Meeting Abstract
Phenotypic covariation occurs within and among morphological, behavioral, physiological and life history traits at four biological scales: within individuals, within populations, among populations, and among species. However, the causes of trait covariation often change over these scales, making it difficult to use patterns at one scale to infer process at other scales. The proliferation of trait covariation studies has resulted in increasingly specialized terminology that is used to describe many patterns of phenotypic correlation, often independent of the scale of covariation and of any explicit structural, functional, or evolutionary hypotheses about why and how traits are related. We use a functional performance framework to identify different trait relationships and then we link these to fitness landscapes in order to explore hypotheses about the form and evolution of trait relationships. Our approach generates predictions about how covariance structure should change within and among populations under different forms of selection. This framework is not trait- or taxon-specific and is relevant to studies from individuals to species. It also contributes to our understanding of phenotypic evolution by allowing us to better classify the performance consequences of trait linkage and use these to develop hypotheses about genetic architecture, evaluate the interplay between selection and constraint, and explain the persistence of maladaptive traits.
