Multivariate analysis in relation to breeding system in okra

Abstract

Fifty okra genotypes of different eco-geographical origin maintained at Indian Institute of Vegetable Research, Varanasi were evaluated to study the genetic divergence for fruit yield/plant and its 9 component traits following multivariate and canonical analyses. The genotypes were grouped in 15 clusters and confirmed by canonical analysis. Sixty eight per cent genotypes (34/50) were genetically close to each other and grouped in 5 clusters (I, II, III, IV, V), while apparent diversity was noticed for 32 per cent of the genotypes who diversed into rest 10 clusters (VI-XV). Inter-cluster distance ranged from 21.16 to 80.0. The maximum was between X and XV followed by X and XIII (79.69) and IX and X (77.09). The genotypes in cluster X, XIII, and XV had greater potential as breeding stock by virtue of high mean values of one or more component characters and high statistical distance among them. Based on findings of high cluster mean of component trait and inter-cluster distance among clusters, a breeding plan has been discussed.