We have long been interested in understanding how populations adapt to both fixed and changing environments. One important factor that affects adaptation is the movement of individuals between groups and subpopulations, or dispersal. In this study, we use mathematical models and computer simulations to investigate how limited dispersal affects the process of adaptation.

We found that, contrary to previous findings, limited dispersal can actually speed up adaptation in some cases. Specifically, when a beneficial trait is controlled by a recessive gene, limited dispersal can increase the frequency of homozygous individuals, which in turn exposes the recessive gene to selection and speeds up the process of adaptation. However, this effect only occurs when dispersal is not too limited, otherwise the inflation of effective population size caused by dispersal limtation outcompetes any other possible effect of selection. We also found that the mode of adaptation (i.e. “hard” or “soft” sweeps) can affect the consequences of dispersal limitation and the signatures that those sweeps cause in genomes.

Overall, our study highlights the complex interplay between dispersal, selection, and genetic drift in the process of adaptation. By shedding light on these factors, we hope to contribute to our understanding of how species respond to changing environments, and how we can better conserve biodiversity in the face of global change.

The first publication of this work with my PhD supervisor, Charles Mullon, is available on Genetics.

Tandem Repeat sequences (TRs) are a significant source of genetic structural variation within and between individuals. These sequences occur when a short DNA motif is repeated head-to-tail along a chromosome. Variation in TRs is thought to reflect a balance between purifying selection, amplification, unequal recombination, and genetic drift.

The influence of these evolutionary processes on TRs is well understood in large haploid populations where individuals mate randomly. But many organisms are capable of selfing, which leads to inbreeding and an excess of homozygosity with relevant consequences for the evolutionary processes that shape variation in TRs. In this project, we use mathematical modeling to investigate the evolutionary dynamics of homologous TRs under the joint actions of purifying selection, amplification, unequal recombination, and genetic drift in a population of partially selfing diploids.

With my PhD supervisor Charles Mullon, we find that the equilibrium distribution of TRs in the population depends on the interaction between unequal recombination and selfing. When the selfing rate is low and homologous sequences within individuals tend to be different, unequal recombination reduces variation. However, when selfing is common and homologous sequences within individuals are more similar, unequal recombination increases the variation in TRs. In addition, because selfing increases variation between individuals relative to within individuals, purifying selection tends to be stronger in selfers compared to outcrossers. As a result, selfers have on average shorter TRs, in spite of experiencing increased genetic drift.

In subdivided populations, different ecological processes drive intraspecific phenotypic variation at different spatial scales. While spatial heterogeneity promote distinct morphs in different patches leading to local adaptation, negative-frequency dependent selection arising from local competition favours polymorphism within patches. The emergence and maintenance of these different phenotypes is not straightforward when traits are polygenic as hybridization breaks up adaptive allele combinations. In this case, selection acts on genetic architecture to concentrate allelic effects onto as few independently segregating loci as possible. This concentration may be achieved via different genetic mechanisms such as amplifiers of allele effect size or suppressors of recombination, but it remains unclear whether selection favours different mechanisms in different ecological settings.

In this project, together with my PhD supervisor Charles Mullon and the postdoc at our lab Ewan O. Flintham, we use computational models to study the evolution of a trait’s genetic architecture through various modifiers of allelic effects and of recombination, varying the extent to which polymorphism in the trait is favoured by spatial heterogeneity or local competition. We show that owing to more frequent hybridisation, local competition exerts stronger selection on genetic architecture than spatial heterogeneity. Furthermore, these different ecological processes favour different levels of genetic dominance and different patterns of recombination. In particular, local adaptation often involves large regions of recombination suppression when local competition generally sees multiple smaller regions scattered across the genome.