The Multivariate Analysis of Morphologic Traits, Grain Yield and Yield Componets of Native and Improved Rice Varieties

Studies of genetic diversity of plant populations are very important tools for plant breeders. In this study, 110 rice genotypes, including 39 Iranian local, 20 Iranian improved and 51 foreign improved varieties were studied for 14 agronomic traits in the the rectangular lattice design at the Rice Research Institute of Iran (Rasht). The analysis of variance showed that there were significant differences between cultivars for all the studied traits that indicates high genetic diversity among cultivars. Correlation analysis showed that grain yield was highly correlated with tillers per plant (rp =0.669 and rg =0.711), single panicle yield trait (rp =0.544) and number of filled grains per panicle (rp=0.454 rg=0.455), respectively. The results of factor analysis showed that seven main factors including panicle type, plant yield, plant height, plant shape, leaf shape, awn and seed shape, explain 77.68 percents of the total variance. Also the KMO coefficient was 0.64 that showed the relative preference of using of analysis factor. Cluster analysis using ward's minimum variance method, classified cultivars and characters in three and six groups, respectively. The accuracy of the cluster analysis was measured with the discriminant function and the respected amount was about 95 and 91 percent for the grouped varieties and characters. Results of factor analysis and cluster analysis confirmed each other.