Relative controls on stream temperature from shade, land use, and water management in basins of the Pacific Northwest, USA
Citation:
Fuller, M., N. Detenbeck, P. Leinenbach, R. Labiosa, AND D. Isaak. Relative controls on stream temperature from shade, land use, and water management in basins of the Pacific Northwest, USA. Society for Freshwater Science Annual Meeting, Salt Lake City, Utah, May 19 - 23, 2019.
Impact/Purpose:
Stream temperatures are driven by both natural processes (e.g., weather) and anthropogenic impacts (e.g., land-use change/alteration). This research addresses the need to better understand patterns of stream temperatures across river networks and how natural and anthropogenic controls balance with each other. To do this, we use statistical spatial stream network models and a model selection process in which various land-use, climate, and water management processes compete and/or combine with each other to see which predict the stream temperature patterns best across study basins of the Pacific Northwest, USA. The results from this work provide insight into how restoration of streams and their uphill landscapes could be managed for restoring cold-water habitat for cold-water species (e.g., Pacific salmon).
Description:
Stream temperature is controlled by many factors including a landscape’s geology, climate, and human alterations. Our study compares the relative importance of these factors in controlling stream temperature using a statistical modelling framework for three study regions in the Pacific Northwest, USA. We tested dozens of competing and alternate covariates which include representations of landscape geology, climate, and land use, as well as covariates with potential management implications to see which were significant in predicting observed stream temperatures in the study basins. To do this, spatial stream network models were fit with these covariate combinations and a modified best subsets analysis selected among different models (using Akaike Information Criterion – AIC – as a model diagnostic statistic) to find a suite of best models. Using this suite of best models, model-averaging methods predicted stream temperatures under a variety of management plan scenarios. These predictions demonstrate how much water temperatures might shift under different restoration scenarios (e.g., channel width restoration, reduced agricultural water use, or riparian shade restoration). The results of this research provide information on the alternate mechanisms potentially controlling stream temperature and could help advise decisions for future temperature management plans.