Abstract
The rapid global loss of biodiversity calls for a robust understanding of how populations will respond demographically to environmental change. The viability of many populations will depend on their ability to adapt to rapidly changing environmental conditions. Heritability of traits under selection is known to affect the efficacy of adaptation. However, the effects of the genetic architecture underlying the heritability of selected traits (e.g., the number and effect sizes of the causal loci) on population viability remain poorly understood. We found from a wide range of deterministic theoretical models and individual-based simulations of divergent life histories (approximating corals and large mammals) that the genetic architecture of a selected trait can strongly affect population viability during selection associated with a rapidly shifting phenotypic optimum. Polygenic architectures (i.e., many loci, each with a small phenotypic effect) appear to confer greater adaptation and higher population viability than genetic architectures including large-effect loci responsible for 50-90% of the initial heritability of selected traits. Our results also suggest that the viability of populations with large-effect loci can depend strongly on initial allele frequencies, with already-frequent and very rare positively-selected alleles conferring low adaptive potential and viability compared to moderately low large-effect allele frequencies. These results uncover a crucial role of the genetic architecture underlying the heritability of fitness traits in determining eco-evolutionary dynamics and population viability during rapid environmental change.