Knowledge of the genetic divergence in a set of genotypes is of great importance in breeding programs to avoid similar genetic combinations, with a consequent increase of heterotic expression in hybrids and genetic gains in segregating generations. We characterized morpho-agronomically 23 genotypes of sweet potato of the Embrapa Hortaliças' germplasm bank, to use these characteristics to assess the genetic variability among the genotypes by application of the methods Hierarchical Cluster Analysis (HCA) and Principal Component Analysis (PCA) and to estimate population parameters. The experiment was carried out at Fazenda Água Limpa, owned by Universidade de Brasilia ( Brasilia University, UnB), Brasilia, Brazil. Plants were cultivated in field conditions on a randomized blocks design with four replications. Eight characteristics of the roots were evaluated. The coefficients of genetic (CVg) and environmental (CVe) variation, the broad sense heritability (ha 2) and the ratio CVg/CVe indicated a favorable situation for the selection of the traits marketable yield, total yield, diameter and length of sweet potato roots. Most of the analyzed genotypes are similar with respect to the morpho-agronomic descriptors used. However, crosses between genetically dissimilar parents such as the CNPH 69 clone, which stood out by high marketable and total yield; the CNPH 80 clone, presenting good shape and moderate resistance to soil insects and the CNPH 60 clone, which showed roots with ideal diameter for commercialization, can be recommended for inclusion in intercross programs aiming to generate superior genotypes.
Ipomoea batatas; dissimilarity; Hierarchical Cluster Analysis; Principal Component Analysis; population parameters; breeding.
