UC San Diego

Understanding what factors shape the magnitude of divergence under rapid evolution is critical. Phenotypic divergence, specifically, can be influenced by selective forces such as the environment (climate, geographic distance) as well as non-selective forces (genetic composition of the founding population). Threespine sticklebacks provide a unique opportunity to study the magnitude of divergence under rapid evolution. Marine sticklebacks independently colonized freshwater habitats at the end of the last ice age (~12,000 years ago) generating multiple replicate pairs that represent a natural experiment. However, studies rely on untested assumptions that marine sticklebacks are not phenotypically varied and are unchanged from their ancestors. Here, we test how differential environmental conditions impact the estimated magnitude of phenotypic divergence and parallelism of independent freshwater stickleback populations. We find that marine variation is comparable to freshwater variation. Importantly, the inferred magnitude of phenotypic divergence of each freshwater population is dependent on marine sampling location. The geographic distance and environmental similarity between the marine and freshwater pairs explain a significant degree of variance. When estimating the degree of parallelism among freshwater pairs, marine reference also affects the inferred magnitude. The observed pattern suggests the choice of reference population and its geographic distance are important aspects to consider when estimating freshwater divergence and parallelism. Results also implicate a signature of local adaptation and/or isolation by distance. We provide recommendations for choosing ecologically relevant marine references in future studies on this important evolutionary model system.