Variabilidade genética de progênies de feijão-comum do grupo preto do quarto ciclo de seleção recorrente para produtividade

Abstract

Grain yield remains the primary objective of plant breeding programs. Due to the quantitative and polygenic nature of this trait, recurrent selection methods are the most appropriate for achieving rapid and effective genetic gains, as they promote the continuous accumulation of favorable alleles. In common bean (Phaseolus vulgaris L.), successful cultivars consistently integrate high yield with commercial grain quality. The present study aimed to estimate genetic parameters and assess the genetic diversity of common bean progenies from the third cycle of recurrent selection for grain yield, within the black bean market class, to guide the composition of the subsequent selection and recombination cycle. To obtain experimental populations, 13 parental genotypes were utilized. In the initial generations, 500 progenies were evaluated at the S0:1, S0:2, and S0:3 stages, with 35 progenies selected for advancement to the S0:4 generation. The 35 S0:4 progenies, along with four control cultivars, were evaluated in field trials arranged in a randomized complete block design (RCBD) with three replications. The experimental plots consisted of three-row plots, each three meters in length, across five environments located in the states of Goiás and Paraná. The following agronomic traits were assessed: grain yield (GY) (kg ha⁻¹), 100-seed weight (100M) (g), sieve yield (SY) (%), lodging (LOD), and plant architecture (PAR). Additionally, disease resistance evaluations were conducted for Fusarium oxysporum (FOP) and anthracnose (AN). The genotypic characterization was performed using molecular data obtained from microsatellite (SSR) markers. Statistical analyses included analysis of variance, estimation of genetic parameters, and the calculation of genetic, phenotypic, and environmental correlations. The SSR marker data were used to estimate genetic divergence and population structure. For the selection of superior progenies, the Kennard-Stone algorithm and the Mulamba & Mock selection index were applied. The overall means for GY, 100M, and SY were 2846 kg ha⁻¹, 21.36 g, and 86.70%, respectively. Analysis of variance revealed a significant effect for genotypes, demonstrating genetic variability among progenies, as well as a significant genotype-by-environment interaction, indicating differential progeny responses across the tested environments. Heritability estimates were high for GY (67.97%) and 100M (90.67%) and intermediate for SY (74.29%). The expected genetic gain based on individual trait selection was 6.01% for GY, 3.47% for 100M, and 2.77% for SY. Molecular analysis revealed inconsistent clustering patterns and a lack of clear genetic structuring. Among the different selection strategies evaluated, the Mulamba & Mock index with economic weighting yielded the highest selection gains for GY (5.65%), 100M (1.29%), and SY (1.23%). Based on these results, ten superior progenies were selected for inclusion in the next recombination cycle. These progenies will be used to develop new elite lines and to generate the population for the subsequent cycle of recurrent selection.