Uso dos métodos AMMI e GGE biplot para análises de adaptabilidade e estabilidade de híbridos de milho e estratificação ambiental

• Autor(a) principal: Oliveira, Rogério Lunezzo de
• Data de Publicação: 2009
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFLA
• Texto Completo: http://repositorio.ufla.br/jspui/handle/1/4399
• topic: Zea mays; Interação genótipo-ambiente; Análise de componentes principais; Genotype-environment interaction; Principal components analysis; CNPQ_NÃO_INFORMADO

This work was carried out for evaluating stability, adaptability and environmental stratification, through the AMMI and GGE biplot methods by using maize hybrids trials data evaluation and for comparing the efficiency of these methods. Grain yield data originating from final trials of maize hybrids evaluation supplied by the Monsanto company were used and the study was carried out in the states of Minas Gerais, São Paulo, Paraná, Goiás, Mato Grosso do Sul, Bahia and Distrito Federal . The data was obtained from the evaluation of twenty-three maize hybrids in two agricultural harvests, 2005/2006 and 2006/2007, and eleven sites, which were: Barreiras-BA, Brasília-DF, Chapadão do Sul-MS, Iraí de Minas-MG, José Bonifácio-SP, Passos-MG, Presidente Olegário-MG, Rio Verde-GO, Rolândia-PR, Três Corações-MG, Uberaba-MG. The data were submitted, first, to the individual and joint variance analyses. Following that, to the adaptability, stability and environmental stratification analyses, based on the winning genotypes, through the AMMI and GGE biplot methods. Hybrid 15 was the highest yielding. However, by analyzing the AMMI2 graph, hybrid 10 was the one which presented the best combination between adaptability and stability, and the GGE biplot analysis led to the choice of hybrid 16 as being that one which presented the best combination between adaptability and stability. With the environmental stratification there was the formation of two groups of environments, in the AMMI methods (model AMMI1) and GGE biplot, and three groups in the AMMI2 model. The stratification permitted a reduction of 28% in the number of sites, considering the two methods. The GGE biplot captured a larger portion of the sum of the squares of genotype(G)+ interaction (GxA) when compared to the AMMI1model. Furthermore, the GGE graph was more efficient and of easier interpretation, when compared to the AMMI2 graph. From the obtained results, it is concluded that it is possible to reduce the number of evaluation sites. The grouping obtained by the GGE biplot method was similar to that of the AMMI1 model. The GGE biplot analysis was more efficient than the AMMI analysis.