Population dynamics in extreme environments: shared histories of strong selection lead to predictable patterns across separate high-elevation populations of small mammals

Many frameworks exist to identify species that are vulnerable to environmental change, but we still generally lack the ability to anticipate species' responses when changes occur. Exploring patterns of trait evolution along environmental gradients could help this process: conditions become more predictable and less variable at the ends of gradients, with past studies hypothesizing that consistently harsh environments will result in predictable evolutionary patterns. To test this hypothesis, we characterized genetic and phenotypic variation, as well as selection on this variation, in deer mice (Peromyscus maniculatus) along two transects spanning from 2120-4200m in the Colorado Rocky Mountains, U.S.A. We found that both genetic and phenotypic variation declined with elevation, that high elevation populations had higher rates of survival, and that fewer traits influenced survival at high elevations. These results suggest extreme environments may lead to predictable consequences for populations and provide a basis for forecasting population vulnerability to environmental change.