42 J. Hortl. Sci. Vol. 12(1) : 42-48, 2017 Genetic divergence assessment in Kale (Brassica oleracea L var. acephala (DC.) Alef.) by using the multivariate analysis S R Singh, N Ahamed, Dinesh Kumar, K K Srivatsava, Sabeena Yousuf and Abid Mir ICAR-Central Institute of Temperate Horticulture, Old Air Field Rangregth Srinagar - 190 007, J&K E-mail: srajparmar@gmail.com ABSTRACT A total of 87 genotypes collected from different geographical areas of Kashmir valley evaluated at one site to determinate genetic variability. Considerable diversity was found in different traits of horticultural importance. High coefficient of variation and wide range and mean differences of studied traits indicated the existence of wide genetic variability. Three principal component having eigen value more than one and cumulatively accounted for 84.85 percent of total variability of evaluated horticultural traits. The leaf weight, leaf length, leaf width, leaf yield / plant and yield (q/ha) were major contributing traits towards the first principal component. Similarly number of number of leaves/plant was impotent contributed traits toward principal component -II, whereas plant height was main contributing traits to principal component -III. The maximum inter cluster D2 value (731.04) was observed between cluster IV and cluster -I and followed by between cluster -V and cluster- I (677.29) and between cluster II and I (430.13).It indicated that genotypes belongings with these groups were genetically most divergent and may be use for hybridization to get better segregants. Key words: Kale, genetic diversity, principal component analysis, single linkage cluster analysis. INTRODUCTION Kale (Brassica oleracea L) is one of important leafy vegetable grouped into cooked greens belongs to cole group. This is a preferred and widely grown vegetable of Kashmir valley due to cold hardiness, higher yield and better nutritive value. It is only available as a fresh vegetable in valley during extreme cold temperature and snowing period when the area is cut off with rest of the country due to heavy snowfall. The crop is grown in valley since long and have been improved by growers through a selection. Some genotypes have become popular in the valley either by the name of grower or by the name of growing locality such as G M Dari, Khaniyari and Cowdari. A wide range of genetic diversity exists due to cross pollinating nature of crop and long growing history. Large succulent and curly leaves are characteristics of consumer preference. However, no such commercial variety is available in the region. Thus development of high yielding variety with preferable quality is the need of the region.Improvement in yield and quality is normally achieved by selecting the genotype with desirable character combinations existing in the nature or by creating the diversity with hybridization. Selection of genetically diverse parents in any purposeful breeding programme on basis of divergence would be more promising to get the heterotic F1,s and broad spectrum of variability in segregating generation ( Meena and Bahadur, 2015). In a varietal breeding for a particular growing conditions, it is essential to know about the local genetic population since their the relationship among the yield component are balanced and are in harmony with climatic and edaphic factors. Multiva ria te a nalysis is an effective tool for characterization and classification of plant genetic resources when a large number of accessions are assessed for several traits. Principal component analysis (PCA), one of multivariate analysis methods, depicts the tr a its which wer e decisive in genotype differentiation(Kovacic,1994). It enables an easier *ICAR-Central Institute of Subtropical Horticulture, Remankhera, Kakori Lucknow. Original Research Paper 43 J. Hortl. Sci. Vol. 12(1) : 42-48, 2017 understanding of impact and relationship among the different traits. However PCA alone would not give an adequate character representation in term of relative importance when multiple characters are considered simultaneously (Shalini et al,2003). To complement the results of such multivariate analysis, Single Linkage Cluster Analysis(SLCA) is employed to classify the variation. SLCA is an agglomerative technique which shows the patterns of exact genotype position in population (Ariyo and Odulaja,1991) by sorting them in distinct groups. T hus, present investigation was undertaken to assess the nature and magnitude of genetic diversity in kale accessions of Kashmir valley for different morphological traits which could be utilized in further improvement programme. MATERIAL AND METHODS Eighty seven ka le a ccessions (Brassica oleracea L var. acephala) collected from growing spots of Kashmir valley and some heterotic selection from kale lines bred at ICAR-CITH were evaluated (Table-1) . Seeds were sown in nursery in mid of August in each year and 30 days old seedling was transplanted at 45x 60 cm2 apart in 3.0x2.70m2 bed. The experiment was carried out during 2012 and 2013 at experimental farm of ICAR- Central Institute of Temperate Horticulture, Srinagar Jammu and Kashmir in randomized block design with three replications. Geographical position of the experimental site lies between latitude of 34005 N and longitude of 74050 E at an altitude of 1640 m amsl. The average maximum 19.63%C and minimum 6.52%C temperature, annual precipitation 160.72 mm and relative humidity 58.35%, evaporation 2.45mm and soil characteristics viz. pH= 6.81 and EC = 0.36 dSm”1 recorded during the cropping season. Recommended uniform agronomic and cultural practices were adopted to obtain better phenotypic expression of the characters. A total of seven quantitative traits representing to vegetative characteristics of plants related to yield and yield Table 1. Accession used in study along with code number Genetic Diversity in Kale 44 attributes were measured. Data collected on the quantitative characters were analysed using the SAS microsoft windows 9.2 (SAS Institute, 2011). Genetic diversity was studied following the Mahalanobis (1936) generalised distance (D2) extended by Rao (1952). Average intra cluster distance was calculated by following formula as suggested by Singh and Choudhry (1985). PCA a nd SLCA wer e used for the determination of genetic variation and percentage similarity within the genotypes. The PCA produce Eigen – Vectors and principal component score were used to assess the relative discriminatory power of its axis and their associated characters. The cluster procedure was used to produce a distinct group of 87 genotypes on the basis of genetic relationship while using the character variation. SLCA summarized the position of accessions analysed the position of accessions into a dendogramme at an interval of 5% level of dissimilarity starting from 100 % of level of dissimilarity. RESULTS AND DISCUSSION The eighty seven genotypes evaluated varied significantly for all horticultural traits except to average leaf weight (Table 2). The phenotypic variability revealed by coefficient variation (%) was highest for average leaf weight followed by leaf yield /plant and q/ha, number of leaves per plant which was also substantiated by wider range and mean difference values. The coefficient of variation varied from 15.12 for leaf length to 33.45 for average leaf weight. High coefficient of variation and wide range and mean differences of studied traits indicated a wide range of genetic variability, which reflects the potential of improvement in kale. Similar type of variability in germplasm of kale has been reported by Maria et al., 2002. To extract principal factors which do not require the assumption of normal distribution of proportion, principal component analysis was used a nd 8 7 ka le genot yp es b a s ed on degr ee of divergence of seven morphological traits were grouped into three principal components having Eigen va lue mor e t ha n one a nd cumu la tively accounted for 84.85 percent of total variability (Table 3). The first principal component contributed 44.59 % of total variation and was positively loaded with impotent horticultural traits viz., average leaf weight, leaf length, leaf width, leaf yield /plant and yield (q/ha), where as negatively loaded with number of leaves /plants. The second principal component responsible for 26.93 percent of total multivariate variation was positively loaded with number of leaves/plant, leaf yield per plant and yield (q/ha) where as negatively loaded with pla nt height, number of leaves and leaf width. The principal component III accounted for 13.32 % of total variation and was positively loaded with plant height ,number of leaves /plant and yield per plant where as negatively loaded with leaf weight, leaf length and leaf width. The positive and negative loading of quantitative characters reflect the positive and negative correlation trend between the components and variables and suggesting that theses principal component may be used to summarize the variables. The characters with largest absolute value closer to unit within the first component influencing the clustering than those to lower absolute value closer Table 2.Variability for different metric traits in kale genotypes J. Hortl. Sci. Vol. 12(1) : 42-48, 2017 Singh et al 45 J. Hortl. Sci. Vol. 12(1) : 42-48, 2017 Table 3. Latent vectors for seven traits of 87 genotypes of kale to zero. Thus in present study the differentiation of genotypes in different principal component was because of high contribution of few characters rather than small contribution of each characters. The characters positively in first three principal component could be in consideration while selecting the genotype with appr opriate tr aits a nd yield potential. The principal component analysis has also been used for studying the genetic variability in germplasm of many species (Ahmed et al.,2015, Singh et al., 2013). The Plot of PC- I versus PC - II indica ted the that some groups of isola ted genotypes clearly define the diversity among the evaluated germplasm. The genotypes CITH-KC- 23, CITH-KC-25, CITH-SAG-24, 2011 KLVr-12, New SAG- 27(5), Kashmiri Local, KC-12, CITH - KC-6, CITH-KC-14, CITH-44, and 2011/KLVr-5 were most divergent (Fig-1). Usually it is customary to use one importent variable from theses identified groups for improvement programme. Hence PC-I for leaf length , leaf width and leaf yield per plant should be first choice which has a largest positive loadings for these traits. Number of leaves per plant for second principal component and plant height for third principal component. The results of study are useful as it furnish the information about the groups where certain traits are more important, allowing to breeder to execute the breeding programme for specific tar get. The biologica l implica tion of principal component analysis can be quantified by contribution of different variables in each PC as revealed by eigen vector. The clustering score a mong the component a xes suggest that some relationship exist among the individuals within the cluster but do not provides the clear position of genotypes . Based on Single Linkage Cluster Analysis, the genotypes were grouped into five clusters quantifying the share genotypes and cluster mean of all traits (Table 4). The maximum number of genotypes (81 nos) were accommodated in cluster I followed by cluster - II (3Nos) and cluster III, IV and V (1 no. in each) contributing 93.24, 3.4, 1.41, 1.14 and 1.14 % respectively. On basis of cluster mean the cluster -IV was important for maximum number of leaves per plant, leaf yield per plant and yield (q/ha) cluster- II was important for average leaf weight ,cluster- III, cluster three for plant height and cluster - V for leaf length and leaf width. The cluster having high mean values of traits would contribute more positively in their offsprings if used as a parent. Rehman and Mansur (2009) and Ahmad et al, 2015 also suggested that the cluster having highest mean values may be used for hybridization programme to get better segregants. Genetic Diversity in Kale 46 Table 4. Cluster means values for 7 important horticultural traits along with number and Proportion genotypes falling in each cluster Singh et al J. Hortl. Sci. Vol. 12(1) : 42-48, 2017 Fig. 1 Bi-plot for 1st and 2nd PC for genotypes of kale in relation horticultural traits D 2 va lue estima te of genetic diver gence suggested the resolution for 87 kale genotypes in distinct five clusters with wide range of diversity in experimental material for a majority of traits (Table- 5). The maximum inter cluster D2 value (731.04) was observed between cluster IV and cluster -I and followed by between cluster -V and cluster- I (677.29) and between cluster II and I (430.13). It indicates that genotypes belongings with these gr oups wer e genetically most divergent. The selection of divergent parents based on theses cluster distance may be used in selecting the parents for the hybridization and formulating a comprehensive strategy to get a superior hybrid or superior segregants in kale. The findings are 47 Genetic Diversity in Kale J. Hortl. Sci. Vol. 12(1) : 42-48, 2017 in conformity with finding of Maria, et al., 2002, Singh, et al., 2013 and Ahmed et al.,2015 who had also indicated the accessions among the cluster separated by high D2 cluster values used in hybridization programme to obtain a wide spectrum of variation among the segregants. Dendogram drawn from the Single Linkage Cluster Analysis by using the Euclidian distance depicted the relationship and exact position of genotype in the cluster (Fig.2) All the genotypes were distinct from each other at 100 % of dissimilarity and farmed 17 cluster at 75% of dissimilarity and farmed 3 cluster at 50% of dissimilarity . The dissimilarity range from 50 to 100 % among the delineated genotypes large enough to suggest the variability for cr op impr ovement (Denton and Nwangburuka,2011) CITH-KC-23,CITH-KC-45, CIT H-SAG-24, 2011 KLVr-12, New SAG- 27(5),Kashmiri Local,KC-12, CITH -KC-6, CITH- KC-14, CITH- SAG-3 and 2011/KLVr-5 were most divergent in cluster position and may be use for hybridization to get better segregants. Ahmed et al.,2015 also reported such variability by using the single linkage cluster analysis. This genetic diversity analysis would be useful to avoid the selecting parents from genetically homogenous cluster and maintain a broad genetic base for future breeding programme. Table 5. Average intra and inter cluster distances (D2) of kale genotypes Fig 2. Dendogram depicting the genetic relationship among the kale genotypes based horticultural traits produced by ASH analysis (scale-Euclidian distance) 48 (MS Received 11 September 2016, Revised 20 May 2017, Accepted 24 June 2017) J. Hortl. Sci. Vol. 12(1) : 42-48, 2017 REFERENCES Ahmed,N.,Singh,S.R., Lal S., Mir K. A., Asima, A.,Habib, K. and Salmani, M. 2015. Assessment of genetic diversity in brinjal (Solanum melongena L.) genotypes using multivariate analysis. Indian J. Hort. 71:494-498. Ario, O. J. and Odulaza, A. 1991. Numerical analysis of variation among accessions of okra. (A. esculentus L. Moench). Malvaceae. Ann. Bot. 67:527-531 Denton, O. A. and Nwangburuka, C. C. 2011.Genetic Variability in Eighteen Cultivars of Solanum anguiviam L. using Principal component Analysis (PCA) and Single Linkage Cluster Analysis (SLCA). An of Biol. Res. 2 :62-67 Kovacic, Z. 1994. Multivariate analysis. Faculty of Economics, University of Belgrade (In Serbian) 293 p. Mahalanobis, P. C. 1936. On the generalized distance in Statistics. Proc. Nati. Inst. Sci. India. 2: 49-55. Maria, E.C., Ana, P., Soengas, P. and Amando. O . 2002. Morphlogical characterization of kale population from North Western Spain. Euphytica: 129:25-32. Meena, O. P. and Bahadur, V. 2015. Breeding potential of indeterminate tomato (Solanum lycopersicum L) accessions using D2 analysis. SARRAO J. Bree. and Gen 47(1)49-59. Rahman. M. M. and Mansur. A. M. 2009. Genetic divergence analysis of lime. J. Ban. Agri. Univ. 7: 33–37. Rao, C. R. 1952. Advanced Statistical Methods in Biometrics Research. John Wiley and Sons, New York, pp. 357- 69 SAS Istitute.2011.SAS enterprise guide.Version9.2SAS, Institute,Cary NC,USA. Shalini, M., Sharma. S., Gupta, M. M., Shashi, K. 2003.Evaluation of an Indian germ plasm collection of medicinal plants Bacopa monnieri (L) Pennell by use of multivariate aprochess.Euphytica.133:255-65. Singh, R. K. and Choudhry, B. D. 1985. Biometrical Methods i n Quant i tat i ve Gene t ic A nal y si s . Ka l ya n i Publishers, New Delhi, p. 318. Singh, S. R., Ahmed, N., Lal, S., Ganaie, S. A., Mudasir, A., Nusarat, J. and Asima, A. 2013. Determination of genetic diversity in onion (Allium cepa L) by multivariate analysis under long day conditions. Afri. J. Agri. Res. 8: 5599-606. Singh et al 00 Content JH June 2017.pdf 00 SPH-Journal November new 12.pdf 01 SPH-Journal November new 01.pdf 02 SPH-Journal November new 04.pdf 03 SPH-Journal November new 05.pdf 04 SPH-Journal November new 07.pdf 05 SPH-Journal November new 09.pdf 06 SPH-Journal November new 10.pdf 07 SPH-Journal November new 11.pdf 08 SPH-Journal November new 13.pdf 09 SPH-Journal November new 03.pdf 10 SPH-Journal November new 06.pdf 11 SPH-Journal November new 08.pdf