Using regional scale flow–ecology modeling to identify catchments where fish assemblages are most vulnerable to changes in water availability

Funding information U.S. Department of the Interior Southeast Climate Center; U.S. Department of Agriculture Forest Service Eastern Forest Environmental Threat Assessment Center; U.S. Geological Survey National Water Census Abstract 1. Streamflow is essential for maintaining healthy aquatic ecosystems and for supporting human water supply needs. Changes in climate, land use and water use practices may alter water availability. Understanding the potential effect of these changes on aquatic ecosystems is critical for long-term water management to maintain a balance between water for human consumption and ecosystem needs. 2. Fish species data and streamflow estimates from a rainfall-runoff and flow routing model were used to develop boosted regression tree models to predict the relationship between streamflow and fish species richness (FSR) under plausible scenarios of (1) water withdrawal, (2) climate change and (3) increases in impervious surfaces in the Piedmont ecoregion of North Carolina, U.S.A. Maximum monthly flow, the fraction of total flow originating from impervious surface runoff, coefficient of monthly streamflow variability, and the specific river basin accounted for 50% of the variability in FSR. This model was used to predict FSR values for all twelve-digit Hydrological Unit Code catchments (HUC-12s) in the North Carolina Piedmont under current flow conditions and under water withdrawal, climate change and impervious surface scenarios. 3. Flow–ecology modeling results indicate that predicted FSR declined significantly with increased water withdrawals. However, the magnitude of decline varied geographically. A “hot-spot” analysis was conducted based on predicted changes in FSR under each scenario to understand which HUC-12s were most likely to be affected by changes in water withdrawals, climate and impervious surfaces. Under the 20% withdrawal increase scenario, 413 of 886 (47%) HUC-12s in the study area were predicted to lose one or more species. HUC-12s in the Broad, Catawba, Yadkin and Cape Fear river basins were most susceptible to species loss. 4. These findings may help decision making efforts by identifying catchments most vulnerable to changing water availability. Additionally, FSR-discharge modeling results can assist resource agencies, water managers and stakeholders in assessing the effect of water withdrawals in catchments to better support the protection and long-term conservation of species. Accepted: 9 October 2017