Análise multiambientes visando a recomendação regionalizada de clones de cana-de-açúcar

• Ano de defesa: 2011
• Autor(a) principal: Leite, Mauro Sergio de Oliveira
• Orientador(a): Peternelli, Luiz Alexandre
• Banca de defesa: Cecon, Paulo Roberto , Carneiro, Antônio Policarpo Souza , Silva, Felipe Lopes da , Carneiro, Pedro Crescêncio Souza
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Doutorado em Genética e Melhoramento
• Departamento: Genética animal; Genética molecular e de microrganismos; Genética quantitativa; Genética vegetal; Me
• País: BR
• Palavras-chave em Português: Saccharum; Cana-de-açúcar
• Palavras-chave em Inglês: Saccharum; Sugarcane
• Área do conhecimento CNPq: CNPQ::CIENCIAS AGRARIAS::AGRONOMIA::FITOTECNIA::MELHORAMENTO VEGETAL
• Link de acesso: http://locus.ufv.br/handle/123456789/1338

The objectives of this study were to evaluate the regionalization of final competition trials of sugarcane clones using REML/BLUP in the context of mixed models, aiming at the regionalized recommendation, as well as to use the GGE biplot analysis approach on winner genotypes, aiming at determine a pattern of trials stratification. Data from 215 trials were analyzed in randomized complete block design with three replications, consisting of 75 different clones and 13 standard varieties, divided into five years of planting trials. The sites were previously grouped into five or six regions based on the characterization of the "Edaphoclimatic Environment" of each site. Variance components estimates and genotypic values predictions were obtained by the REML/BLUP procedure using the software SELEGEN-REML/BLUP and employing the model 53. Estimated mean to five harvests for tons of cane and pol per hectare (TCHe5c and TPHe5c) were evaluated. The GGE biplot analysis was performed considering the simplified model for two principal components, local centered, using the free software R. In the context of mixed models is possible to group different edaphoclimatic regions, since genotype x location interaction is high in all regions and showed to be lower when the regions are grouped two by two. In the joint analysis for all regions, seeking for broad adaptation recommendation, 13 genotypes with regional adaptation are discarded. There is a genetic gain 0.7% and 1.1% higher for TCHe5c and TPHe5c, respectively, when the genotype recommendation for regional adaptation is compared to the recommendation of genotypes broadly adapted to all regions. In the GGE biplot analisys two stable regions was observed, in other words, regions that showed coincidence of locations in at least three of the five years evaluated. Different edaphoclimatic regions were often grouped in the same region or strata obtained by GGE biplot analysis. Both studies agreed in the fact that some edaphoclimatic regions showed similar performance of genotypes, allowing the stratification of the experimental network into a smaller number of regions.