EVALUATION OF GENOTYPE VERSUS ENVIRONMENT INTERACTION IN SUGARCANE USING MIXED MODELS

• Autor(a) principal: Bastos, Irlane Toledo
• Data de Publicação: 2007
• Outros Autores: Barbosa, Márcio Henrique Pereira, Resende, Marcos Deon Vilela de, Peternelli, Luiz Alexandre, Silveira, Luís Cláudio Inácio da, Donda, Leandro Rodrigues, Fortunato, Alessandro Antônio, Costa, Paulo Mafra de Almeida, Figueiredo, Izabel Cristina Rodrigues de
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Pesquisa Agropecuária Tropical (Online)
• Texto Completo: https://revistas.ufg.br/pat/article/view/3077
• topic: Mixed models; REML/BLUP; stability; MHPRVG method; Modelos mistos; REML/BLUP; estabilidade; método MHPRVG

The objective of this study was to evaluate yield, adaptability, and stability of sugarcane genotypes using the harmonic mean of relative performance of genotypic values method (MHPRVG), and to compare it with the method of differences of the segmented straight line weighted by the residual coefficient of variation (DRRB-CV), based on the Pi statistic from Linn & Binns. Employing the augmented block design, with five blocks, seventy clones and two checks were evaluated. Two experiments were carried out in each of seven environments in Minas Gerais State. Evaluations were performed in plant and ratoon. Considering the selection for sugar yield, adaptability, and stability, three clones performed better: RB947520, RB957712, and P86-42. The MHPRVG method was highly correlated with the parameters of the DRRB-CV method, and with the genotypic values that took into account the genotype x environment mean interaction. All the correlations between these statistical measurements were greater than 0.90. However, the MHPRVG method has the advantage that the results appear on the same scale of the characters evaluated, which could be directly interpreted as genetic values. Another advantage for clone selection is that this method takes into consideration simultaneously the criteria yield, adaptability, and stability. KEY-WORDS: Mixed models, REML/BLUP, stability, MHPRVG method.