Abstract
SignificanceTraits of wild animals can change over contemporary timescales, but concluding that evolution played a role requires demonstrating that trait change is linked to genetic change. This is because while selection acts on organisms’ traits, evolution in the strict sense is a process resulting in changes to the genome. But natural selection operating in natural ecosystems rarely acts in a single direction, and many factors that cause selection vary through time. We study wild stickleback in a well-studied lake to characterize how the genetics of correlated traits respond to different types of selection (e.g., directional or fluctuating). Our study clearly demonstrates how evolutionary processes cause trait change in the wild on a contemporary timescale.
A central goal in evolutionary biology is to understand how different evolutionary processes cause trait change in wild populations. However, quantifying evolutionary change in the wild requires linking trait change to shifts in allele frequencies at causal loci. Nevertheless, datasets that allow for such tests are extremely rare and existing theoretical approaches poorly account for the evolutionary dynamics that likely occur in ecological settings. Using a decade-long integrative phenome-to-genome time-series dataset on wild threespine stickleback (Gasterosteus aculeatus), we identified how different modes of selection (directional, episodic, and balancing) drive microevolutionary change in correlated traits over time. Most strikingly, we show that feeding traits changed by as much 25% across 10 generations which was driven by changes in the genetic architecture (i.e., in both genomic breeding values and allele frequencies at genetic loci for feeding traits). Importantly, allele frequencies at genetic loci related to feeding traits changed at a rate greater than expected under drift, suggesting that the observed change was a result of directional selection. Allele frequency dynamics of loci related to swimming traits appeared to be under fluctuating selection evident in periodic population crashes in this system. Our results show that microevolutionary change in a wild population is characterized by different modes of selection acting simultaneously on different traits, which likely has important consequences for the evolution of correlated traits. Our study provides one of the most thorough descriptions to date of how microevolutionary processes result in trait change in a natural population.