Citation:

Leibowitz, S., R. Hill, M. Weber, D. Thornbrugh, EricW Fox, Tony Olsen, J. Flotemersch, J. Stoddard, AND D. Peck. SSWR 3.01B.1: National maps of watershed integrity and stream condition. SSWR webinar on “National maps of watershed integrity and stream condition”, NA, August 01, 2016.

Impact/Purpose:

As part of the Clean Water Act, the EPA is required to report on the integrity of the Nation’s water resources. To help support EPA’s Strategic Goal 2, “Protect and Restore Watersheds and Aquatic Ecosystems,” SSWR 3.01 was tasked with developing national maps of watershed integrity and stream condition. Watershed integrity was defined as the capacity of a watershed to support and maintain the full range of ecological processes and functions essential to sustainability. Stream condition is determined nationally based on results from the National Rivers and Streams Assessment (NRSA). This presentation provides an overview of these national mapping studies and makes the following four points: (1) The Index of Watershed Integrity assesses risks to key watershed functions; (2) NRSA data can be leveraged to predict probability of stream condition at individual stream and river reaches for all perennial rivers, including those not sampled; (3) Together, watershed integrity and probability of stream condition can inform restoration and protection efforts at multiple scales; and (4) Production of maps was made possible due to EPA’s StreamCat dataset. In order to make these points, the presentation first provides an introduction and background, followed by an overview of the StreamCat dataset. The presentation then separately describes the watershed integrity and stream condition studies. Finally, the presentation addresses future efforts. Along with three study manuscripts, this presentation contributes to the FY16 SSWR Annual Performance Reporting (APR) product 3.01B.1, “National maps of watershed integrity and stream condition and report and webinar describing these.” The approaches and results from this work could be important for three programs within the Office of Water: the National Aquatic Resource Surveys, the Healthy Watersheds Program, and the Biocriteria Program.

Description:

This presentation reports on two separate studies conducted under SSWR 3.01B as part of an FY16 Annual Performance Reporting (APR) product for the Office of Management and Budget. Three separate but related studies were conducted. The first study used EPA’s StreamCat dataset to quantify the IWI, and mapped this for the conterminous US. The authors also developed a related ICI, which was developed based on local stream segments (i.e., upstream areas were excluded). Regression analyses were used to evaluate the ability of the IWI and ICI to predict site-level indicators derived from EPA’s National Rivers and Streams Assessment. Results from this study show high integrity in the western US, intermediate integrity in the southern and northeastern US, and the lowest integrity in the upper midwest and lower Mississippi Valley. Although related to the IWI, the ICI could be useful for local applications where information on a mainstem river and its upstream catchments is not required. The IWI was able to account for a quarter of the national variation in a water quality metric that was derived using data from EPA’s National Rivers and Streams Assessment. While the IWI in its present form could be useful for management efforts, limitations concerning the absence of data for certain stressors needs to be taken into account. The second study used random forest models to predict the probability of good condition for the macroinvertebrate multi-metric index (MMI) at approximately 1800 NRSA sampling sites, using landscape information from StreamCat. They then used StreamCat data to apply this model to sites that were not sampled by NRSA. In all, probability of good condition was predicted for 1.1 million perennial stream segments. This study found that the random forest models correctly predicted the biological condition of 78% of sites nationally. Regional evaluations also suggested good model performance. Results are presented as maps of predicted probabilities of good condition, given upstream and nearby landscape settings. Inversely, the maps can be interpreted as the probability of a stream being in poor condition, given human-related alterations to the landscape. The presentation also references a third study that provides further assurance that the modeling approach of the second study is robust.

Record Details: