Obtenção de estimativa de herdabilidade em populações implantadas sem delineamento experimental

• Ano de defesa: 2016
• Autor(a) principal: Silva Júnior, Vitor Passos da
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Lavras; Programa de Pós-Graduação em Genética e Melhoramento de Plantas; UFLA; brasil; Departamento de Biologia
• Palavras-chave em Português: Herdabilidade; Genética quantitativa; Plantas - Melhoramento genético; Eucalipto; Genética de populações; Heritability; Quantitative genetics; Plant breeding; Eucalyptus; Population genetics; Genética Quantitativa
• Link de acesso: http://repositorio.ufla.br/jspui/handle/1/11134

Some populations were established in the past years without any experimental design in Forest breeding, which has challenged breeders to estimate genetic parameters. In this context, an estimation method for the heritability (h2) was proposed taking into account only the location of the trees in the field. Thus, this study aimed to compare the method of Sakai and Hatakeyama (1963) with the conventional method, based on analysis of variance, using for this, a Eucalyptus population established in experimental design. Also compare the efficiency of the method from simulated data with known h2. This study was carried out in a Eucalyptus breeding population composed of 49 clones. The trials were established in eight sites in the states of Espírito Santo, Bahia, São Paulo and Mato Grosso do Sul, Brazil, following a randomized block design with 30 replicates in seven sites and 40 in the other one, in single tree plots. Growth traits analyzed were circumference at breast height (CBH) and total height (H) at three years. Analysis of variance were performed for each site. Scenarios with heritabilities of 0.3, 0.4 and 0.8 and sample numbers of 10, 20 and 30 were simulated 100 times for each population. The method in study was consistent in cases with high heritabilities, however, in simulated data with low heritabilities, the model was not efficient. The results suggested that the model is highly dependent on the soil heterogeneity index (b) and the coefficient of determination (R²) is not good-enough criteria to identify the best value to be assigned to the constant b.