An integrated drought early warning system to support cold-water fisheries management

Abstract Low streamflow resulting from drought and human water use is a threat to freshwater fisheries throughout western North America, creating a need for effective tools to aid management and fisheries use. We developed a drought early warning framework that forecasts river conditions that could result in temporary fishery closures and restrictions across Montana’s renowned trout fisheries. This tool provides reach-specific, probabilistic decision-support for proactive management under drought conditions. We calibrated statistical models using historical streamflow and stream temperature data from U.S. Geological Survey (USGS) gages to predict the likelihood of exceeding low-flow and high-temperature fisheries closure thresholds within 1- to 4- week outlooks during the warm season (May-Oct). These forecasts demonstrated consistently high model performance in predicting low-flow fishing closure criteria across all rivers (4-week outlook: mean R2 = 0.96, IQR = 0.94-0.97) and variable but often high model performance for temperature-based (“hoot-owl”) restriction criteria (mean R2 = 0.80, IQR = 0.55-0.86), particularly at sites which frequently experience higher stream temperature conditions. This drought early warning system—available via the web-based tool FishCast (https://usgs.gov/apps/fishcast)—provides daily-updated seasonal forecasts to help anticipate effects to recreational fishing opportunities. By quantifying forecast uncertainty and providing reach-specific probabilities, the system supports proactive management, improves coordination between fisheries and water managers, and provides stakeholders with actionable information on probable future impacts to fishing access within a framework that is broadly applicable to other freshwater systems and water resource management challenges.