Caracterização de acessos de Coffea arabica L. do banco ativo de germoplasma de Minas Gerais

Abstract

Germplasm banks are sources of genetic variability and, consequently, sources of desired agronomic traits in coffee genetic improvement programs. The aim of this work was to study the genetic variability and select accessions of arabica coffee from the Active Germplasm Bank of Minas Gerais at EPAMIG in Patrocínio - MG through agronomic characteristics. Twenty-five accessions of Coffea arabica L. were studied through morphoagronomic evaluations. In the second half of 2017, growth evaluations were carried out, with measurements of stem diameter, plant height and crown diameter. The productivity of accessions in the 2016/2017, 2017/2018, 2018/2019 and 2019/2020 harvests was evaluated. Immediately after harvesting, a sample of coffee fruits was separated to count the percentages of ripe, green, raisin/dry and floating fruits. With the productivity data, percentage of ripe, green and raisin/dry fruits and percentage of floating fruits, the variance components and the prediction of random effects were estimated using the REML/BLUP procedure. Productivity data were submitted to analysis of variance and Tukey test (p<0.05) was applied to compare means. With the vegetative characteristics, the genetic divergence between the accessions was evaluated by means of multivariate analyses, using the generalized Mahalanobis distance and, also, the accessions were grouped by the UPGMA hierarchical method. There was genetic variability among the accessions studied for the traits yield and percentage of fruits raisin/dried, indicating possibilities of obtaining genetic gains through selection in this group of accessions. There is potential for selection gain and consequent improvement in the means of the evaluated characters, especially in accessions MG0223 (Pacamara), MG0438 (Mundo Novo x S795 UFV 335-04), MG0194 (Caturra Amarelo Colombiano), MG0420 (Mundo Novo x S795 UFV 315 -04) and MG0036 (Yellow Bourbon). The hierarchical method UPGMA was efficient for the establishment of accession groups in the analysis of genetic divergence.