The present study was aimed to assess genetic variation among rice germplasm lines employing a multivariate biometrical approach, viz., Principal Component Analysis. Analysis of variance as per Augmented Block Design indicated the presence of sufficient genetic variation among rice germplasm lines, while estimates of components of variation revealed a maximum contribution of genotypic variance to the phenotypic variance, suggesting its exploitation through selection and hybridization. Simultaneously, correlation estimates revealed a significant positive association of grain yield with panicle length, 1000 grain weight and grain length indicating suitability of these traits for indirect selection. D2 statistics grouped germplasm lines into sixteen clusters and among these cluster I consists of 67 germplasm lines forming the largest cluster followed by cluster III (26 lines), cluster II (18 lines), cluster IV (15 lines), cluster V (13 lines), cluster VI (5 lines), cluster VIII (3 lines), cluster VII and cluster IX (2 lines each), while clusters X, XI, XII, XIII, XIV, XV, XVI (1 line each). Inter cluster distances were found to be higher than intra cluster distances and the maximum inter cluster distance was observed between cluster IX and VI (73.47) followed by cluster IX and X (73.14).
Principal component analysis transformed eleven interrelated variables into four major principal components having an eigen value of more than 1, thereby, indicating that these components are responsible for a higher magnitude of variance in the population (76.60 per cent). Among these principal components, the first component accounted for 32.80 per cent of the total variation while, the second, third and fourth components explained 18.50 per cent, 13.90 per cent and 11.40 per cent of total variation, respectively. Factor loadings of the principal components revealed that principal component 1 had high positive loadings for grain length, panicle length, length/breadth ratio, 1000 grain weight and plant height whereas, principal component 2 had high positive loadings for total number of tillers per plant and number of effective tillers per plant indicating that the first two principal factors can be collectively designated as yield attributing factors. Principal component 3 had high positive loadings for grain breadth, 1000 grain weight and plant height, whereas, principal component 4 had high positive loadings for 1000 grain weight, days to maturity and days to 50% flowering. PCA biplots revealed that germplasm lines viz., GPL-1, GPL-4, GPL-131, GPL-128, GPL-20, GPL-127, GPL-135, GPL-100, GPL-130 were found to be superior performers for desirable traits and can be used as parents in hybridization programme.