View Program - 2022 Population, Evolutionary, and Quantitative Genetics Conference

Phenotypic adaptation can occur rapidly as a response to a sudden - natural or artificial - change in the environment. In the early phase of adaptation, the evolution of gene frequencies will be mainly driven by directional selection. While molecular population genetics focuses on the dynamics of single loci, quantitative genetics has a more trait-centered view. We combine these two approaches to describe adaptation of complex traits by studying the evolutionary dynamics at individual loci. For a finite panmictic population, we derive accurate approximations for the distribution of newly arising beneficial mutations as a function of time, first for a single locus. Then, using an infinite sites model and assuming unlinked loci (sites) loci contributing additively to a quantitative trait under weak non-epistatic directional selection, we derive highly accurate approximations for the evolutionary dynamics of the phenotypic mean and variance. Thus, we examined the effect on polygenic adaptation of selection, random genetic drift, and mutation rate and mutational effects drawn from a distribution. We also present results for the number of segregating sites during adaptation. The mathematical model is based on a combination of branching process theory (for the initial stochastic phase) and deterministic theory. Although diffusion theory leads to simple expressions for many quantities such as fixation probabilities and times, time-dependent results seem to be out of reach. Our approach yields this time dependence and is especially accurate in the initial phase of adaptation. However, we also derive highly accurate results for the stationary phase under long-term selection, i.e., when the phenotypic variance has stabilized. These refine classical results. Analytic approximations are tested by comprehensive simulations based on a Wright-Fisher model. As an application, we explored when the response of the trait is mainly caused by selective sweeps at few loci and when it is due to subtle allele-frequency shifts at many loci. We found that the selection strength determines primarily the rate of adaptation. The central parameter determining the relative importance of sweeps vs. shifts is the population-wide mutation rate. However, unequal mutational effects of different loci blur this distinction, and intermediate patterns occur in a wider parameter range if mutational effects have large variance.

Response of Quantitative Traits to Directional Selection in Finite Populations