Multivariate analysis of diversity in land varieties of Assam over varying nitrogen levels in rice

We evaluated twenty-six rice genotypes of Assam, including landraces and modern varieties, for ten quantitative traits in three different nitrogen regimes under shallow lowland conditions during Kharif 2014 and 2015 following randomized complete block designs and assessed the variability using multivariate methods. The principal component analysis (PCA) revealed that the first four principal components (PCs) explained about 78.7% of the total variation among the twenty-six rice genotypes. The two-dimensional projections for PC1 and PC2 suggested that the traits, viz ., 1000-grain weight, panicle length, stem dry weight, and effective tillers were the principal discriminatory traits. Mahalanobis D 2 statistics grouped the genotypes into three clusters, and 1000-grain weight contributed the most towards divergence (26.8%), followed by stem starch content (15.6%) and panicle length (13.5%). The promising genotypes of clusters I and II might be used as potential donors in the hybridization program to exploit heterosis and to isolate desirable recombinants from the advanced generations for the lowland rice ecosystem. From our analysis, it was inferred that, the genotypes, namely Mal Bhog, Kabori, Jahinga and Lothabor with higher 1000-grain weight, spikelet fertility, stem starch content, and panicle length, respectively would serve as potential donors for increasing grain yield in rice.