Regional Assessments of Aquatic Intactness: Sensitivity of Inferences to Differences in Data Sources, Scoring, and Aggregation

Charged with managing over one-quarter of land in the United States (> 2.5 million km²), federal land management agencies are increasingly prioritizing restoration, conservation, and development through regional assessments of aquatic intactness (e.g. see the BLM’s new public land rule). Such assessments seek to infer ecological integrity (e.g., water quality, habitat connectivity, biological condition) from readily accessible measurements of land uses and surface disturbances. While ideally these assessments would be driven by direct measures of ecological condition using field data, the cost and scale of these analyses frequently prevent use of field data alone. Additionally, field data does not include information on large scale anthropogenic threats to aquatic intactness such as percent of the watershed with agricultural land use, which are critical to consider when prioritizing management actions. Despite rising popularity, neither the accuracy nor the comparability of regional desktop based aquatic assessments have been well tested, and no standard framework for their construction has emerged.

For this project we will review regional assessments of aquatic intactness conducted to date in the U.S. based on primarily GIS data and compare assessments from the upper Colorado River basin to compare how differences in scope, data types, and methods of scoring and aggregation can affect final assessment outcomes. We will attempt to validate assessment results for this region by assessing how strongly field measures of biological condition are associated with desktop based aquatic intactness scores. Our comparisons should: