Multivariate techniques for the determination of genetic diversity in sesame using RAPD markers

Abstract

The objective of this work was to compare different multivariate techniques in the characterization of 35 sesame genotypes using 769 RAPD makers. Genetic distances were obtained from the arithmetic complement of the Jaccard coefficient, and were evaluated by single linkage, complete linkage and unweighted arithmetic average (UPGMA) agglomerative methods, Tocher optimization method and principal coordinate analysis. The clustering structure was altered by the different methods. Adopting the same genetic distance (0.36) as value of cut, four groups were discriminated in single linkage, 13 in complete linkage and 11 in UPGMA; the Tocher's optimization methods formed four groups. Among the hierarchical clustering methods, UPGMA showed the best adjust for estimated and originals distances (CCC = 0.89). Principal coordinates analyses confirmed the low diversity among genotypes. The highest divergence occurred between Seridó 1 and Arawaca 4 cultivars, and the minor, between VCR-101 and GP-3314 genotypes. The three first principal coordinates responded for 35.13% of the variability, and 18 autovalues were necessary to explain 81% of the genetic variation. UPGMA, Tochers' optimization and the principal coordinates analyses are complementary in the clusters formation.