A striking mismatch between genomic and phenotypic selection responses– the role of pleiotropy in complex environments

Abstract Background A major focus of evolutionary biology is the characterization of the genetic basis of adaptation. A widely used approach builds on the comparison of allele frequencies among populations adapted to different environments, reasoning that differentiated loci uncover the genetic basis of adaptation. Results We evolved replicated Drosophila simulans populations in novel laboratory conditions that differed only in their temperature regime. One group of populations evolved under a constant temperature of 23 °C, while the other experienced temperature fluctuations around the same mean. Consistent with a temperature-specific selection response, we find that 93% of the SNPs with significant allele frequency changes, are restricted to one of the two temperature regimes. Fitness measurements in each regime, however, do not support temperature-specific selection responses. Rather, replicate populations evolved in the two temperature regimes exhibit a similar fitness increase relative to the ancestral populations. Notably, this pattern is observed in both temperature regimes. We use computer simulations to test the hypothesis that this apparent discrepancy between genomic selection signatures and fitness increase can be explained by pleiotropy. Our simulations confirm that when populations share a common stressor, such as the novel laboratory environment in our experiments, temperature-triggered pleiotropic effects could lead to a temperature-specific genomic selection response, despite the phenotypic response being dominated by the shared stressor. Conclusions Our findings highlight the importance of considering pleiotropy during multivariate selection for the interpretation of allele frequency changes in complex environments.