Resumo: Responsible for approximately 70% of the world?s coffee exports, Brazil is increasingly concerned about the quality of the coffees produced, given the growing demand for so-called specialty coffees. With this, the breeders need, besides the agronomic characteristics, to consider the physical and sensorial quality of the beans in the breeding programs. However, the greater the number of characteristics to be considered in the selection process, the higher the difficulty in selecting superior genotypes. In this context, multivariate analyzes can help to overcome this problem. In the light of the facts, the objective was to select Coffea arabica genotypes with a high simultaneous potential of variables of commercial interest, in three municipalities belonging to the Matas de Minas region?MG, Brazil, through factor analysis, using their scores as criteria or indices of selection for genotype identification. Multivariate analyzes were performed for each environment individually and, by commonality, three factors were established for each environment. The factors were interpreted as sensorial quality, sieve and vigor, in a similar way in the three environments. The interaction genotype by environment was maintained even after the summary of the variables in factorial complexes. The genotypes Catucaı´ Amarelo 24/137 and H419-3- 3-7-16-4-1 excelled in relation to the factorial complexes, besides showing good adaptability and stability, consequently, they present great potential to improve the coffee production performance in the region of Matas de Minas.

Ano de publicação: 2019

Tipo de publicação: Artigo de periódico

Unidade: Embrapa Café