A large collection of maize germplasm is introduced annually to Lesotho from CYMMIT in Zimbabwe for evaluation of adaptability and yield performance. This collection is not characterized for degree of similarities and dissimilarities using morphological and other markers. The study was conducted with the objectives of (a) estimating genetic distance among maize cultivars using cluster analysis and (b) identifying morphological characters with high discriminatory power to segregate maize cultivars. The study was conducted at National University of Lesotho, Experimental farm. Randomized Complete Block Design was applied with ten treatments and three replications. Data collected using Descriptor compiled by International Board of Plant Genetic Resource Unit included number of leaves per plant, tassel colour, number of cobs, silk colour, stem colour, plant height, number of ears, ear length, cob diameter, number of kernels, kernel arrangement, kernel colour, shape of upper surface, kernel type, leaf length and tassel length. Data were subjected to GENSTAT software package to generate cluster analysis and perform principal component analysis. The results of cluster analysis revealed two big groups, of which one consisted of six cultivars and another consisted of four cultivars. Besides, there was one outlier. Two big groups were further divided into sub-groups, Three principal component analyses were used to analyze the results, which constituted 65.37% of the total variation. The first one showed variation of 26.93%, the second one showed 20.65% while the third one had 17.79%. The first principal component was constituted by ear length, tillering, maize height, total number of kernels, cross-section of cob and stem colour. The characters comprising second principal component were kernel type, kernel row arrangement, silk colour, number of ear and number of kernel row. Lastly, the character influencing separations along third principal component were number of kernel row, silk colour and number of leaves. The study was able to distinguish the cultivars.
Keywords: Cluster analysis, Lesotho, Maize, Principal Component Analysis
