Nonlinear genotypic response to macro- and microenvironments

Environmental variation can be due to macro- and microenvironments. Whereas macroenvironments, such as climate, density, and nutritional levels, are distinguishable, microenvironments including external random errors or internal 'accidents' of an organism cannot be well specified. A quantitative genetic model is proposed to estimate the genetic control of response of genotypes to these two kinds of environments. The model extends Gimelfarb's additive-multiplicative model for genotype x environment interaction by considering both macro- and microenvironmental variation and the mechanistic basis of genotypic response to a particular macroenvironmental factor. It is further extended to estimate genetic correlations between quantitative traits under the additive-multiplicative model. An example from a forest tree was used to illustrate the application and power of the new model. In many aspects, the model displays remarkable advantages over the traditional analysis of variance used to study genotype x environment interaction.