Polygenic selection to a changing optimum under self–fertilisation

Abstract

Many traits are polygenic, affected by multiple genetic variants throughout the genome. Selection acting on these traits involves co–ordinated allele–frequency changes at these underlying variants, and this process has been extensively studied in random–mating populations. Yet many species self–fertilise to some degree, which incurs changes to genetic diversity, recombination and genome segregation. These factors cumulatively influence how polygenic selection is realised in nature. Here, we use analytical modelling and stochastic simulations to investigate to what extent self–fertilisation affects polygenic adaptation to a new environment. Our analytical solutions show that while selfing can increase adaptation to an optimum, it incurs linkage disequilibrium that can slow down the initial spread of favoured mutations due to selection interference, and favours the fixation of alleles with opposing trait effects. Simulations show that while selection interference is present, high levels of selfing (at least 90%) aids adaptation to a new optimum, showing a higher long–term fitness. If mutations are pleiotropic then only a few major–effect variants fix along with many neutral hitchhikers, with a transient increase in linkage disequilibrium. These results show potential advantages to self–fertilisation when adapting to a new environment, and how the mating system affects the genetic composition of polygenic selection. showing that there are two competing effects. First, it can expose mutations to selection more quickly, strengthening adaptation to a changing environment. Conversely, it can reduce the efficacy of selection through weakening the efficacy of recombination. We then use multi–locus stochastic simulations to investigate outcomes under more realistic scenarios, and find that high selfing can lead to higher fitness in the long–term, in contrast to classic expectations. We also investigate how many traits each variant influences, a property known as pleiotropy. If pleiotropy is absent we see that under very high levels of self-fertilisation, populations fix mutations with opposite effects on a trait. If pleiotropy is present then we instead see only a few major-effect genetic variants fixing in the population, alongside many neutral mutations. These findings provide insights into how natural populations adapt to changing environments.