Application of Multivariate Analysis to Access Selected Rice Germplasm Phenotypic Diversity

Abstract

Without variability, it is not possible to conduct a plant breeding program. Germplasm is hence the critical first step in initiating a breeding program. Diversity in O. glaberrima accessions is enormous. It needs to be organized and characterized in order to facilitate its use by plant breeders. Two principal components (PRIN1) and (PRIN2) accounted for most of the variability observed in characters studied. PRIN 1 accounted for 56% of the phenotypic and morphological variation. The PRIN 1 was loaded on plant height, number of panicle, biomass wet weight, panicle wet weight and grain yield traits. PRIN 2 accounted for 23% of the variation. PRIN 2 was loaded on biomass wet weight, biomass dry weight, panicle wet weight, panicle dry weight, and harvest index traits. The test for univariate statistics described individual variable to explore pattern of response to variation showed strong statistical significantly (P<0.001) on phenotypic differences in all the variables that were measured. These traits studied are the most important contributing to the overall variability. The dendogram produced grouping that defined nine distinct clusters and minimum genetic distance between clusters varies from 0 to 5. All the selected Oryza glaberrima accessions and the Oryza sativa were distributed across the nine clusters respectively.