Divergent gene pools in rice improvement.

The objective of the present study was to investigate the genetic architecture of yield in diverse populations of rice. Multivariate analysis by Mahalanobis's D(2) statistic and canonical (vector) analysis revealed that panicle weight, days to maturity, plant height and seed size were the important forces contributing towards divergence. One hundred rice strains were grouped into nine clusters with the help of D(2) and canonical analysis. The grouping pattern of the varieties were quite at random indicating that the geographical and genetic diversity were not related.