Abstract: Thirty accessions of cowpea from Ebonyi, Enugu and Kogi States, Nigeria were used for the study. Seeds of the accessions were randomized in a plot measuring 50×50 meters for three growing seasons. Our results on yield and yield-related traits showed that three principal components were extracted, which contributed 82.23% of the total variability. It revealed that number of seeds pod-1, 100-seed weight, pod length, days to 50% maturity, seed yield, number of leaves plant-1 contributed significantly to the total genetic variability while for proximate composition, four principal components accounted for 93.75% of the total genetic divergence. Cluster analysis revealed that accessions were grouped not necessarily based on geographical location but genetics. Selection for high yielding accessions should be done on cluster 2 as we recommend selection and hybridization of accessions from cluster 1, 2, and 3 for optimal benefit.
Keywords: Germplasm collection, cowpea, selection, hybridization, multivariate analysis.