Thermal regimes in rivers and streams are fundamentally important to aquatic ecosystems and are expected to change in response to climate forcing as the Earth's temperature warms. Description and attribution of stream temperature changes are key to understanding how these ecosystems may be affected by climate change, but difficult given the rarity of long-term monitoring data. We assembled 18 temperature time-series from sites on regulated and unregulated streams in the northwest U.S. to describe historical trends from 1980-2009 and assess thermal consistency between these stream categories. Statistically significant temperature trends were detected across seven sites on unregulated streams during all seasons of the year, with a cooling trend apparent during the spring and warming trends during the summer, fall, and winter. The amount of warming more than compensated for spring cooling to cause a net temperature increase, and rates of warming were highest during the summer (raw trend = 0.17°C/decade; reconstructed trend = 0.22°C/decade). Air temperature was the dominant factor explaining long-term stream temperature trends (82-94% of trends) and inter-annual variability (48-86% of variability), except during the summer when discharge accounted for approximately half (52%) of the inter-annual variation in stream temperatures. Seasonal temperature trends at eleven sites on regulated streams were qualitatively similar to those at unregulated sites if two sites managed to reduce summer and fall temperatures were excluded from the analysis. However, these trends were never statistically significant due to greater variation among sites that resulted from local water management policies and effects of upstream reservoirs. Despite serious deficiencies in the stream temperature monitoring record, our results suggest many streams in the northwest U.S. are exhibiting a regionally coherent response to climate forcing. More extensive monitoring efforts are needed as are techniques for short-term sensitivity analysis and reconstructing historical temperature trends so that spatial and temporal patterns of warming can be better understood. Continuation of warming trends this century will increasingly stress important regional salmon and trout resources and hamper efforts to recover these species, so comprehensive vulnerability assessments are needed to provide strategic frameworks for prioritizing conservation efforts.

Mountain streams provide important habitats for many species, but their faunas are especially vulnerable to climate change because of ectothermic physiologies and movements that are constrained to linear networks that are easily fragmented. Effectively conserving biodiversity in these systems requires accurate downscaling of climatic trends to local habitat conditions, but downscaling is difficult in complex terrains given diverse microclimates and mediation of stream heat budgets by local conditions. We compiled a stream temperature database (n = 780) for a 2500-km river network in central Idaho to assess possible trends in summer temperatures and thermal habitat for two native salmonid species from 1993 to 2006. New spatial statistical models that account for network topology were parameterized with these data and explained 93% and 86% of the variation in mean stream temperatures and maximas, respectively. During our study period, basin average mean stream temperatures increased by 0.38 degrees C (0.27 degrees C/decade), and maximas increased by 0.48 degrees C (0.34 degrees C/decade), primarily due to long-term (30-50 year) trends in air temperatures and stream flows. Radiation increases from wildfires accounted for 9% of basin-scale temperature increases, despite burning 14% of the basin. Within wildfire perimeters, however, stream temperature increases were 2-3 times greater than basin averages, and radiation gains accounted for 50% of warming. Thermal habitat for rainbow trout (Oncorhynchus mykiss) was minimally affected by temperature increases, except for small shifts towards higher elevations. Bull trout (Salvelinus confluentus), in contrast, were estimated to have lost 11-20% (8-16%/decade) of the headwater stream lengths that were cold enough for spawning and early juvenile rearing, with the largest losses occurring in the coldest habitats. Our results suggest that a warming climate has begun to affect thermal conditions in streams and that impacts to biota will be specific to both species and context. Where species are at risk, conservation actions should be guided based on considerations of restoration opportunity and future climatic effects. To refine predictions based on thermal effects, more work is needed to understand mechanisms associated with biological responses, climate effects on other habitat features, and habitat configurations that confer population resilience.

The distribution and future fate of ectothermic organisms in a warming world will be dictated by thermalscapes across landscapes. That is particularly true for stream fishes and cold-water species like trout, salmon, and char that are already constrained to high elevations and latitudes. The extreme climates in those environments also preclude invasions by most non-native species, so identifying especially cold habitats capable of absorbing future climate change while still supporting native populations would highlight important refugia. By coupling crowd-sourced biological datasets with high-resolution stream temperature scenarios, we delineate network refugia across >250 000 stream km in the Northern Rocky Mountains for two native salmonids-bull trout (BT) and cutthroat trout (CT). Under both moderate and extreme climate change scenarios, refugia with high probabilities of trout population occupancy (>0.9) were predicted to exist (33-68 BT refugia; 917-1425 CT refugia). Most refugia are on public lands (>90%) where few currently have protected status in National Parks or Wilderness Areas (<15%). Forecasts of refuge locations could enable protection of key watersheds and provide a foundation for climate smart planning of conservation networks. Using cold water as a 'climate shield' is generalizable to other species and geographic areas because it has a strong physiological basis, relies on nationally available geospatial data, and mines existing biological datasets. Importantly, the approach creates a framework to integrate data contributed by many individuals and resource agencies, and a process that strengthens the collaborative and social networks needed to preserve many cold-water fish populations through the 21st century.