Predição genômica do desempenho de híbridos de milho considerando a interação genótipos por ambientes

• Ano de defesa: 2017
• Autor(a) principal: Pires, Luiz Paulo Miranda
• 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: AMMI-Bayesiano; Milho - Melhoramento genético; Parentesco genético; Ensaios multiambientes; AMMI-Bayesian; Corn - Genetic improvement; Genetic kinship; Multi-environment trials; Fitotecnia
• Link de acesso: http://repositorio.ufla.br/jspui/handle/1/29463

In plant breeding, especially in maize, the genotype interaction effect by environment is present in all stages of hybrid evaluation, from the initial to the pre-commercial stages. This is an important difficulty in the cultivar recommendation in multiple environments and also in the hybr ids prediction. It means one of the main challenges for the professionals involved in the process of obtaining new cultivars. In maize breeding programs, genotypes from various backgrounds are naturally incorporated into each cycle, while so many others ar e discarded, making further accurate predictions is more difficult as a result of increased commercial germplasm complexity. In addition to substantial losses related to heritability decline, just because the phenotypic variance can increase by the interaction component. This circunstantially removes the phenotypic value of the actual genetic value of the hybrids under evaluate. At this scenario, several studies have been carried out in order to increase the contributions of genomic prediction in the genotypes interaction context by environments. Effectively predicting hybrids in the interaction context can approximate what are produced by the research with the needs of commercial breeding programs, since the costs related to genotyping of materials have declined in recent years, while costs related to phenotyping are increasing and limiting for increase breeding programs. In addition, several factors promote the umbalance in multienvironment trials, not only related to loss of plots, but also loss of whole environments of evaluation by climatic factors, insufficient amounts of seeds from some hybrids, so that they are tested in all environments, or even inclusion and exclusion derived from breeders criteria during the hybrids evaluation. Thus, new model proposals that are efficient in conditions of umbalance and applications of prediction processes with molecular markers information can contribute in the reduction of financial expenses and identifications of superior genotypes in maize breeding programs. With this study, we will try to create information about the efficiency of the genomic kinship information for hybrids prediction using GBLUP-AMMI model with Bayesian approach, besides studying the genotypes interaction with environments and realization of inference about the contribution of genotypes to the interaction.