Increasing the Accuracy of Analyzing GXE Interaction and Integrating the Information to 'P. Radiata' Breeding Program

Abstract

The importance of genotype by environment (GxE) interaction can help breeders to determine an optimal breeding program. The efficiency of a breeding program relies on precise information relating to the GxE interactions. Alternative techniques were applied on three series of controlled mating experiments data to address different issues of GxE interactions. The GxE interaction for different traits, different genetic levels and different ages were examined. Four classical methods were used to analyze GxE interactions from population to individual levels. REML approaches based on mixed linear models were used to test the sources of GxE interactions, and to measure the GxE interactions by correcting for the heterogeneity of variances. Using pedigree information, the additive and non-additive genetic variances were partitioned and therefore the concepts of GxE interactions were extended for different genetic levels. The techniques of GGE biplot analysis and factor analytical model (FA2) were used to verify the patterns of GxE interactions in different trial series. The spatial analysis was explored to identify the possible causes of spatial patterns and investigate the microenvironments and genotype by microenvironment interaction within each trial. The theoretical genetic gain was predicted for specific vs. broad adaptation strategies. ... The methods used in this study provide different types of information. Compared with traditional methods, the REML approach shows its power for interpreting GxE interactions. Factor analytic and spatial model can be further explored to provide more underlying information relating to the occurrence of GxE interactions.