Evolutionary Significance of Variation

Abstract

In this chapter, we consider the causes and evolutionary consequences of phenotypic variation among individuals in random-mating, unstructured populations. We focus on quantitative traits because they illustrate well the difficulties in determining genetic versus environmental causes of phenotypic variation. This is an important step when aiming to make precise predictions concerning phenotypic change in traits that influence individual longevity and reproduction. We also describe several recent conceptual advances concerning the evolutionary significance of population structure. Variation among individuals in quantitative traits is typically measured as total phenotypic variance (the variance estimated from all phenotypes measured in a population). This parameter has three convenient properties; first, total phenotypic variance can be partitioned into components that are themselves variances attributable to different causes, and second, these components are additive, summing to the total phenotypic variance. These attributes allow one to identify and to compare the magnitudes of different sources of variance to determine their relative evolutionary and ecological importance. Third, even when total phenotypic variance in a trait is high, resulting in a high degree of overlap among the means of different Variation among individuals in quantitative traits is typically measured as total phenotypic variance (the variance estimated from all phenotypes measured in a population). This parameter has three convenient properties; first, total phenotypic variance can be partitioned into components that are themselves variances attributable to different causes, and second, these components are additive, summing to the total phenotypic variance. These attributes allow one to identify and to compare the magnitudes of different sources of variance to determine their relative evolutionary and ecological importance. Third, even when total phenotypic variance in a trait is high, resulting in a high degree of overlap among the means of different populations or genotypes, the statistical control of one or more variance components often permits the detection of significant differences among phenotypic means. The proportion of total phenotypic variance accounted for by genotypic versus environmentally induced causes determines the degree of resemblance between parents and offspring or between other types of relatives (e.g., clonal replicates, siblings, half-siblings, maternal lineages). Given that a high degree of resemblance among relatives is a criterion for natural selection to cause evolutionary change, a major goal of evolutionary ecologists is to measure these variance components in wild populations.