Abordagem bayesiana na seleção de clones e progênies de batata-doce

Abstract

The bayesian inference as a tool for the identification of superior sweet potato genotypes is an important strategy for the development of breeding programs, as it allows the use of prior knowledge from previous experiments. Sweet potatoes have the potential to combat malnutrition and ensure food security for the population. A versatile culture that can be used both in human and animal food, and as a raw material for industry. Therefore, the objective was to select clones and superior progenies and to study the variability existing in the germplasm bank of the Institute of Agricultural Sciences (ICA), of the Federal University of Minas Gerais (UFMG), using the bayesian approach. 24 clones and 16 progenies of sweet potato halfsiblings were evaluated in two experiments regarding root and branch productivity, root shape and resistance to soil insects. The data were analyzed using bayesian inference, with an informative priori, obtained from data from previous experiments. To assess the genetic dissimilarity of the 24 clones, quantitative and qualitative data were used in a multivariate analysis using the Bayesian approach. The clones UFVJM40, UFVJM06, UFVJM09 and CAMBRAIA and progenies F-UFVJM15, F-UFVJM56, FUFVJM09, F-CAMBRAIA and F-UFVJM40 stood out from the others and can be used as parents in future breeding programs. There was genetic variability between sweet potato genotypes. UFVJM05, UFVJM09, UFVJM31, UFVJM37, UFVJM40 are dissimilar to the others and can make crosses with the selected superior genotypes. Bayesian inference was efficient in assessing dissimilarity and in the selection of sweet potato clones and progenies.