The measurement and interpretation of genotype-environment interactions

TLDR: The regression analysis developed by Finlay and Wilkinson to investigate genotype-environment interactions and to assess genotypes for their adaptation to a range of environments is reviewed and conclusions were extended to consider variation in several environmental factors.

Abstract: The regression analysis developed by Finlay and Wilkinson to investigate genotype-environment interactions and to assess genotypes for their adaptation to a range of environments is reviewed. Their analysis used the mean yield of many genotypes to provide a measure of the environment; it was not measured in physical terms. To reveal aspects of their analysis it was applied to data of the response of genotypes to variation in a single precisely measured environmental factor. The conclusions were extended to consider variation in several environmental factors. The effects on the regression statistics that occur with different samples of genotypes, sub- and super-optimal environmental conditions, differences in periods of growth, changes in the scale of measurement and the occurrence of several stress factors are outlined. The study byBreese (1969) of genotype-environment interaction inDactylis glomerata is considered against a background of these effects.