Parâmetros genéticos, métodos de condução de população segregante e estratégias de seleção de genótipos de soja

Abstract

The study of inheritance for agronomic characters and the selection of more effective methods are fundamental in breeding programs for the development of soybean progenies. This thesis was subdivided into six chapters, the first of which brings the theoretical reference related to culture, economic importance and soybean breeding. The second chapter aimed to obtain estimates of the general and specific combining abilities of five soybean parents through the partial diallel, in order to identify the most promising genotypes for agronomic traits and the best combinations for the development of superior strains. Six hybrid, non-reciprocal and five parent combinations (BRSGO Chapadões, TMG 4185, TMG 4182, BR/MG 46 Conquista and UFUS 7910) were analyzed using the partial diallel proposed by Geraldi and Miranda Filho (1988). Through diallel analysis, it can be concluded that the BRSGO Chapadões and UFUS 7910 parents were considered genetically superior and indicated to participate in crosses. The BRSGO Chapadões x UFUS 7910 cross was considered the best combination. The third chapter aimed to analyze the inheritance of agronomic traits in four soybean segregating populations. F2 segregating populations and their respective parents were analyzed from generations under field conditions. The evaluated agronomic characteristics were: number of days for flowering, height of the plant at flowering, number of nodes at flowering, number of days at maturity, plant height at maturity, number of total nodes at maturity, number of productive nodes, total pods and grain yield per plant. The estimated heritability and the minimum number of genes evidenced the quantitative nature of the agronomic traits, mainly for the number of total pods and for the production of grains. In addition, it was verified the presence of transgressive individuals for all characters evaluated, making possible the selection in sowing conditions outside the conventional period. The fourth chapter aimed to evaluate the agronomic performance of soybean progenies developed using the modified genealogical method. In this study, 19 F3:5 progenies, developed using the modified pedigree method, and four controls were evaluated in a randomized complete block design with three replicates. According to the results, the modified pedigree method allowed the development of progenies with high productive potential, with grain productivity above 4000 kg ha-1. The progenies UFUSL3, UFUSL9, UFUSL12, UFUSL13, UFUSL16, UFUSL17, UFUSL18 and UFUSL19 were considered superior to the other genotypes and indicated for productive performance assays. In the fifth chapter, the objective was to compare the modified bulk conduction methods, descendants of a single pod and pedigree in the development of soybean progenies for grain yield. In this work, 45 soybeanF4:5 progenies were evaluated in a hierarchical design, with three replicates. Genetic progress was calculated aiming to identify the best method of conducting segregant populations for the grain yield character. The development of soybean progenies for productivity was influenced by the adopted method of conduction, presenting the capacity to produce genotypes with different productive potentials. The bulk and descendant methods of a single pod presented the same potential in the development of soybean strains compared to the pedigree, because they present a greater number of productive lines in relation to the parental ones. The sixth chapter was made in order to identify the best strategy for the selection of soybean genotypes. A total of 33 F4:5 progenies were evaluated in a randomized complete block design with three replicates. The selection and prediction of gains were made through direct selection and through the following selection indices: sum of ranks, multiplicative and distance of the genotype-ideotype. In this study, direct selection for productivity and genotype-ideotype distance selection index presented the highest selection gains for grain productivity.