Estratégias para seleção de progênies de sorgo sacarino com base em múltiplos caracteres

Abstract

Sweet sorghum is a bioenergy crop that presents stalks rich in sugar content, similar to sugar cane. Correlation studies have shown several agroindustrial traits commonly assessed in the genotype evaluation trials are highly correlated with the breeding target trait 1 st generationethanol yield, such as the production in tons of brix per hectare (TBH). In this way, the aim of this work was to evaluate the prediction accuracy using the univariate and multivariate mixed model approaches and compare the genetic gains by different strategies of progeny multi-trait selection. We selected 196 half-sib progenies from the zero-cycle base population of the intrapopulational recurrent selection program of Embrapa Maize and Sorghum aiming to increase sugar yield in sweet sorghum. The experiments were carried out in two environments in the 14 x 14 lattice experimental design. The traits measured were flowering time (FLOW), plant height (PH), green mass production (GMP), total soluble solids content (TSS) and tons of brix per hectare (TBH). We analyzed the data using a univariate and multivariate mixed model approaches, and four progeny selection strategies: direct selection based on TBH, FAI / BLUP index, Mulamba and Mock index, and additive index. Significant genetic variance was observed for all traits in each environment, and in the joint analysis for almost all, except TSS. The multivariate approach provided estimates of genetic parameters and predictions of the progeny genetic values more accurate and greater genetic gains for all traits than the univariate approach. The selection for TBH and the FAI/BLUP index resulted in balanced genetic gains estimates for the traits in both the univariate and multivariate approaches, allowing the identification of progenies that associate high agroindustrial performance.