Genetic diversity and trait association between growth, yield and seed component of Jatropha curcas (L.) source collection from Indian sub-continent.

Seventy-two Jatropha curcas clonal accessions, collected from different parts of India, were grown in randomized block design and evaluated for 13 characters. These showed significant differences in growth, yield and seed characteristics, except for number of primary branches per plant. The accessions AFRI-35, AFRI-52 and AFRI-56 showed above mean value for 12 growth, fruit and seed yield characters except for number of primary branches, seed width and seed thickness. Accessions AFRI-12, AFRI-20, AFRI-21 and AFRI-39 showed above mean value for 11 traits. All the growth traits except for number of primary branches were positively correlated with fruit and seed yield. The magnitude of genotypic correlation coefficient (rg) was higher than their corresponding phenotypic coefficient of correlations (rp) for almost all the growth and yield characters, indicating the strong genetic linkage among all the characters. Kernel oil percentage was positively correlated with 20-seed weight and seed/kernel ratio. Seed/kernel ratio and oil percentage were negatively and non-significantly correlated with seed length and width. Principal component analysis revealed that plant height, collar diameter, number of secondary branches, crown diameter, fruit yield, seed yield, 20-seed weight, seed length, seed width, seed thickness, S/K ratio and oil percentage could be used as characters to distinguish the germplasm entries. K-means clustering revealed that trees from different geographic regions were grouped together in a cluster as well as trees from the same geographical area were placed in different clusters, suggesting that geographical diversity did not go hand in hand with genetic diversity. In addition, clustering identified promising accessions with favourable traits for future establishment of elite Seedling Seed Orchard (SSO) and Clonal Seed Orchard (CSO) for varietal and hybridization programmes.   Key words: Biodiesel, clustering, genetic variation, growth performance, principal component analysis.