Genetic Correlations: The Quantitative Genetics of Evolutionary Constraints

Abstract

Dictionary definitions of “constraint” generally indicate some aspect of restricting or confining the possible states or actions of individuals or systems. In the disciplines of linear and dynamic programming, a constraint is a function or inequality specifying the range of permissible values of variables. If the constraints were removed, the dynamical and equilibrium properties of a system of equations may be quite altered. A constraint may be either fixed, specifying an absolute limit on a particular variable, or it may be dynamic and restrict the joint behavior of two or more variables. For the purpose of this discussion, a genetic constraint will be defined functionally as those aspects of the inheritance of traits that prevent natural selection from resulting in a steepest ascent approach of the mean phenotype to an optimum. Genetic constraints can be fixed, as in the case when genetic variation for a particular phenotype does not exist, or dynamic, as in the case of genetic correlations depending on allele frequencies and linkage disequilibria. The primary genetic reason to be suspicious of the adaptationist program (Gould and Lewontin 1979) is the prevalence of genetic correlation. The intent of this chapter is to examine constraints as themselves being evolved traits, just as the processes of Mendelian genetics are themselves the products of evolution. The utility of specifying an explicit relationship between phenotypic characters and fitness based on demographic or physiological principles will be demonstrated.