COS 145-9 - Developing stressor-watershed function relationships to refine the national maps of watershed integrity

Thursday, August 10, 2017: 4:20 PM
E142, Oregon Convention Center
Zachary Johnson1, Scott Leibowitz2 and Ryan A. Hill1, (1)Western Ecology Division, ORISE c/o US EPA, Corvallis, OR, (2)US EPA, Corvallis, OR
Background/Question/Methods

Human-induced ecological change within stream ecosystems resulting from various alterations is of major scientific concern due to the many social and ecological services provided by these systems. Previous work by the EPA developed a definition of watershed integrity as the capacity of a watershed to support and maintain the full range ecological processes and functions essential to sustainability. To aid in the evaluation of the conditions of the nation’s streams and rivers, an index of watershed (IWI) and catchment integrity (ICI) were mapped for the conterminous USA using an EPA spatial dataset of landscape indicators (StreamCat). First-order approximations of relationships between stressors and six watershed functions (hydrologic regulation, regulation of water chemistry, sediment regulation, hydrologic connectivity, temperature regulation, and habitat provision) were used in the initial national maps. The work presented here builds on these previous efforts by evaluating and improving methods used in establishing the national maps of watershed integrity.

Results/Conclusions

In this study, we developed metrics associated with the functional watershed integrity components and evaluated correlations graphically and through simple statistics using regional datasets from the Calapooia River (Oregon), Little Miami River (Ohio), and Narragansett Bay (Rhode Island) watersheds. We then used boosted regression trees to model potential stressors used in the initial maps against our developed metrics. Although preliminary, we found that many of the relationships between stressors and the six watershed functions are non-linear. These relationships display threshold behaviors, where a relatively small range in stressor values produces relatively large changes in a watershed function. This suggests that the IWI and ICI can be improved by substituting these empirical functions for the initial first-order approximations. We discuss the potential implications of these non-linear relationships for refining the national maps of watershed integrity.