Evolution of Genetic Variance-Covariance Structure

One of the features of organisms that makes the study of biology so compelling is their apparent complexity. Molecular, cellular, developmental, physiological, neurological, and behavioral systems are each fascinating in and of themselves, but it is their interaction that generates what we see as the organism as a whole. By necessity, biologists have a tendency to break down organisms into their component parts to see how they work. Yet, no trait is an island (Dobzhansky 1956). In particular, we need to consider the organism in its entirety when looking at evolutionary change. This perspective is necessitated simply by the fact that it is the whole individual, not specific traits or particular alleles, that lives and dies. The whole individual is therefore the central unit of selection (Lewontin 1970). How can we deal with the complexity inherent in trying to integrate biological function and evolutionary change across the whole organism? Quantitative genetics provides an attractive potential framework because it attempts to summarize important features of the genetic system while focusing firmly at the level of the phenotype. The multiple trait (multivariate) formulations of quantitative genetics that we will discuss here can be seen both as a conceptual framework for integrating multiple systems at many levels and as a tool for making quantitative predictions about evolutionary change. Whether it succeeds on either count is the subject of active debate (Barton and Turelli 1989). Ways of successfully building upon the traditional statistical framework of quantitative genetics is one of the exciting areas of future growth within the field of evolutionary genetics.