Efeito competicional em testes de progênies de eucalipto

Many studies have recommended the use of small plots for forest experiments, although they do not consider the inter-genotype competition increase. If this competition is not isolated from the mathematics model, it can lead to incorrect selection of genetic materials. The aim of this work was to evaluate the effect of seven competition covariates in two Eucalyptus spp. progeny tests. Data from the two half-sib eucalyptus progenies were analyzed, using the randomized blocks design. The seven analyzed covariates were HegyI's competition index (IC), self-competition (AT), alo competition (AL), self-competition mean (MAT), alo competition mean (MAL), and arithmetic means of four (M4) and eight (M8) nearest neighbors. Individual and combined analyses of covariates were used for the wood volume trait. All the variance components and the changes caused by covariates use were evaluated. The competition affects the results of eucalypt progeny analysis in different ways, according to its type, self or alo competition. Most influential covariates were MAT, MAL and IC. Most promising results of competition effects reduction were observed for the IC/MAT covariates inclusion in eucalypt progeny tests.