Unraveling the long-term population evolution under fluctuating environmental condition

Abstract

Evolutionary biologists have been focusing on uncovering the effects of fluctuating environmental conditions on genetic change in populations. However, the extent to which these periodic forces associated with environmental alteration influence organisms at both genomic and phenotypic levels remains unclear. Furthermore, despite compelling genomic and phenotypic evidence of seasonally fluctuating forces from recent studies, there is a disconnect between empirical findings and theoretical models regarding the mechanisms underlying those observations, due to limited evidence on the scale and processes that generate stable genome-wide oscillations. This dissertation aims to shed light on the genetic and ecological factors that generate seasonal polymorphic alleles and to identify parameters that produce stable, long-term oscillatory patterns in allele frequencies. To address longstanding challenges, the studies in this dissertation combine simulation and empirical genomic data, integrating quantitative and population genetics to simulate a population under various selection regimes. The forward evolutionary software SLiM is used to incorporate genetic parameters, such as heritability and genomic architecture, as well as ecological parameters, including selection pressure, season length, and demographic changes (e.g., recurrent bottlenecks). Empirically, the Evolve and Resequence method is applied to evolve budding yeast cells (Saccharomyces cerevisiae) exposed on fluctuating treatments different laboratory temperature regimes. The timeseries genomic data obtained are used to validate the simulation models and assess the credibility of parameter values that produce oscillations in allele frequencies. Unlike early models that focused on selection acting directly on loci, the current model evaluates individual fitness based on the shift in the seasonal optimum relative to the mean phenotype. Furthermore, it applies spectral analysis to detect periodicity, indicating cyclical selective environments. Overall, the study shed light on the conditions that sustain oscillations in allele frequencies over time and highlighted the conditions under which demographic changes may impede the detection/underestimation of the fluctuating selection effect. While spectral analysis successfully identifies periodic patterns in allele frequencies, this approach requires an enormous amount of data across multiple time points, which can be challenging to generate empirically. Therefore, selecting appropriate time points will be required to confidently use this approach, as long as the sequence cost remains relatively high. All in all, the study confirms the ability of fluctuating environmental conditions to generate genome-wide seasonally shift signals through simulations and experiments, and it also shows the underlying parameters that are relatively easy to predict from natural populations; hence, contributing to the ongoing debates of uncovering the source of variations in natural populations and predicting the evolutionary outcomes from constantly changing environments.