The estimation of genetic variability in germplasm collections is important not only for the conservation of genetic resources, but also for plant breeding purposes. Accessions in a germplasm bank are studied based on quantitative and qualitative descriptors. However, these data are rarely considered simultaneously in joint analyses. This work aimed to study the genetic diversity among 56 Capsicum chinense accessions coming from the Germplasm Collection of the Universidade Estadual do Norte Fluminense Darcy Ribeiro (UENF). We used 44 morpho-agronomic descriptors, 37 qualitative and seven quantitative, and the Grower's algorithm for the joint analysis. We used a completely randomized design with three replications and three-plant plots. Plants grew in 5 L pots. There was phenotypic variability among the chili accessions studied, mainly in fruits. Marked differences were observed in size, shape, color and total soluble solids and vitamin C contents of fruits. We used UPGMA to perform the clustering, since it was the method with the higher cophenetic correlation coefficient (r = 0.82). Accessions felt into six classes. Gower's algorithm was more efficient in clustering when qualitative instead of quantitative data were considered. It indicates that qualitative data played a crucial role in explaining the observed groupings. Joint analysis of quantitative and qualitative data resulted in greater efficiency in the determination of genetic divergence among the accessions evaluated. Therefore, such analysis is definitely a viable and important tool for understanding the variability within germplasm banks.
Capsicum chinense; multivariate analysis; germplasm; plant genetic resources
