The missing E: the role of the environment in evolution and animal breeding

SUMMARY: In the development of applied quantitative genetics to date it was assumed that natural selection was absent. Hence, an important part of the environment has been ignored and breeders have concentrated on estimated breeding values. The same is implicit in molecular efforts to modify genotypes of animals directly. Furthermore, explanations of evolution imply that genes are responsible for evolutionary changes. These consequences of this assumption are now seen to be wrong. Beilharz (1998b) shows that the resources available in any environment limit the phenotypes that can develop there. This is because natural selection is always acting on phenotypes and has already selected organisms to be as good as it is possible to be on the available resources. This paper describes the necessary theoretical changes to quantitative genetics and the consequences that follow for evolutionary theory and animal breeding. A component of environment (Er, regional environment) describing that environmental variation to which natural selection adapts populations of organisms, is introduced. Variation among levels of Er reflects different amounts of resources available for animals to live on. At each level of Er genotypes are selected that have maximum phenotypic fitness exactly at that level, at the expense of genotypes with either lower fitness or a fitness demanding higher levels of resources to realise its potential. Changing environments (levels of resources) favour new phenotypes whose resource demand matches the changed resources available. The role of genes is to produce those variants that have phenotypes able to follow the environmental change. The paper also shows how genotype-environment interactions have resulted from the matching of genotypes to different levels of Er in earlier times. In the breeding of domestic animals, genetic improvement of production traits, without raising Er, must be accompanied by trade-offs (antagonisms) among production traits and with fitness. To obtain improvement with minimal negative side effects, breeding goals must be specified more comprehensively that to date, e.g. by also taking into account the duration of productive life. The selection criteria and estimates of heritability and genetic correlations used in the estimation of breeding values must be appropriate for the particular level of environment in which the genetic improvement is to be made.