International Journal of Aquatic Biology (2013) 1(5): 202-208 ISSN: 2322-5270 Journal homepage: www.NPAJournals.com © 2013 NPAJournals. All rights reserved Original Article Evaluation of water quality using TOPSIS method in the Zaringol Stream (Golestan Province, Iran) Zohreh Mazaheri Kohanestani*1, Rasoul Ghorbani 1, Abdolazim Fazel2 1Department of Fisheries, Gorgan University of Agricultural Sciences and Natural Resources, P.O. Box 45165-386, Gorgan, Iran. 2Department of Fisheries, University of Guilan, Guilan, Iran. Article history: Received 8 May 2013 Accepted 21 August 2013 Available online 2 5 October 2013 Keywords: Pollution TOPSIS method Water quality Zaringol Stream Abstract: In order to evaluate water quality condition, biochemical oxygen demand (BOD), dissolved oxygen (DO), temperature, pH, turbidity, total suspended solid (TSS), phosphate (PO4-), nitrate (NO3-) and fecal coliform were measured seasonally from 9 sites from November 2009 to August 2010 in Zaringol Stream. Water quality condition was estimated using TOPSIS method. Comparison of TOPSIS values in different sampling stations showed the minimum (0.230) and maximum values (0.604) are in points 1 and 5, respectively. According this result, point 1 had the best water quality condition and point 5 had the lowest quality. Also, Seasonal results of TOPSIS values showed that the maximum value was found in spring. Discharge of effluents from land uses located along the stream specifically, trout farms and starting agricultural activity and production process in spring and summer are most important reasons for decreasing of water quality. TOPSIS estimates values ≤0.5 for almost stations and seasons. It shows Zaringol Stream has an average water quality. Introduction Rivers play an important role in watersheds for carrying off wastewater and run off from farmland and are the most susceptible water body to pollutants (Yu et al., 2003; Singh et al., 2004; Wang et al., 2007). The constant discharges of wastewater and seasonal surface run-off have a strong effect on the river, water quality, human health and aquatic organisms (Kazi et al., 2009). Water quality zoning is critical to identify preferred potential usages of water resources and source of pollution to optimize water usage and management, especially, in rivers and streams (Simenove et al., 2003; Karimian et al., 2007). Making decision in the field of water resource management represents a task of high importance and responsibility (Kirilov et al., 2009). Choosing the best option is technically challenging as there is * Corresponding author: Zohreh Mazaheri Kohanestani E-mail address: zohremazaheri_65@yahoo.com no scientific tools to predict future impacts of alternative management actions (Gao and Hailu, 2011). Since determining of water quality condition is done with numerous different parameters could effect on water quality with different ways as it seems antithetic sometimes, managers use valid index to combine different parameters and calculate a value for taking decision more easily such as NSFWQI, WQI, etc. (Liou et al., 2003; Heernandez- Romero et al., 2004; Simoes et al., 2008). These indexes are comprehensive and common in water quality zoning (Jonnalagadda and Mhere, 2001), though they can throw us in making mistakes when their value are in border ranges of two water quality classes (for example the value belongs to class "bad" but is also so near to the next water quality class "average") . 203 Mazaheri Kohanestani et al/ International Journal of Aquatic Biology (2013) 1(5): 202-208 Multivariate criteria decision making (MCDM) is one of research fields of management sciences have expanded in different applied researches based on its need recently and It makes it possible to take decision wisely (Saati et al., 2007). TOPSIS is the most famous classical MCDM technique described by Hwang and Yoon (1981) for the first time. In TOPSIS technique, the basic solution method is defining positive and negative ideal (non-ideal) solution (Biorani and Ghofran, 2009). Positive ideal solution includes the best available value of parameters while the non-ideal one is made of the worst available value of parameters. Finally, the best answer has both the shortest distance from the ideal solution and the longest from the non-ideal (Saati et al., 2007). Simplicity, rationality, comprehensibility, good computational efficiency and ability to measure the relative performance for each alternative in a simple mathematical form are some of the advantages of TOPSIS methods (Roszkowska, 2010). Zaringol Stream with 22 kilometer length is one of the Gorgan-Rud brunches and has important role in water supply of agricultural, aquaculture and domestic usages (Abdoli and Rahmani, 2002). In recent years, great changes have taken in marginal regions lead to severe effect of its water quality though, there is no report on water quality because of limited hydrometric station along the stream. So, the aim of this study is to evaluate water quality using TOPSIS method in Zaringol Stream, the Golestan province. Materials and Methods This study was carried out in Zaringol Stream- Golestan Province. Water was sampled from 9 sites along the stream during December 2009 to September 2010, seasonally (Fig. 1). According to Figure 1, after station 1 as the nearest station to the spring and a less polluted station, there are two trout farms located at station 2 and 5 and their actual capacities are 15 and 7 tons, respectively. Other stations are located along the stream where the agriculture is the dominated land use. Nine water quality parameters were measured including biological oxygen demand (BOD5), dissolved oxygen (DO), nitrate (NO-3), phosphate (PO-4), temperature (ΔT), pH, turbidity, fecal coliform and total suspended solid (TSS) by Water checker u-10 and Spectrophotometer. Decision Matrix (m×n) was made as: A1, A2, …, Am are alternatives; C1, C2, …, Cn are evaluation factors; Gij means evaluation rate of Ai related to factor Cj and W is evaluation factor weight were calculated by AHP (eigenvector) method (Gao and Hailu, 2011) (Table 1). 𝑊𝑖 = 𝑤1, 𝑤2, … , 𝑤𝑛 , ∑ 𝑊𝑖=1 Next, data were normalized to remove variation of each factor (Pimerol and Romero, 2000). 𝑝𝑖𝑗 = 𝐺𝑖𝑗 √∑ (𝐺𝑖𝑗)2𝑚𝑖1 ∀𝑖 = 1, ,2, … , 𝑚 𝑗 = 1,2, … , 𝑛 Figure 1. Site of water sampling-Zaringol Stream. 203 204 Mazaheri Kohanestani et al/ International Journal of Aquatic Biology (2013) 1(5): 202-208 Then weight of evaluation factors was affected by multiplying its vector to decision matrix and provided a normalized weighted decision matrix. Positive ideal (A+) and non-ideal (A-) solutions were determined and distance index calculated. "K" and "l" are associated with benefit and cost criteria, respectively (Hwang and Yoon, 1981). 𝐴+ = {(𝑀𝑎𝑥 𝑣𝑖𝑗 𝑗 ∈ 𝐽), (𝑀𝑖𝑛 𝑣𝑖𝑗 𝑗 ∈ 𝐽′); 𝑖 = 1,2, 𝐾, 𝑚} = {𝑣1 +, 𝑣2 +, . . , 𝑣𝑛 +} 𝐴− = {(𝑀𝑖𝑛 𝑣𝑖𝑗 𝑗 ∈ 𝐽), (𝑀𝑎𝑥 𝑣𝑖𝑗 𝑗 ∈ 𝐽′); 𝑖 = 1,2, 𝐾, 𝑚} = {𝑣1 −, 𝑣2 −, . . , 𝑣𝑛 −} 𝐽 = {𝐾|𝐾 = 1,2, 𝐾𝑛} 𝐽′ = {𝑙|𝑙 = 1,2, 𝐾𝑛} parameter Unit Weight DO Saturate (%) 0.17 Fecal coli form Colony/100ml 0.16 pH --- 0.11 BOD5 ppm 0.11 ΔT ◦C 0.1 NO3 ppm 0.1 PO4 ppm 0.1 Turbidity NTU 0.08 T.S.S ppm 0.07 Table 1. Weight of water quality parameters used in TOPSIS method. Parameter/station 1 2 3 4 5 6 7 8 9 Log Turbidity (NTU) 1.46a 2.36b 2.16ab 2.33b 2.49b 2.08ab 2.22b 2.05ab 2.03ab TSS (ppm) 0.20a 1.06c 0.80b 0.87b 0.95b 0.82b 0.95b 0.84b 0.91b PO4 (ppm) 0.36ab 0.88b 0.25ab 0.16a 0.2a 0.12a 0.08a 0.06a 0.11a NO3 (ppm) 1.38a 2.18a 1.40a 0.63a 2a 2.08a 2.28a 1.25a 2.35a pH 8.75a 8.58a 8.75a 8.77a 8.36a 8.56a 8.66a 8.37a 8.30a T (◦C) 14.05a 16.65a 17.18a 20.82a 16.98a 18.68a 19.3a 17.95a 19.35a DO (%) 66.50b 53.25a 66.25b 66b 54.75a 59.25ab 59.50ab 58.50ab 58.75ab BOD5 (ppm) 1.78a 2.75bc 2.38ab 2.23ab 2.83bc 3.30c 3.45c 2.65bc 2.68bc Fecal coli form×105 (counts/100 ml) 1a 2.2bc 2.13bc 1.63ab 2.75c 2.2bc 1.93bc 1.55ab 1.40ab Data presented as mean. Similar letter shows no difference between stations. Table 2. Water quality parameters of different stations - Zaringol Stream. Station 𝑑𝑖 + 𝑑𝑖 − 𝐶𝑖  Rank 1 0.0538 0.0161 0.230 9 2 0.0286 0.0401 0.584 2 3 0.0336 0.0269 0.468 5 4 0.0383 0.306 0.444 6 5 0.0303 0.0461 0.603 1 6 0.0303 0.0335 0.525 4 7 0.0269 0.0371 0.580 3 8 0.0404 0.0221 0.353 8 9 0.414 0.0257 0.383 7 Table 3. Results of TOPSIS method of different stations-Zaringol Stream. 205 Mazaheri Kohanestani et al/ International Journal of Aquatic Biology (2013) 1(5): 202-208 Finally after calculating the distance from positive ideal solution (𝑑𝑖+), non-ideal solution (𝑑𝑖−) and relative nearest vicinity from ideal answer (𝐶𝑖*), values were arranged as an ascending order (Hwang and Yoon, 1981). Maximum value (nearest one to 1) has the highest preference. 𝑑𝑖 + = √∑(𝑉𝑗 − 𝑉𝑗 +) 2 𝑛 𝑗=1 ∀𝑖 = 1,2, … , 𝑚 𝑑𝑖 − = √∑(𝑉𝑗 − 𝑉𝑗 −) 2 𝑛 𝑗=1 ∀𝑖 = 1,2, … , 𝑚 𝐶𝑖 + = 𝑑𝑖 − (𝑑𝑖 − + 𝑑𝑖 +) ∀𝑖 = 1,2, … , 𝑚 0 ≤ 𝐶𝑖  ≤ 1 Data checked for normality distribution with the Kolmogorov-Smirnov test. Spatial and temporal variation of water quality parameter were analyzed using one-way ANOVA and Duncan's post-hoc test, assuming a significant level of α=0.05 by SPSS 17 software package. Results Data on water quality parameters of different stations are given Table 2. Considering that the point 1 does not expose to the pollution resources, so it was considered as the test station. According to Table 2, maximum turbidity belongs to stations 2 and 5. Also maximum values of total suspended solid, phosphate were measured in station 2. Dissolved oxygen decreased in station 2 and 5 significantly. Some of water quality parameters had no significant difference between different stations (i.e. pH, temperature and nitrate). As the aim of calculation was to detect polluted stations in Zaringol Stream, so it should be considered that the nearest value to 1 shows more water pollution (Table 3). Maximum and minimum relative nearest vicinity from ideal answer (identification polluted points) calculated for station 5 and 1 respectively. In the other hand the most polluted station is 5 and the least one is number 1. Comparison water quality parameters between different seasons show significant increase in turbidity, temperature, BOD5 and fecal coli form in spring and NO3 in summer. pH decreased in summer, Parameter/station Spring Summer Autumn Winter Log Turbidity (NTU) 2.24b 2.13ab 1.59a 2.40b TSS (ppm) 0.6a 1.01b 1.11b 0.57a PO4 (ppm) 0.39a 0.33a 0.15a 0.11a NO3 (ppm) 1.73b 2.28b 2.51b 0.38a pH 8.56b 8.33a 8.67b 8.71b T (◦C) 25.1c 21.44b 10.76a 14.23a DO (%) 56.78a 63a 60.33a 61.11a BOD5 (ppm) 2.93b 2.71ab 2.91b 2.12a Fecal coli form×105 (counts/100 ml) 2.43b 1.91b 1.60a 1.51a Data present as mean. Similar letter shows no difference between stations. Table 4. Water quality parameters at different seasons - Zaringol Stream. Table 5. Results of TOPSIS method of different seasons - Zaringol Stream. Seasons 𝑑𝑖 + 𝑑𝑖 − 𝐶𝑖  Rank Spring 0.0238 0.0685 0.742 1 Summer 0.0443 0.0457 0.507 3 Autumn 0.0459 0.0501 0.522 2 Winter 0.0674 0.0210 0.237 4 205 206 Mazaheri Kohanestani et al/ International Journal of Aquatic Biology (2013) 1(5): 202-208 too. PO4 and COD and Salinity did not show significant difference between seasons (Table 4; P<0.05). Relative nearest vicinity from ideal answer in different stations show that winter, summer, autumn and spring rank from high water quality to low, respectively. Discussion Results of water quality parameters at different stations of Zaringol Stream showed that the minimum values of total suspended solids, turbidity, biological oxygen demand, fecal coli forms and maximum value of dissolved oxygen were measured in station 1. According to these results, it was the least polluted station, as it is mentioned before. Results of TOPSIS ranking show the minimum value (or best water quality condition) in station 1, so it confirms our claim. After station 1, a decline could be observed in water quality and some of critical parameters such as biological oxygen demand and fecal coli forms which show water pollution clearly, increased significantly. Similar to the results of water quality parameters, an increasing trend can be observed in the TOPSIS values. Based on the TOPSIS ranking, station 2 had the second grade of water pollution and station 5 was the most polluted station and had the nearest TOPSIS value to 1. Since stations 2 and 5 are located after trout farms and the concentrations of some ions like phosphate (significantly) and nitrate (not significantly) increased in these stations, it can be infer that entrance of farms effluents had an important effect on water quality condition. The TOPSIS values showed that station 5 was more polluted than station 2. It is reported that the severity of effect of fish farms effluents on water quality depends on volume and concentration of substances, flow rate of water and time of effluent discharge (Pillay, 2004). Although the actual capacity of first fish farm was more than second ones (located before station 5), station 5 was more polluted because the water loaded a lot of pollution from station 2. Also adding a branch to the main stream flow and bringing a new volume of substances was another probable reason for the highest TOPSIS value in station 5. Comparison of increasing rate of TOPSIS value between station 1 to 2 (0.354) and station 4 to 5 (0.159) confirms that first farm discharge more effluents into the stream than second one. The increasing trend in TOPSIS values did not resume along the stream and water quality trend to better condition in other stations. Self-purification power of stream improved water quality condition along the stream; though others land uses discharge along the stream prevented it to go back to the first condition again completely. There are some reports on environmental effect aquaculture (Manoochehri et al., 2010; Uzbilek Kirkagac et al., 2009; Pulatsu et al., 2004; Mmochi et al., 2002) which confirm effect of fish farm effluents on water quality condition. Our results agree with them. The TOPSIS values show maximum value and the worst condition in spring. Due to climate condition, rain fall decrease in warm season, so because of decreasing of stream flow, concentrations of contaminated components increase and generally stream is in acute condition. In addition to agricultural activity and reproduction period of trout farm start in spring and continue in summer. Therefore water quality of warm seasons may affect more than cold seasons. Based on Relative nearest vicinity from ideal answer ranged from 0.23 to 0.6 and the theoretical highest value (Ci* =1) we can conclude that Zaringol water quality condition is almost average because most of the TOPSIS values of stations and seasons are ≤0.5. In summary, the results of our research showed TOPSIS method declared water quality variations clearly as they were expected based on analyzing water quality parameters and can determine polluted stations. Therefore this is an applied simple method and can be used for managers to make decision easily while they are faced to several complicated parameters. 207 Mazaheri Kohanestani et al/ International Journal of Aquatic Biology (2013) 1(5): 202-208 References Abdoli A., Rahmani H. (2002). Study on diet composition of Neogobius fluviatilis, N. melanostomus in Madarsoo stream, Golestan national park. Journal of Agriculture Science and Natural Resources of Gorgan, 8: 3-16. (in Persian). Biorani H., Ghofran A. (2009). Determining and using technique for order preference by similarity ideal Solution (TOPSIS) for ranking of civic region based on crim. Detective, 2: 109-128. (in Persian) Gao L., Hailu A. (2011). Evaluating the effects of area closure for recreational fishing in a coral reef ecosystem: The benefits of an integrated economic and biophysical modeling. Ecological Economics, 70: 1735-1745. Hernandez-Romero A.H., Tovilla-Hernandez C., Malo E.A., Bello-Mendoza R. (2004). Water quality and presence of pesticides in a tropical coastal wetland in southern Mexico. Marine Pollution Bulletin, 48: 1130–1141. Hwang C.L., Yoon K. (1981). Multiple attribute decision making: Methods and application, Springer, 139 p. Jonnalagadda S.B., Mhere, G. (2001). Water quality of the Odzi River in the eastern highlands of Zimbabwe. Water Research, 35: 2371-2376. Karimian A., Jafarzadeh N., Nabizaheh R., Afkhami M. (2007). Zoning of water quality bases on WQI index, Zohreh river case study. International of Journal Water Engineering, 18: 53-62. (in Persian) Kazi T.G., Arain M.B., Jamali M.K., Jalbani N., Afridi H.I., Sarfaraz R.A. (2009). Assessment of water quality of polluted lake using multivariate statistical techniques; a case study. Ecotoxicology and Environmental Safety, 72: 301-309. Kirilov L., Bournaski E., Iliev R. (2009). Modeling and decision analysis with multiple objectives of water quality management problems. Problems of Engineering Cybernetics and Robotics, 60: 55- 65. Liou S.M., Lo S.L., Hu C.Y. (2003). Application of two stages fuzzy set theory to river quality evaluation in Taiwan. Water Research, 37: 1406– 1416. Manoochehri H., Abedi M., Hashemi M., Boustani F. (2010). An assessment of fish processing plant discharge effluent on the Tizab River. World Applied Science Journal, Engineering and Technology, 70: 145-147. May A.M., Mutasem E., Mark D.S., John N.L. (2006). Factors influencing development of management strategies for the Abou Ali River in Lebanon. Science of the Total Environment, 362: 15- 30. Mmochi A.J., Dubi A.M., Mamboya F.A., Mwandya A.W. (2002). Effects of fish culture on water quality of an integrated mariculture pond system. Western Indian Ocean journal and Marine Sciences, 1: 53-63. Pillay T.V.R. (2004). Aquaculture and the environment. Blackwell science, 212 p. Pimerol J.C., Romero S.B. (2000). Multi-criteria decision in management: principles and practice, Kluwer Academic Publisher, 150 p. Pulatsu S., Rad F., Koksal G., Aydin F., Caglan Karasu Benli A. (2004). The impact of Rainbow Trout farm effluents on water quality of Karasu stream, Turkey. Turkish Journal of Fisheries and Aquatic Sciences, 4: 9-15. Roszkowska E. (2011). Multi-criteria decision making models by applying the TOPSIS method to crisp and interval data, In: T. Trzaskalik, T. Wachowicz (Ed.), Multiple Criteria Decision Making. The Publisher of the University of Economics in Katowice. pp: 343-363. Saati S., Hatami Marbini A., Makoei A. (2007). Group decision by TOPSIS fuzzy. Applied Mathematics Journal of Lahijan University, 4: 21- 34. (in Persian) Simeonov V., Stratis J.A., Samara C., Zachariadis G., Voutsa D., Anthemidis A. (2003). Assessment of the surface water quality in Northern Greece. Water Research, 37: 4119-4124. Simoes F., Moreira A.B., Bisinoti M.C., Gimenez S., Santos M. (2008). Water quality index as a simple indicator of aquaculture effects on aquatic bodies. Ecological Indicator, 38: 476-480. Singh K.P., Malik A., Mohan D., Sinba S. (2004). Multivariate statistical techniques for the evaluation 207 208 Mazaheri Kohanestani et al/ International Journal of Aquatic Biology (2013) 1(5): 202-208 of spatial and temporal variations in water quality of Gomi River (India): a case study. Water Research, 38: 3980-3992. Uzbilek Kirkagac M., Pulatsu S., Topcu A. (2009). Trout farm effluent effects on water sediment quality and benthos. Clean, 37: 386-391. Wang X.L., Lu Y.L., Han J.Y., He G.Z., Wang T.Y. (2007). Identification of anthropogenic influences on water quality of rivers in Taihu watershed. Journal of Environmental Sciences-China, 19: 475-481. Yu S., Shang J. (2003). Factor analysis and dynamics of water quality of the Songhua River, Northeast China. Water, Air and Soil Pollution, 144: 159-169. .