Multiple regression and classificatory analysis in Vigna Radiata (L.)

Abstract

Abstract Thirty genotypes of Vigna radiata (L.) Wilczek obtained from AVRDC (Asian Vegetable Research and Development Centre), Taiwan were subjected to multivariate statistical analysis of regression and classification with regard to eight principal plant characters. Multiple regression analysis revealed that the characters under study contributed 71% of the total variation towards the regression of yield. It was also observed that number of pods, days to maturity, length of pod and 100-seed weight are relatively important in that order to the rest of the plant characters. The thirty genotypes were classified into seven distinct clusters based on the D 2 -values. Among the seven group constellation, two clusters each consisted of a single genotype viz., V-2273 (most of the plant characters on the lower range) and V-3404 for which pod length and 100-seed weight were found to be maximum. Of the total divergence, days to maturity contributed a maximum of 61% towards divergence.