Implicações da interação clones por cortes na seleção de Genótipos de cana-de-açúcar de um programa regionalizado

Abstract

The existing genotype by environment interaction in sugarcane makes difficult the identification of superior genotypes in the final stages of breeding programs. Besides, multivariate methods are widely used due to their high capacity to capture a greater proportion of studied effects for the existing interaction in this crop. Among these methods, the Additive Main Effects and Multiplicative Interaction (AMMI) has been found to be good tool due to its easy graphic display of superior genotypes, stability, as well as the megaenvironments clustering for recommendation purposes. In this study, data about Pol Tons per Hectare (PTH) obtained in the final stages of the 2003 Serie of the Sugarcane Technology Center were used to identify superior genotypes, stable, mega-environments for recommendation, specific adaptation, as well as crop interactions. Two harvests were made between the years of 2010 and 2011, the first on plant sugarcane, and the second on the ratoon sugarcane. The combined analysis between harvests was statistically significant at 5% probability for the environment (E), genotype (G) and harvests (H), as well as for the following interactions: G×E, E×H, G×H and E×G×H; showing greater diversity in the genotypes and environments discrimination during harvests. The overall average was 12.77 PTH, the coefficient of variation was 8.70%, and the selective accuracy was 98.63%; showing great levels of experimental quality. The genotypes G4, G14 and G16 were statistically superior; however, they did not show good stability on the average of harvests. There was specific adaptation between the environment E7 with the genotype G4, as well as between E5 and G5. In general, the environments clustering was not consistent during harvests, with the exception of the environments E1 and E4, which showed similarities in the genotypes discrimination.