Yield Trait Association Studies through Principal Component Analysis in Bi-Parental Population Derived from indica/tropical japonica Inter Sub-Specific Cross in Rice (Oryza sativa L.)

In the present investigation, a mapping population consisting of 189 lines derived from an inter sub-specific cross between an indica cultivar, ‘Samba Mahsuri’ and a tropical japonica accession, ‘Cuba 65’ was investigated to assess the trait contribution to variability employing principal component analysis (PCA). The study aimed at determination of inter-relationships among various yield attributing traits vis-à-vis individuals of the mapping population. Out of eleven principal components (PC), first four PCs exhibited Eigen value more than one accounting to a total of 68.73% cumulative variance among the characters. The first PC accounted for the highest variability (PC1 31.6%) to the total variability, followed by 14.51% (PC2), 12.69% (PC3) and 9.91% (PC4). The highest positive Eigen value observed for the tiller number (0.33) and panicle number (0.32) in PC1 indicated their pronounced effect in the overall variation in the population. The analysis revealed the traits that contribute maximum for the variation. PCA also revealed negligible effect of days to fifty percent flowering, grain size and spikelet fertility to the variability in the mapping population. Although all the studied traits except grain size showed significant positive association with grain yield; negative correlation was observed between some of the component traits. Panicle weight, thousand grain weight and grain number showed negative association with tiller number and panicle number. The latter two traits contributed to the maximum variability in the population. The trait associations identified in the present investigation will help in determining the selection criteria to be adopted in the development of high yielding cultivars involving indica/tropical japonica populations.