Estratégia fuzzy aplicada a seleção de linhagens de feijoeiro

Abstract

Common beans from the black commercial group are the second most produced in Brazil. It is consumed mainly in the southern region of the country and the states of Rio de Janeiro and Espírito Santo. Due to its national importance, bean breeding programs have been dedicated to developing highly productive lines, resistant to the main pathogens that affect the crop, plants with upright architecture, and, especially, with grains of high commercial quality. Providing these lines is not a simple task, since these characters are quantitative. To achieve success, these genotypes must be extensively evaluated for the main agronomic characters in experiments under different environmental conditions. Multi-trait and multi-environment selection are challenging due to the negative correlation between traits and genotype-by-environment interaction. Innovative methodologies such as fuzzy logic, which is based on the knowledge of the specialists to develop decision-making systems, can assist in the selection of lines in breeding programs. Therefore, the objective of this work was to employ a fuzzy controller to assist in the selection of black bean lines with high productive potential and commercial grain quality. A total of 162 black bean lines and seven controls were evaluated for the productivity and commercial aspect of the grains in two experiments scheduled in the 2022 rainy and dry 2023 seasons. Individual and joint analyses of variance/deviance were carried out for both traits. A Fuzzy controller was also developed to assist in selecting strains regarding grain productivity and appearance. This controller is currently simultaneous selection under genotype by environment interaction, which was observed for both traits. Using this methodology, 16 lines of black beans were selected with high productivity potential and commercial grain quality.