Genotype-Environment Interaction Concepts for Field Experimentation

Abstract

A reference set of experiments based on sets of random environments available in time and space was characterized and used as a framework for considering genotype-environment interactions in quantitative genetic studies. Biases, introduced when data collected for y years at 1 sites were analyzed as a set of random environments, were evaluated. The ramifications involved primarily the magnitudes of the estimates for two intraclass correlations. Experimental estimates for the correlations were obtained from published genotype-environment studies conducted primarily in North Carolina. With the procedures presently used to select test sites, the sites and years had similar effects on relative genotypic responses, and one should be able to analyze associated environments as random without unduly biasing his estimates of genetic components.