Genetic gain by different selection criteria in soybean segregant populations

Abstract

In breeding programs, the complexity of economically outstanding characters, most of which highly influenced by the environment, makes selection difficult. Using genetic parameters, like heritability and gain with selection, it is possible to identify superior genotypes in early generations. Therefore, the aim of the present research was to compare different selection criteria by the estimated gains and the selected progenies, determining the superior and more similar methods. Augmented block design was used with three additional checks. Twelve hundred genotypes were evaluated. The most expressive expected gains were obtained with direct selection. However, the indices have shown to be more appropriate for selecting superior progenies because of their higher total gains, distributed among all the evaluated characters. The Sum of Ranks Index allowed the highest gains in most of the situations analysed.