Multivariate analysis in sorghum (Sorghum bicolor) germplasm for yield contributing quantitative traits

Multivariate analyses was carried out for 13 yield-contributing quantitative traits observed in 102 sorghum germplasm accessions evaluated at Department of Millets, TNAU, Coimbatore during Summer, 2021. The Euclidean distances obtained were subjected to Ward clustering algorithm which grouped the genotypes into eight clusters. The clusters VI and VII were observed as the largest and cluster VIII was the smallest. Clusters II and VIII were found to possess elite germplasm with superior traits namely, flag leaf length, flag leaf breadth, flag leaf area, number of leaves, 100-seed weight and grain yield per plant. Principal component analysis revealed that yield contributing traits were separated into 13 principal components of which four components had eigen values more than one. Seventy-three percentage of total variation was observed in the first four principal components. This suggests that the first principal component is the major contributor to the total variation in the population. Traits viz., plant height, days to 50% flowering, stem diameter, flag leaf length, flag leaf breadth, flag leaf area , number of leaves, leaf length, leaf breadth and days to maturity contributed for maximum of 34 % variation in the PC1 with 4.45 eigen value. Divergence of second PC was mainly contributed by plant height, days to 50% flowering, leaf length, panicle length and days to maturity. Eigen value for PC2 was 2.63 and contributed about 20% of total variation. Contribution of divergence by PC3 was observed to be 10% with eigen value of 1.31. Traits responsible for variation in PC3 were leaf length, leaf breadth, days to maturity and hundred seed weight. Total variation contributed by PC4 was 9% and had 1.13 eigen value. Major variation contributing traits in PC4 were flag leaf length, flag leaf breadth and flag leaf area. Results of this study revealed the presence of variability for all the traits under study and identified accessions from clusters II and VIII as elite sources for trait specific improvement.