The Bayesian approach in the evaluation of dissimilarity in sweet potato clones

Abstract

The availability of genetic variability in germplasm banks is fundamental for the success of plant breeding programs. The Bayesian inference combined with multivariate techniques allows to identify these sources of variability, assisting in decision making in plant breeding. Therefore, the objective was to evaluate the genetic dissimilarity among 24 sweet potato clones using qualitative and quantitative data, using the multivariate approach using Bayesian inference. In the morphological characterization, 24 morphological descriptors and data of productivity of roots and branches were used. For statistical analysis, multivariate analysis using the Bayesian inference was used. There was genetic variability among sweet potato clones. There is great dissimilarity between the UFVJM05, UFVJM09, UFVJM31, UFVJM37, UFVJM40 clones with the others. These divergent clones can be used in crosses in order to obtain progenies with high genetic variability. The multivariate approach using Bayesian inference was efficient in assessing dissimilarity.