Determination of Genotypic Stability and Adaptability in Wheat Genotypes Using Mixed Statistical Models

The objective of this study was to evaluate the adaptability and stability of wheat genotypes simultaneously in unbalanced Multi-Environment Trials (MET) in four different regions of Brazil, using the method of harmonic means of the relative performance of genetic values. Mixed model was applied to the analysis of Genotype- Environment Interaction (GEI) in wheat. Grain yield data were obtained from a network of MET carried out at seven locations from 2008 to 2010. A joint of experiments in complete randomized blocks design with some common treatments was used in all 21 experiments. Adaptability and stability parameters were obtained by several different methodologies, based on prediction, Harmonic Mean, and of the Relative Performance of Genotypic Values (HMRPGV). These methodologies ranked in a very similar way the studied genotypes and indicated the genotypes CD0950, CD0857, CD0667, CD0915, CD0914, CD0669, CD0859, and CD0851 as the superior ones for grain yield, adaptability, and stability in all environments. Dourados-MS (2010) was the worst environment with lowest mean (1,560.26 kg ha-1) and São Gotardo–MG (2008) was the best environment with highest mean (5,687.08 kg ha-1). The genotype more stable by HMRPGV across 21 environments tested was CD085; in the best environment, it was ranked the sixth (6,319.30 kg ha-1), but changed your values in the worst environment and was ranked the fifth (2,051.53 kg ha-1). The HMRPGV proved to be a practical and useful statistical tool in the determination of the Value for Cultivation and Use (VCU), particularly in the selection of genotypes’ reliability when genotypes are selected for the environments evaluated. This method has the advantage of providing results that are directly interpreted as breeding values ​​for yield, stability, and adaptability.