Divergência genética entre acessos de Capsicum chinense Jacq. coletados no sudoeste goiano

Abstract

Collecting and characterization of Capsicum chinense Jacq accesses are very important to recover and preserve genetic resources of this species; however, it is necessary to quantify the existing divergence for its use in breeding programs. The objective of this work was to verify the genetic divergence of 137 accesses of Capsicum chinense Jacq collected in southwest Goiás through 25 morphological descriptors suggested by IPGRI, and to quantify them using the colorimetric method in absorbance/gram of the capseine present in fruits. The experiment was carried out in a green house, at the University of Goiás Jataí campus, arranged in a completely randomized bloc design, with four replicates of one plant in an eight kilogram soil pot. Data was subjected to the analysis of variance and the 100 seed weight was significant at 5% of probability, and the other traits significant at 1% probability. The same occurred with the means grouping by Scott-Knott at 5% probability, with the formation of one more group for all descriptors: 10 fruit mass, fruit length average, and fruit pedicel length; which presented a larger number of groups. In the multivariate statistic the optimizing grouping of Tocher was performed for all 137 accesses, with the inclusion of 95% of the accesses in the first group. Regrouping of the group I accesses was chosen, indicating the formation of 19 new groups, showing a large intragroup heterogeneity. 16 groups were formed by the hierarchical UPGMA method, with a cut at approximately 60% from the distance. The first three main components were able to explain 43.68% of the total variation; however, even with low explanation it was possible to observe agreement between the grouping techniques used. ACP was complemented with the projection of CP1, CP2 and CP4. There were no drastic changes, enabling a 7.85% increase in the explanation. The variables with the largest contribution, by the Singh method were plant height and plant canopy diameter; both related to plant size. Based on what was proposed by Julliffe, 14 descriptors were likely to be discarded. The analyses performed contributed for the generation of information regarding the variability present in the accesses tested with subsidies for further studies of the species assessed.