Interação genótipo x ambiente em testes de procedências e progênies de Tachigali vulgaris nos estados do Pará e Amapá

Abstract

Studies aimed at the genetic improvement of Tachigali vulgaris suggest good prospects for gains with the selection of the best genotypes. However, the selection of superior genetic materials is hampered by the genotype x environment (G x E) interaction. Studies that evaluate the G x E interaction are essential to assess the potential for improvement of species in different soil and climatic conditions. The objective of this work was: (1) to evaluate the existence of a G x E interaction for the characters mean annual increment (MAI), bifurcation (BIF) and survival (SOB) in three tests of provenance and progenies of T. vulgaris in northeastern Brazil; (2) to select the best T. vulgaris families for the MAI, using two methods, MHPRVG and GGE biplot and (3) to select the best individuals of T. vulgaris for the MAI. The tests were installed in 2011 and 2012 in a randomized block design, with four blocks and six plants per plot. At the age of ten years (test 1) and nine years (tests 2 and 3) all plants were evaluated for MAI, BIF and SOB. For the analyses, three groupings of locations were made. The first analysis involved all the 3 sites and 17 families were used. The second analysis involves tests 1 and 2, and 52 families were used. The third analysis involved tests 2 and 3, for which 24 families were used. The analysis of the G x E interaction for the MAI, BIF and SOB variables was carried out using model 51 of the SELEGEN-REML/BLUP 1.0 software and the analysis of the GGE biplot was carried out with the aid of the R software. The selection gain was estimated based on the ten best families and on a number of individuals that guaranteed a minimum effect size of 30. For the selection of the best families, the predicted genotypic values of MAI for each family were multiplied by the predicted genotypic values of SOB in order to correct for the increased growth with increasing space as a result of mortality. The results indicated a simple interaction for MAI and complex for BIF and SOB, considering all environments together. The analysis of the pairs of environments indicated good perspectives of gains with the selection of the best families, with a simple G x E interaction for BIF and SOB in the pair of environments 1-2 and for the MAI in the pair of environments 2-3, suggesting the formation of improvement zones. It was possible to obtain a minimum effective population size of 30 for the individual selection, considering the individual environments, in pairs and altogether, with greater prospects of gains in the pairs of environments (11.05% for the pair 1-2, selecting 120 individuals and 28.86% for the pair 2-3, selecting 84 individuals). The ten best families with survival correction, according to the MHPRVG analysis, were: G12, G15, G14, G9, G24, G7, G44, G45, G10 and G23. The GGE biplot analyzes identified the G45 family with the best performance for the different environments and environment 2 as the most discriminating and representative. We conclude that there are families and individuals with high productive potential and that they present good stability and adaptability to different locations in the states of Pará and Amapá in northeastern Brazil.