Genetic Divergence, Principal Components and K-means Clustering Analyses of Some Agronomic Characteristics of Eleven Castor (Ricinus communis L.) Accessions

Morphological Characterization of 11 castor (Ricinus communis L. ) accessions was carried out in 2012 at Castor Research field of National Cereals Research Institute (NCRI), Nigeria. The accessions were evaluated for 24 morphological characters at different growth stages. The data generated were subjected to One-way ANOVA, principal components analysis and K-means non-hierarchical clustering analysis. The accessions differed significantly at p less than 0.01 for all characters evaluated, with exception of length of leaf petiole which only significant at p less than 0.05. The first 5 PCs revealed 93.9 % of the total variation, with PC1 and PC2 being responsible for 50.2 % and 22.4 % respectively. PC1 was mainly associated with number of nodes to 1st raceme, weight of 100 seeds and days to maturity. The contributory characters on PC2 axis were height at maturity, number of nodes on the main stem and length of leaf petiole. Specific grouping of the accessions in 5 clusters indicated that cluster 1, 2, 3, 4 and 5 contained 2, 1, 2, 4, 2 members respectively. The members (1, 9) of cluster 1 had the highest yield followed by the only member (10) of cluster 2 while the members of clusters 3 and 4 recorded low yield. This study revealed the distinctiveness of each accession, which could serve as basis for the crop improvement.