Eficiência do índice FAI-BLUP na seleção de genótipos Bourbon

Abstract

The present work aimed to select genotypes of Coffea arabica based on multi-characters by the FAI-BLUP selection index and compare it to the Smith-Hazel index. A total of 20 coffee genotypes were used, of which 17 belong to the Bourbon group and three are widely cultivated genotypes in the State of Minas Gerais, Mundo Novo IAC 502/9, Catuaí Vermelho IAC 144, Icatú Amarelo IAC 3282. The experiments were installed in the towns of Campos Altos, Santo Antônio do Amparo, Patrocínio, Lavras and Três Pontas located in the state of Minas Gerais. The experimental design was did in randomized complete blocks with three replications and ten plants per plot, installed in December 2005, by the space 3.5m x 0.8m. The experimental data were taken in the 2007/2008, 2008/2009 and 2009/2010 agricultural years. Agronomics, sensorial, chemical and physico -chemical characteristics were evaluated. The individual and joint statistical analyzes were performed for each characteristic. In the joint analysis, the residual variance and covariance matrix was modeled with different. The Schwarz criterion (BIC) was used to choose the best matrix. We estimated The E-BLUP (Empirical Best Linear Unbiased Prediction) of the genotypic values that were used in the FAI-BLUP selection, all analyzes were performed in the R software using the Asreml package. The use of the FAIBLUP selection index in the selection of multiple characters in Coffea arabica genotypes has been shown to be efficient and may be an alternative in Coffea arabica breed ing programs.