Repeatability estimation for mangaba selection using mixed models

The Mangaba (Hancornia speciosa Gomes) is a fruit tree rich in such vitamins, iron, calcium and protein, and plays an important social, cultural and economic role for farmers in the Cerrado and Caatinga biomes of Brazil. The aim of this study was to estimate genetic parameters and genetic gains in the agrotechnological attributes of a breeding population of the mangaba from Embrapa Meio Norte. The variables under analysis were: fruit weight (FW), fruit length (FL), fruit diameter (FD), skin weight (SkW), number of seeds per fruit (NSF), seed weight (SeW), total soluble solids (TSS), total titratable acidity (TTA), TSS/TTA ratio, pulp weight (PW) and percentage pulp (%P). The data were analysed using the REML/BLUP procedure of the Selegen software. High and intermediate estimates of the components of repeatability were found for the variables of the populationunder study. It can therefore be inferred that these variables can be used in genetic studies, with a view to an increase in the efficiency of the selection process and to a progressive advance in the breeding population of this important native species. As such, taking repeatability as the limit to be assumed for heritability, the values obtained can be considered satisfactory in terms of genetic progress, showing good reliability for the phenotypic value as an indicator of genotypic value and significant repeatability of the expression of the evaluated trait. Genetic gains were seen for genotypes 6 and 12, considering the silmultaneous selection of eight highlighted genotypes based on the most important attributes of the improvement program, and which can contribute to the productive sector.