BAER, C.F.; FULLER, R.C.; TRAVIS, J.*; University of Florida; Florida State University; Florida State University: When and How Selection Experiments Might Actually Be Useful

Selection experiments have traditionally been motivated by a desire to understand either (1) the traits underlying fitness in a given environment or (2) the genetic architecture of a trait (or traits). Comparative biologists have pursued the first to understand adaptation, whereas breeders and quantitative geneticists have pursued the second to understand the constraints surrounding character evolution. Here we discuss the use of selection experiments in understanding the evolution and genetic architecture of complex traits, i.e. traits that either are affected by many genes or that are coupled through pleiotropy with other traits. By “genetic architecture” we refer to the pattern of genetic (co)varaiance between traits within and between populations and higher taxa and to the underlying genetic causes – pleiotropy and/or linkage disequilibrium fostered by natural selection. We describe two uses for selection experiments. First, selection experiments can be used to understand the evolutionary divergence of complex traits. By re-running the evolutionary process we can begin to understand the relative likelihoods of different evolutionary outcomes. Second, selection experiments can be used to elucidate the genetic architecture underlying complex traits. We review the strategies for selection experiments that seem best suited to each application and outline the practical considerations that must be confronted. We conclude that selection experiments can provide an important tool to help dissect genetic architecture, but care must be taken to (1) properly frame the question of interest, and (2) design an experiment that can answer the question, at least in principle.