Estratégias de seleção de linhagens de feijoeiro para multicaracteres

Abstract

Common bean (Phaseolus vulgaris L.) is a crop of great importance in the world agricultural scenario. However, the carioca bean represents the most consumed and produced commercial type of common bean in Brazil. Due to the greater preference for carioca bean, breeding programs for this crop have been dedicated to obtaining cultivars of this commercial group with adaptation to the different environmental conditions observed in the States. Recurrent selection has been a key strategy that has been used to achieve this goal. The Federal University of Lavras (UFLA) conducts two recurrent selection programs for carioca common beans, one targeting grain yield and the other resistance to angular leaf spot. Factors such as quantitative inheritance of economically important traits and genotype-by-environment interaction are constant challenges encountered by breeders in identifying superior lines. The success of recurrent selection programs is associated with the proper evaluation of the lines obtained throughout this process. In this final phase of the program, intermediate trials are carried out in a larger number of locations, seasons, and years in order to identify those lines with the potential to be included in trials of value for cultivation and use (VCU). At this stage, traits such as grain yield, plant architecture, and resistance to pathogens are evaluated. Another challenge faced by crop breeding programs is the selection of these lines for mult-traits in multi-environments. Thus, the objective of this study was divided into two chapters: (1) select bean lines from UFLA's recurrent selection programs aiming at grain yield and resistance to angular leaf spot to compose VCU trial; (2) to verify the applicability of lines selection strategies for mult-traits in multi-environments based on factor analysis and genotype-to-ideotype distance index. In these studies, 35 bean lines from the two recurrent selection programs for carioca beans at UFLA were evaluated. The traits evaluated were grain yield (PG), angular leaf spot severity (MA), and plant architecture (ArQ). The experiments were conducted in Minas Gerais, between 2017 and 2019. In the first chapter, the existence between lines and environments interaction of a simple nature for MA and ArQ and of a complex nature for PG was verified. Based on the analysis of adaptability for PG and the simultaneous selection of traits, seven lines with wide adaptability for PG and with the potential to compose VCU trials of carioca bean in the State of Minas Gerais were identified. In chapter 2, the result of the selection of the 20% best lines was compared in terms of the multi-traits evaluated in multi-environments through the use of methodology factor analysis, genotype-ideotype index, and adaptability according to Annichiarico. The analysis of individual factors has allowed the maximization of the selection of lines for groups of evaluations with a similar pattern of response of the lines regarding specific traits. The use of the genotype-ideotype index associated with the analysis of factors presented a balance in the selection of lines both in conditions of the same trait in multi-environments and of multi-traits in multi-environments.