Multivariate techniques were used to evaluated the genetic divergence among 56 accessions of 'chili' and sweet pepper from the germplasm collection of Universidade Estadual do Norte Fluminense. Eleven quantitative descriptors proposed by International Plant Genetic Resources Institute were utilized in a field experiment carried out in Campos dos Goytacazes, Rio de Janeiro State, Brazil, in a randomized block design with three replications and ten plants per plot. Generalized Mahalanobis distance (D²) was used as dissimilarity measure. Canonical variated analysis, cluster analysis using Tocher's optimization method and distances in the plan were applied. The variables: fruit length, fruit diameter, number of seeds per fruit, fruit average weight, plant height, plant canopy width, 1000seed weight, days to flowering, days to fruiting, fruit number per plant and fruit weight per plant were evaluated. There was significant difference among accessions for all descriptors evaluated. General agreement among all multivariate techniques used in the work was observed and it was possible to separate the accessions in eight distinct groups, indicating that there is genetic variability for the evaluated traits. The highest generalized distance of Mahalanobis was 266.42. Accessions have the potential to be used as parents in artificial crosses to obtain progenies with higher heterosis. Through Canonical variable analysis we observed that crosses with the greatest heterotic potential had been 56x43; 34x08, and 59x41.
Capsicum spp.; germplasm; canonical variables; Tocher's method; distance in a plan; genetic resources
