Estratificação ambiental pela análise da interação genótipo x ambiente em milho

Abstract

Resumo -O objetivo deste trabalho foi realizar a estratificação de ambientes de cultivo de milho nos Estados do Paraná, Minas Gerais e Bahia, por diferentes métodos, bem como determinar o grau de associação e divergência entre os métodos de estratificação. Foram avaliados quatro métodos: o tradicional de Lin; o de dissimilaridade ambiental; o de decomposição da interação genótipo x ambiente (GxA) em partes simples e complexa; e o de análise de fatores. Esses métodos foram aplicados a 48 híbridos experimentais de milho, avaliados em 11 ambientes de cultivo nos três Estados, divididos em dois conjuntos de experimentos. Além dos híbridos, foram avaliadas seis cultivares comerciais, utilizadas como testemunhas, comuns aos dois conjuntos. Verificou-se a predominância de interação GxA do tipo complexa. A decomposição da interação GxA em partes simples e complexa e a análise de fatores são métodos fortemente associados entre si, mas moderadamente associados aos demais. Além disso, esses métodos são mais rigorosos no processo de estratificação ambiental e ponderam de maneira mais eficiente a magnitude da interação GxA.Termos para indexação: Zea mays, caracterização ambiental, métodos de estratificação. Environmental stratification by analysis of genotype x environment interaction in maizeAbstract -The objective of this work was to perform the stratification of maize cropping environments in Paraná, Minas Gerais, and Bahia States, Brazil, using different methods, and to determine association and divergence levels between the stratification methods. Four methods were evaluated: the Lin's traditional one; the one of environmental dissimilarity; the genotype x environment (GxE) interaction decomposition in simple and complex parts; and the factor analysis. These methods were applied to 48 maize experimental hybrids, evaluated in 11 cropping environments in the three Brazilian states, divided into two sets of experiments. Besides the hybrids, six commercial cultivars, common to both sets, were used as controls. Genotype x environment interaction was predominantly complex. The decomposition of GxE interaction in simple and complex parts and the stratification by factor analysis were strongly associated to each other, and moderately associated to the other methods. Furthermore, these methods are more rigorous in the stratification process, weighting more efficiently the magnitude of the GxE interaction.

Factor analysis applied to the G+GE matrix via REML/BLUP for multi-environment data

“…Garbuglio et al (2007) broke down the interaction in simple and complex parts for all the pairs of environments and observed that the environments grouped in the same factor contained a large part of the interaction as simple. These results were also observed by Mendonça et al (2007) and Ribeiro and Almeida (2011). These results confirm the efficiency of the groupings obtained via factor analyses after rotation.…”

“…This result may be observed in Table 3, where analysis with two factors grouped 4 environments on the first axis and 3 on the second axis and, in the other situations, there was no grouping but rather a scattering of the environments among the axes. Fritsche-Neto et al (2010b) and Ribeiro and Almeida (2011) obtained similar results, in which environments were spread out by the factors and, as a result, no grouping was obtained.…”

“…In the study of Murakami and Cruz (2004), the first two axes accounted for 81%, 67.3%, and 60.31% of the total variation accounted for in their three crop seasons. Garbuglio et al (2007) obtained 65.8% of the total variation accounted for in the first two axes, and Ribeiro and Almeida (2011) observed that the first two axes contained 64.73% of the total variation. Brian (2005) reports that in situations with few factors used, these few factors will have many high loadings, whereas in situations with many factors, its may be fragmented, hindering interpretation of a convincing nature.…”

“…Various authors have made use of this method to perform environmental stratification. Ribeiro and Almeida (2011), comparing environmental stratification by different methods, have concluded that factor analysis is more realistic and informative and that it more efficiently considers the magnitude of the GxE interaction of genotypes in maize. Garbuglio et al (2007), like Murakami and Cruz (2004), observed the efficiency of environmental stratification by factor analysis in studying adaptability on maize.…”

Components of variance and inter-relation of important traits for maize (Zea mays) breeding

“…The maize breeding encouters interference of G × E interaction, such as the difficulty of finding the bests genotypes, phenotypic stability and adaptability, the nature of the interaction of simple or complex and a differential answer of the characters that determinate the potencial productive of the culture (Ribeiro and Almeida, 2011). Conversely, some researches have pointed out that the grains mass, the ear dimensions, grains number per plant (Lopes et al, 2007), the plant stature , the prolificity (De Carvalho et al, 2001), the number of grains per ear rows (Balbinot et al, 2005) are determinant to increase the grains yield in maize.…”