Partitioning of Rice (Oryza sativa L.) Genotypes Based on Morphometric Diversity

Abstract

Aims: The Objectives of this study were to partition the rice genotypes into different clusters, identification of heterotic groups, and most important traits contributing to divergence to utilize them for specific objective-oriented breeding programs in future. Study Design: The experiment was set out in randomized complete block design with three replications. Place and Duration of Study: The experiment was carried out at experimental farm of Department of Genetics and Plant Breeding, Bangladesh Agricultural University, Bangladesh during Aus (Kharif) season of 2012. Methodology: The diversity among sixty rice genotypes and contribution of thirteen traits towards diversity were analyzed using Mahalanobis’s D2 statistics and principal component analysis. Results: Analysis of variance showed highly significant variation among the genotypes for all the traits. Cluster analysis based on D2 values exhibited seven distinct clusters. The highest intra-cluster distance (21.95) was observed in cluster II whereas that was lowest (7.62) for cluster VI. Maximum inter-cluster distance was observed between cluster III & VII (46.75) followed by cluster II & VII (42.91), cluster V & VII (38.48), and cluster III & VI (30.87). In all cases inter-cluster distance was higher than the intra-cluster distance suggesting wider diversity among the genotypes. All the short duration genotypes with high yield, high tiller number per hill and more filled grain per panicle were grouped in cluster VII whereas tall, long duration genotypes with low yield, wider flag leaf area, long panicle and more unfilled grain per panicle were grouped in cluster II. Cluster III composed of long duration & moderate yielded genotypes, but cluster V composed of genotypes with long duration and high yield. First three principal components explained about 81% of the total variation. Results of PCA suggested that traits such as number of filled grains per panicle, number of unfilled grains per panicle, flag leaf area, plant height and days to maturity were the principal discriminatory characteristics. Conclusion: The studied rice genotypes showed considerable divergence for most of the traits. These results can now be used by the breeders to develop rice varieties having desirable characteristics and new breeding strategies for rice improvement.