Aplicabilidade do software SELEGEN para avaliação de desempenho produtivo, adaptabilidade e estabilidade de genótipos de arroz irrigado

Abstract

In a breeding program of hybrid rice data from VCU trials are usually analyzed by analysis of variance (ANOVA), which do not allow unbalanced data however the methodology BLUP (Best Linear undebiased predictor) can analyze unbalanced data and indicate beyond genetic constitutions best and more stable performance. This study aims to evaluate the applicability of the BLUP methodology through Software SELEGEN REML/BLUP in a program of hybrid rice. The study was conducted in 2009/2010, 2010/2011 and 2011/2012 seasons at four distinct producing regions of rice in Rio Grande do Sul state, in the SOUTH region, NORTH coast, WEST Border and Capão do Leão District. Data from VCU trials were analyzed including all locations and the three years through ANOVA and through SELEGEN and the differences in the selections of hybrids for grain yield was interpreted. The experimental design was randomized blocks with three replications per location each year and region. The results show that SELEGEN allows selection of hybrids with high yield potential and yield stability among years and locations, and provide an easy interpretation way to unbalanced data. The use of statistical tools that enable the analysis of large amounts of data is extremely important in a hybrid rice breeding program and helps the breeder to seek genetic groups of better combining ability objectively and with focus on the constitutions of high and stable productive potential. In addition, over the years and cycles of selection the data generated can contribute to the prediction of performance saving cost and time in developing superior rice hybrids.