Elevation-dependent warming of streams in mountainous regions: implications for temperature modeling and headwater climate refugia

Abstract

Abstract Climate change is warming stream temperatures with significant implications for species that require cold temperatures to persist. These species often rely on headwater habitats in mountainous regions where elevation gradients in hydroclimatic conditions may induce differential patterns of long-term warming that affect the resistance of refugia. Forecasts from mechanistic and statistical stream temperature models diverge regarding whether this elevation dependence will cause above- or below-average warming in headwaters during warm summer periods, so we examined monitoring records for stream temperature (n = 271), air temperature (n = 690), and stream discharge (n = 131) across broad elevation gradients in a mountainous region of western North America to better understand potential future trends. Over a 40-year period characterized by rapid climate change from 1976–2015, air temperature stations exhibited below-average warming rates at high elevations while stream discharge declined at above average rates. Between climatically extreme years that involved summer air temperature increases >5 °C and discharge declines >70%, temperatures in high-elevation streams exhibited below average increases but otherwise showed negligible elevation dependence during intermediate climate years. In a subsequent example, it was demonstrated that elevation dependent stream warming has a minor effect on the amount of thermal habitat loss relative to the average water temperature increase within a mountain river network. We conclude that predictions of above average warming effects on headwater organisms for this region may be overly pessimistic and discuss reasons why different types of temperature models make divergent forecasts. Several research areas warrant greater attention, including descriptions of elevation-dependent patterns in other regions for comparative purposes, examination of long-term stream temperature records to understand how sensitivity to climate forcing may be evolving, use of new data sources to better represent key processes in temperature models across broad areas, and development of hybrid models that integrate the best attributes of mechanistic and statistical approaches.