Seleção de progênies S2 de milho com abordagem de modelos mistos

• Ano de defesa: 2015
• Autor(a) principal: Cancellier, Leandro Lopes
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Lavras; Programa de Pós-Graduação em Agronomia/Fitotecnia; UFLA; brasil; Departamento de Agricultura
• Palavras-chave em Português: Zea mays; Capacidade combinatorial; Modelos mistos; Desenvolvimento de linhagens; Combining ability; Mix models; Inbred line development; Melhoramento Vegetal
• Link de acesso: http://repositorio.ufla.br/jspui/handle/1/10579

Corn is broadly cultivated from north to south of Brazil. Many hybrids are available to farmers that present high yield potential. Therefore, one of the greatest challenges corn breeders must overcome is to replace the hybrids currently used. One of the main steps in breeding programs is the development of inbred lines, which will later be used for hybrid development. In inbred line development, early selection in the first generations of self-pollination is common. The success of early selection is mainly due to good progeny stability in self-pollination generations, and high correlation among S2 and S6 progeny, tested as topcross hybrids. The adoption of mix models for data analysis is an important tool to improve genotype screening, allowing great flexibility for unbalanced data analysis and offering more accurate genotype values. Therefore, the objective of this study was to select S2 corn progenies, evaluated as topcross hybrids, using mix model methodology. We used 500 S2 progenies obtained from three population, grown and submitted to a screening level of 40%. The resultant progenies were crossed with three testers. The hybrids and controls were tested to evaluate grain yield in five trials, three in Minas Gerais, one in Santa Catarina and one in Paraná, Brazil. The statistical analysis was performed using a mix model approach. We used the REML method to compute the variance component, and BLUP to predict the average. We also predicted BLUPs for General Combining Ability and Specific Combining Ability, as well as Spearman correlation coefficients among the BLUPs. For the overall progenies under study, the dominance effects had more influence over the expression of grain yield, shown by the wider range of SCA values. There was a coincidence of 86% on the selection strategy when made within the three populations regarding selection by the general value of GCA. Among all 444 hybrids, we considered the 133 with the highest SCA values. Considering superior hybrids, the progenies from population C exceeded the amount of expected hybrids by 24.6%, while the reduction was of 30.8% and 20% for A and B. The hybrids with higher BLUP values were crosses between progenies of population C with tester LE84. Regardless of the tester, using BLUP average or SCA, the hybrids ranking will be little changed due to high correlation. The low correlation among testers for SCA and average BLUP indicates the existence of progeny x testers interaction.