future studies should be carried out using different hydrometeorological variables. this study was only limited to the use of canesm-2. future studies can be implemented with other types of rcms and gcms, which are able to represent the climate of the study area well. availability of data and materials (data transparency) the datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. acknowledgements the authors gratefully acknowledge sri lanka institute of information technology (sliit) for its support in conducting this research work. two anonymous reviewers are acknowledged for the comments on the initial manuscript. references agarwal a. babel ms, maskey s. 2015. estimating the impacts and uncertainty of climate change on the hydrology and water resources of the koshi river basin. in: shrestha s., anal a., salam p., van der valk m. (eds) managing water resources under climate uncertainty. springer water. springer, cham, pp. 105-126. https://doi.org/10.1007/978-3-319-10467-6-6 https://doi.org/10.1007/978-3-319-10467-6_6 m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 87 alahacoon n, edirisinghe m. 2021. spatial variability of rainfall trends in sri lanka from 1989 to 2019 as an indication of climate change. isprs international journal of geo-information 10(2-84): 1-18. doi: 10.3390/ijgi10020084 amarasinghe ag. 2020. analysis of long-term rainfall trends in sri lanka using chrips estimates. american journal of multidisciplinary research & development (ajmrd) 2(11): 34-44. bandara im, weerakoon sb. 2020. trend analysis of historical and downscaled future climate data in deduru-oya basin, sri lanka. waso final virtual conference, norway, 15-18 december 2020. basnayake brsb, vithanage jc. 2004. rainfall change scenarios for sri lanka under the anticipated climate change. in: proceedings of the international conference on sustainable water resources management in the changing environment of the monsoon region. colombo, sri lanka. basnayake brsb, rathnasiri j, vithanage jc. 2004. rainfall and temperature scenarios for sri lanka under the anticipated climate change. 2nd aiacc regional workshop for asia and the pacific, manila, philippines, 2-5 november 2004. bates bc, kundzewicz zw, wu s, palutikof jp. (eds.) 2008. climate change and water: technical paper of the intergovernmental panel on climate change. geneva: ipcc secretariat. pp 210. chandrapala l. 1996. long term trends of rainfall and temperature in sri lanka. climate variability and agriculture, abrol yp, gadgil s, pant gb, eds., narosa publishing house, new delhi, 153-162. chen j, brissette f, chaumont d, braun m. 2013. finding appropriate bias correction methods in downscaling precipitation for hydrologic impact studies over north america. water resources research 49(7): 4187-4205. doi: 10.1002/wrcr.20331. chylek p, li j, dubey m, wang m, lesins g. 2011. observed and model simulated 20th century arctic temperature variability: canadian earth system model canesm2. atmospheric chemistry and physics 11(8): 22893-22907. doi: 10.5194/acpd-11-22893-2011. cui t, yang t, xu c, shao q, wang x, li z. 2018. assessment of the impact of climate change on flow regime at multiple temporal scales and potential ecological implications in an alpine river. stochastic environmental research and risk assessment 32(6): 1849-1866. doi: 10.1007/s00477-017-1475-z dantanarayana m, herath s, weerakoon sb. 2021. improving sub daily scale storm forecasting for kelani river basin based on temporal distribution of rain events. journal of climatology & weather forecasting 9(1): 1-9. dashkhuu d, kim jp chun, ja, lee ws. 2015. long-term trends in daily temperature extremes over mongolia. weather and climate extremes 8: 26–33. de costa wajm. 2008. climate change in sri lanka: myth or reality? evidence from long-term meteorological data. journal of the national science foundation of sri lanka 36: 63–88. http://doi.org/10.4038/jnsfsr.v36i0.8048 de costa wajm. 2012. climate change research in sri lanka – are we investigating enough? journal of the national science foundation of sri lanka, 40(4), 281–282. http://doi.org/10.4038/jnsfsr.v40i4.5041 de silva cs. 2006. impacts of climate change on water resources in sri lanka. 32nd wedc international conference. colombo: sri lanka, 13-17 november 2006. de silva cs, weatherhead ek, knox jw, rodriguez-diaz ja. 2007. predicting the impacts of climate change: a case study of paddy irrigation water requirements in sri lanka. agricultural water management 93(1-2): 19-29. https://doi.org/10.1016/j.agwat.2007.06.003 de silva mmgt, weerakoon sb, herath s, ratnayake ur, mahanama s. 2012. flood inundation mapping along the lower reach of kelani river basin under the impact of climatic change. engineer: journal of the institution of engineers sri lanka 45(2): 23–29. http://doi.org/10.4038/engineer.v45i2.6938 de-silva g, weerakoon sb, herath s. 2016. event based flood inundation mapping under the impact of climate change: a case study in lower kelani river basin, sri lanka. hydrology: current research 7(1): 7–10. dharmarathna w, herath s, weerakoon s. 2012. changing the planting date as a climate change adaptation strategy for rice production in kurunegala district, sri lanka. sustainability science 9(1): 103-111. doi: 10.1007/s11625-012-0192-2 http://doi.org/10.4038/jnsfsr.v36i0.8048 http://doi.org/10.4038/jnsfsr.v40i4.5041 https://doi.org/10.1016/j.agwat.2007.06.003 http://doi.org/10.4038/engineer.v45i2.6938 m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 88 dissanayaka k, rajapakse r. 2019. long-term precipitation trends and climate extremes in the kelani river basin, sri lanka, and their impact on streamflow variability under climate change. paddy and water environment 17(2): 281-289. doi: 10.1007/s10333-019-00721-6 dorji s, herath s, mishra bk. 2017. future climate of colombo downscaled with sdsm-neural network. climate 5(1-24): 1-11. https://www.mdpi.com/2225-1154/5/1/24 dumitrescu a, bojariu r, birsan mv, marin l. manea a. 2015. recent climatic changes in romania from observational data (1961–2013). theoretical and applied climatology 122: 111–119. https://doi.org/10.1007/s00704-014-1290-0 eckstein d, künzel v, schäfer l, winges m. 2019. global climate risk index 2020. who suffers most from extreme weather events? weather-related loss events in 2018 and 1999 to 2018. germanwatch. bonn. germany. eriyagama n, smakhtin v. 2010. observed and projected climatic changes, their impacts and adaptation options for sri lanka: a review, in proceedings of the national conference on water, food security and climate change in sri lanka. international water management institute, colombo. esham m, garforth c. 2013. climate change and agricultural adaptation in sri lanka: a review. climate and development 5(1): 66-76. doi: 10.1080/17565529.2012.762333 fernando tk, chandrapala l. 1992. global warming and rainfall variability – sri lankan situation. proceedings of the 47th annual session of sri lanka association for the advancement of science (slass).section e1. pp 138 fernando tk. 1997. recent variations of rainfall and air temperature in sri lanka. proceedings of the 53rd annual session of sri lanka association for the advancement of science (slass).section e1. pp 138. gagnon s, singh b, rousselle j, roy l. 2005. an application of the statistical downscaling model (sdsm) to simulate climatic data for streamflow modelling in québec. canadian water resources journal 30(4): 297-314. https://doi.org/10.4296/cwrj3004297 gunasekara d, yeon k, catherine t, melanie f. 2007. climate change: impacts on australian agriculture. australian commodities: forecasts and issues 14(4): 657–676. doi: https:/doi/10.3316/ielapa.970516347702109 herath hmvv, dayananda rgab, madakumbura mmgd, priyankara wms, weerakon sb. 2016. wrf weather downscaling model coupled with runoff model for anticipatory flood water management of the polgolla barrage. 20th iahr apd congress in colombo, sri lanka. 29-31 august 2016. hussain m, yusof kw, mustafa mru, afshar nr. 2015. application of statistical downscaling model (sdsm) for long term prediction of rainfall in sarawak, malaysia. wit transactions on ecology and the environment 196: 269-278. doi: 10.2495/wrm150231 imbulana kaus, wijesekara nts, neupane br. eds. 2006. sri lanka national water development report. mai&md, un-wwap, unesco and university of moratuwa, sri lanka, paris and new delhi. imbulana n, gunawaradana sk shrestha, s, datta a. 2018. projections of extreme precipitation events under climate change scenarios in mahaweli river basin of sri lanka. current science 114(7): 1495-1509. doi: 10.18520/cs/v114/i07/1495-1509. ipcc. 2007: climate change 2007: impacts, adaptation and vulnerability. contribution of working group ii to the fourth assessment report of the intergovernmental panel on climate change, parry ml, canziani of, palutikof jp, van der linden pj, hanson ce, eds., cambridge university press, cambridge, uk. pp 976. jayatillake h. 2004. will there be sufficient water under internal and external changes? walawe basin (sri lanka). in j. aerts & p. droogers, climate change in contrasting river basins: adaptation strategies for water, food and environment wallingford. uk: cabi. ix, pp 264. jayatillake hm, chandrapala l, basnayake brsb, dharmaratne ghp. 2005. water resources and climate change. proceedings of workshop on sri lanka national water development report. wijesekera nts, imbulana kaus, neupane b. eds. world water assessment programme (wwap): paris, france. https://www.mdpi.com/2225-1154/5/1/24 https://doi.org/10.1007/s00704-014-1290-0 https://www.tandfonline.com/author/gagnon%2c+s%c3%a9bastien https://www.tandfonline.com/author/singh%2c+bhawan https://www.tandfonline.com/author/rousselle%2c+jean https://www.tandfonline.com/author/roy%2c+luc https://www.tandfonline.com/toc/tcwr20/current https://www.tandfonline.com/toc/tcwr20/current https://doi.org/10.4296/cwrj3004297 https://search.informit.org/doi/10.3316/ielapa.970516347702109 https://www.witpress.com/elibrary/wit-transactions-on-ecology-and-the-environment https://www.witpress.com/elibrary/wit-transactions-on-ecology-and-the-environment m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 89 jha m, arnold jg, gassman pw, giorgi f, gu rr. 2006. climate change sensitivity assessment on upper mississippi river basin streamflows using swat. journal of the american water resources association 42 (4): 997–1015. doi: 10.1111/j.1752-1688.2006.tb04510.x jiang t, chen yd, xu cy, chen x, singh vp. 2007. comparison of hydrological impacts of climate change simulated by six hydrological models in the dongjiang basin, south china. journal of hydrology 336(3-4): 316–333. https://doi.org/10.1016/j.jhydrol.2007.01.010 karunanayake c, gunathilake mb, rathnayake u. 2020. inflow forecast of iranamadu reservoir, sri lanka, under projected climate scenarios using artificial neural networks. applied computational intelligence and soft computing 1-11, https://doi.org/10.1155/2020/8821627 katul gg, oren r, manzoni s, higgins c, parlange mb. 2012. evapotranspiration: process driving mass transport and energy exchange in the soil-plant-atmosphere-climate system. review of geophysics 50(3): 1-25. https://doi.org/10.1029/2011rg000366. khalid mm, muhinadeen m, weerakoon sb. 2017. rainfall forecast in the upper mahaweli basin in sri lanka using regcm model. journal of physics conference series 822 (1): 1-10. doi: 10.1088/17426596/822/1/012075. khaniya b, karunanayake c, gunathilake mb, rathnayake u. 2020. projection of future hydropower generation in samanalawewa power plant, sri lanka. mathematical problems in engineering. 1-11. https://doi.org/10.1155/2020/8862067 khattak ms, babel m, sharif m. 2011. hydro-meteorological trends in the upper indus river basin in pakistan. climate research 46(2): 103-119. doi:https://doi.org/10.3354/cr00957 kundzewicz z, robson a. 2000. detecting trend and other changes in hydrological data. united nations educational scientific and cultural organization: world meteorological organization. malmgren ba, hulugalla r, hayashi y, mikami t. 2003. precipitation trends in sri lanka since the 1870s and relationships to el ninosouthern oscillation. international journal of climatology 23(10):1235-1252. https://doi.org/10.1002/joc.921 mann h. 1945. non-parametric tests against trend. econmetrica 13(3):245-259. https://doi.org/10.2307/1907187 ministry of agriculture. 2018. sri lanka council for agriculture research policy, national agricultural research policy and strategy. isbn 978-955-9224-67-9 ministry of environment sri lanka. 2011. climate change vulnerability data book, ministry of environment: colombo, sri lanka. muhinadeen m, khalid mm, weerakoon sb, mutua f. 2016. comparative study of wrf & regcm weather predictions for the upper mahaweli river basin. 7th international conference on sustainable built environment 2016 (icsbe 2016) kandy. 16-18 december 2016. nagahamulla h, ratnayake u, ratnaweera a. 2014. artificial neural network ensembles in time series forecasting: an application of rainfall forecasting in sri lanka. international journal on advances in ict for emerging regions (icter) 6(2), doi:10.4038/icter.v6i2.7151 nagahamulla hrk, ratnayake ur, ratnaweera a. 2011. monsoon rainfall forecasting sri lanka using artificial neural networks. 6th international conference on industrial and information systems. 305309. doi: 10.1109/iciinfs.2011.6038085 nandalal kdw, sachindra da, ratnayake ur. 2012. application of wrf weather model to forecast precipitation in nilwala river basin. engineer 45(1): 51–64. doi:http://doi.org/10.4038/engineer.v45i1.6949 perera a, rathnayake u. 2019. impact of climate variability on hydropower generation in an un-gauged catchment: erathna run-of-the-river hydropower plant, sri lanka. applied water science 9:1-11. https://doi.org/10.1007/s13201-019-0925-9 perera a, ranasinghe t, gunathilake m, rathnayake u. 2020. comparison of different analyzing techniques in identifying rainfall trends for colombo, sri lanka. advances in meteorology 1-10, doi: 10.1155/2020/8844052 premalal khms. 2009. weather and climate trends, climate controls & risks in sri lanka. sri lanka monsoon forum. april 2009. department of meteorology, sri lanka. https://doi.org/10.1016/j.jhydrol.2007.01.010 https://doi.org/10.1155/2020/8821627 https://doi.org/10.1029/2011rg000366 https://doi.org/10.1155/2020/8862067 https://doi.org/10.3354/cr00957 https://doi.org/10.1002/joc.921 https://doi.org/10.2307/1907187 http://doi.org/10.4038/engineer.v45i1.6949 https://doi.org/10.1007/s13201-019-0925-9m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 90 premalal khms. 2013. change and behavior of rainfall pattern in sri lanka with southern oscillation (so) el nino and la nina, saarc meteorological research centre, dhaka, bangladesh punyawardena bvr, cherry nj. 1999. assessment of the predictability of the seasonal rainfall in ratnapura using southern oscillation and its two extremes. journal of the national science council of sri lanka 27(3): 187–195. qin p, xu h, liu m, xiao c, liu l, tarroja b. 2020. climate change impacts on three gorges reservoir impoundment and hydropower generation. journal of hydrology 580, https://doi.org/10.1016/j.jhydrol.2019.123922 rathnayake u. 2019. comparison of statistical methods to graphical methods in rainfall trend analysis: case studies from tropical catchments. advances in meteorology 1-10. https://doi.org/10.1155/2019/8603586 ratnayake u, herath s. 2005. changing rainfall and its impact on landslides in sri lanka. journal of mountain sciences 2: 218–224. https://doi.org/10.1007/bf02973195 ratnayake u, sachindra da, nandalal kdw. 2010. rainfall forecasting for flood prediction in the nilwala basin. proceedings of the international symposium on coastal zones and climate change: assessing the impacts and developing adaptation strategies, monash university, australia. 12–13 april 2010. ruwangika am, perera a, rathnayake u. 2020. comparison of statistical, graphical, and wavelet transform analyses for rainfall trends and patterns in badulu oya catchment, sri lanka. complexity 113. https://doi.org/10.1155/2020/7146593 sarathchandra c, abebe y, wijerathne i., aluthwattha s, wickramasinghe s, ouyang z. 2021. an overview of ecosystem service studies in a tropical biodiversity hotspot, sri lanka: key perspectives for future research. forests 12(5): 1-16. doi:10.3390/f12050540 sen pk. 1968. estimates of the regression coefficient based on kendall’s tau. journal of the american statistical association 63(324): 1379–1389. https://doi.org/10.2307/2285891 shantha wwa, jayasundara jmsb. 2005. study on changes of rainfall in the mahaweli upper watershed in sri lanka due to climatic changes and develop a correction model for global warming. paper presented at the international symposium on the stabilization of greenhouse gas concentrations. hadley centre, met office, exeter, uk. sharannya tm, mudbhatkal a, mahesha a. 2018. assessing climate change impacts on river hydrology – a case study in the western ghats of india. journal of earth system sciences 127:1-11. https://doi.org/10.1007/s12040-018-0979-3 simonovic sp, li l. 2003. methodology for assessment of climate change impacts on large-scale flood protection system. journal of water resources planning and management 129: 361–372. https://doi.org/10.1061/(asce)0733-9496(2003)129:5(361) sumathipala wl, punyadeva nbp. 1998. variation of the rainfall of sri lanka in relation to el nino. in proceedings of the annual sessions of the institute of physics, sri lanka, colombo. taylor ih, burke e, mccoll l, falloon pd, harris gr, mcneall d. 2013. the impact of climate mitigation on projections of future drought. hydrology and earth system sciences 17(6): 2339–2358. https://doi.org/10.5194/hess-17-2339-2013 teutschbein c, seibert j. 2010. regional climate models for hydrological impact studies at the catchment scale: a review of recent modeling strategies. geography compass 4(7):834–860. https://doi.org/10.1111/j.1749-8198.2010.00357.x thambyahpillay g. 1954. the rainfall rhythm in ceylon. university of ceylon review, vol. xii, 4: 224273. http://hdl.handle.net/123456789/1016 trenberth ke, jones pd, ambenje p, bojariu r, easterling d, klein ta, parker d, rahimzadeh f, renwick ja, rusticucci m, et al. 2007. observations: surface and atmospheric climate change. in climate change 2007. the physical science basis, contribution of wg 1 to the fourth assessment report of the intergovernmental panel on climate change, solomon s, qin d, manning m, chen z, marquis mc, averyt kb, tignor m, miller hl, (eds.), cambridge university press, cambridge, new york. https://doi.org/10.1016/j.jhydrol.2019.123922 https://doi.org/10.1155/2019/8603586 https://doi.org/10.1007/bf02973195 https://doi.org/10.1155/2020/7146593 https://doi.org/10.2307/2285891 https://doi.org/10.1007/s12040-018-0979-3 https://doi.org/10.1061/(asce)0733-9496(2003)129:5(361) https://doi.org/10.5194/hess-17-2339-2013 https://doi.org/10.1111/j.1749-8198.2010.00357.x http://hdl.handle.net/123456789/1016 m. riflan et al. sdsm projected future rainfall in three provinces of sri lanka ruhuna journal of science vol 13 (1): 70-91, june 2022 91 vihma t, screen j, brandi tjernström m, newton b, zhang x, popova v, deser c, holland m, prowse t. 2016. the atmospheric role in the arctic water cycle: a review on processes, past and future changes, and their impacts. journal of geophysical research biogeosciences 121(3). doi: 10.1002/2015jg003132 wickramagamage p. 2010. seasonality and spatial pattern of rainfall of sri lanka: explanatory factor analysis. international journal of climatology 30(8): 1235–1245. https://doi.org/10.1002/joc.1977 wilby rl, dawson cw, barrow em. 2002. sdsm – a decision support tool for the assessment of regional climate change impacts. environmental modelling & software 17: 147-159. withanachchi s, köpke s, withanachchi cr, pathiranage r, ploeger a. 2014. water resource management in dry zonal paddy cultivation in mahaweli river basin, sri lanka: an analysis of spatial and temporal climate change impacts and traditional knowledge. climate 2(4): 329-354. doi:10.3390/cli2040329 zhang x, yang f. 2004. rclimdex (1.0) user manual. climate research branch environment canada. available at https://acmad.net/rcc/procedure/rclimdexusermanual.pdf https://agupubs.onlinelibrary.wiley.com/action/dosearch?contribauthorstored=vihma%2c+timo https://agupubs.onlinelibrary.wiley.com/action/dosearch?contribauthorstored=screen%2c+james https://agupubs.onlinelibrary.wiley.com/action/dosearch?contribauthorstored=newton%2c+brandi https://agupubs.onlinelibrary.wiley.com/action/dosearch?contribauthorstored=newton%2c+brandi https://agupubs.onlinelibrary.wiley.com/action/dosearch?contribauthorstored=popova%2c+valeria https://agupubs.onlinelibrary.wiley.com/action/dosearch?contribauthorstored=deser%2c+clara https://agupubs.onlinelibrary.wiley.com/action/dosearch?contribauthorstored=holland%2c+marika https://agupubs.onlinelibrary.wiley.com/action/dosearch?contribauthorstored=prowse%2c+terry https://doi.org/10.1002/joc.1977 https://acmad.net/rcc/procedure/rclimdexusermanual.pdf