Estratificação de ambientes visando otimização da rede de ensaios de híbridos de milho para primeira e segunda safras

• Autor(a) principal: Lepre, André Luiz
• Data de Publicação: 2019
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Repositório Institucional da UFG
• Texto Completo: http://repositorio.bc.ufg.br/tede/handle/tede/10027
• topic: Zea mays; Estratificação de ambientes; Melhoramento de plantas; Milho safrinha; Norma de reação; Environment stratification; Plant breeding; Safrinha season; Reaction norm; Zea mays; CIENCIAS AGRARIAS::AGRONOMIA

The environmental stratification process consists of subdividing heterogeneous regions into more uniform sub-regions with similar ecological characteristics, so that the genotype by environment interactions (GE) it is predominantly of single nature. Environmental stratification procedures can be used to verify whether information generated at the testing locations used for breeding is complementary or redundant. This study was carried out with the objective of stratifying environments used to select corn hybrids in the final stage of breeding in the Central region of Brazil, to optimize the multi environment yield trials, in the first and second growing seasons. Yield data were obtained of 99 and 125 corn hybrids, tested across four years, in 55 and 49 locations, in the first (normal season) and in the second (“safrinha”) seasons, respectively. All yield trials were conducted in a randomized complete block design with two replicates. The yield data were submitted to individual and joint analyzes of the experiments, within each year. The components of variance were estimated and were tested through the Likelihood Test Ratio. The effect of GE interaction it was significant in all evaluated years, except for “safrinha” 2013. For the stratification process, two univariate approaches were used; intraclass correlation and prediction model based on reaction norm; and two multivariate approaches, GGE biplot and Factor Analysis. The most efficient method for eliminating redundant locations it was one chosen for the final stratification. Genetic correlation between pairs of location it was estimated by the univariate approaches. These correlations were used in cluster analysis, applying a SAHN algorithm (Sequential, Agglomerative, Hierarquic, Nonoverlapping clustering), associated with complete linkage method, which established the environmental stratification for the both planting seasons. In the multivariate approaches the environmental stratification were established according to the “who-won-where” approach and the factor loading established within each factor, for GGE Biplot and Factor Analysis, respectively. The verification of the established environmental stratification allowed to identify pairs of locations that were grouped in the same environmental strata, indifferent years of tests, which means redundancy. To optimize the target population of environments, the GGE Biplot method it was chosen which eliminated 12,5% of yield test locations in the both seasons.