Long-term selection experiments: Epistasis and the response to selection

Abstract

Two common features of long-term selection experiments are that, first, there is typically no evidence for selection limits due to exhaustion of genetic variation, and second, selection plateaus are frequently observed that last multiple generations before a response to selection is resumed. These features are usually attributed to the high mutation rates of quantitative traits, and the effects of linkage disequilibrium. Using previously published theoretical results and a simple deterministic model I explore the potential role of gene interaction in generating these patterns seen in the response to long-term selection experiments. I show that epistasis provides a pool of variation that can contribute to an extended response to selection, but that this extended response will, at least under some circumstances, result in intermediate selection plateaus.