Abstract
In a changing climate, organisms can either adapt to the new environment or adopt new geographical distributions that match the new location of the previously experienced conditions. Shifts in distributions can be thus simply predicted from the rate and direction of shifting environmental conditions: the velocity of climate change. In this study we compare distribution shifts in marine species with predictions based on climate velocity. Shifts reported in the scientific literature have been compiled into a database and, by calculating specific velocities for the time period and location of those changes, have been compared with expectations based on simple tracking of shifting thermal environments. Velocity estimates alone cannot predict the likely changes in diversity patterns that result from these geographical shifts. We present a method involving the forward projection of velocity-based trajectories through geographical space that mimics the shifts of many species. The collective behaviour of trajectories identifies areas as sinks and sources for climate-related change in biodiversity, as well as areas of little change and high flux corridors. Sinks are typically found along poleward-facing coasts on land and equatorward-facing coasts in the ocean, as well as on mountaintops and in semi-enclosed seas. Sources occur in locally warmer areas: at the equator in the ocean, and along coasts on land. Sinks represent areas where climate migrants are blocked by geographical or thermal barriers and thus face local extinction. Sources areas are unconnected to warmer environments and thus where no new climate migrants can appear to replace those that leave. Sources and sinks are thus both likely areas for species loss, but through different mechanisms.