A hybrid modeling framework combining HAWQS and structural equation models to support existence valuation studies on aquatic biological assemblages
On this page:
The U.S. Environmental Protection Agency (EPA) is conducting a national stated preference study that seeks to quantify the non-use (existence) portion of total economic value of water quality policies across the conterminous US (CONUS). Previous work identified the ratio of observed-to-expected (O/E) taxonomic composition of macroinvertebrates as an appropriate biological condition indicator to capture existence value of aquatic ecosystems because O/E is 1) available across the CONUS in the EPA National Aquatic Resource Surveys (NARS) data and 2) interpretable by survey respondents. However, to be of use for resource valuation, O/E must be responsive to management actions in models that provide regulatory scenarios. Therefore, we propose developing a hybrid modeling framework to assess how regulatory actions affect aquatic biological integrity through changes in environmental conditions. This framework will support developing management scenarios and predicted biological outcomes that can be integrated into the stated preference surveys.
To test O/E as an endpoint for regulatory scenario assessment, we developed structural equation models with NARS data (NARS-SEM) to predict macroinvertebrate O/E in Western U.S. streams. The models showed that land use activities decreased macroinvertebrate O/E through pathways associated with degraded water quality (increased total nitrogen) and increased fine, unstable stream sediments. However, scenario analyses using NARS-SEM are limited because the predictors in the model are snapshots of conditions at the time of sampling and not directly tied to management actions. Therefore, alternative methods are needed to link NARS-SEMs to policy scenarios.
Scenario analyses can provide a better understanding of the connections between environmental regulations and aquatic ecosystem health. The Hydrologic and Water Quality System (HAWQS) is a process-based watershed modeling system designed to simulate water quantity and quality outcomes for CONUS that can be used to evaluate the effectiveness of various management scenarios (e.g., riparian buffers, green infrastructure) on water quality. HAWQS by itself cannot directly support existence value analyses because it does not predict the condition of aquatic biological assemblages. However, HAWQS water quantity and quality outputs can serve as inputs to other modeling approaches that directly model effects on biological assemblages.
To support existence valuation studies, we propose a hybrid modeling framework that will combine the scenario-modeling capabilities of HAWQS with the biological and ecological insights of NARS-SEM. With this HAWQS-SEM framework, we aim to connect regulatory and management activities to aquatic biological condition. We will discuss the challenges of bringing together two modeling approaches that operate at different spatial and temporal scales and have distinct data requirements. We outline considerations and caveats to implementing the HAWQS-SEM framework such as tradeoffs in spatial extent and model performance. The HAWQS-SEM framework is a promising analytic approach that can explore how specific management activities and future climate scenarios may affect aquatic ecosystem condition, which will be critical to support policy and management decisions.