Watershed characteristics influence winter stream temperature in a forested landscape

2021 
Winter stream temperatures, though infrequently studied, exert important influences on aquatic communities. To quantify effects of watershed physical characteristics on stream winter thermal regime, 54 streams (watershed area = 0.2–7.9 km2; altitude < 1300 m) on the Olympic Peninsula, Washington, USA were monitored hourly for 4 years. During the study, an exceptionally warm winter (2015) was used to evaluate influences of watershed characteristics under climatic conditions similar to those projected for mid-twenty-first century. Four watershed characteristics were hypothesized to influence winter stream temperature: stream size, elevation, solar exposure, and presence of glacial materials overlying bedrock. Larger streams were associated with colder winter water temperatures and higher thermal sensitivity to atmospheric conditions. Elevation—the strongest driver of winter stream temperatures—was negatively correlated to stream temperature, except on the coldest 15 days of winter when it had no influence. Watershed solar exposure had only a marginal influence on how cold streams were in winter but was positively correlated to diel stream temperature variation and thermal sensitivity. Streams in watersheds with glacial material overlying the sedimentary bedrock were colder and had less diel variation and lower thermal sensitivity than streams in watersheds where glacial material was not present. During the warm 2015 winter, the influences of watershed characteristics on temperature tended to be weaker compared with the other years. These insights improve our understanding of how watershed physical characteristics influence stream winter thermal regimes and how these winter thermal regimes vary across landscapes, facilitating development of predictive models, a first step in designing management plans that account for winter thermal habitat needs.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    82
    References
    0
    Citations
    NaN
    KQI
    []