CAIMS Code of Conduct
“All participants are expected to comply with the CAIMS-SCMAI Code of Conduct”
Using machine learning methods and citizen-reported data to predict angler behavior
Recreational angler play an important role in providing researchers and managers with valuable data about the state of fisheries. Angler behavior (e.g., catch rates and angler pressure) can be tracked by mobile phone applications and online platforms at a low cost and in a high temporal and spatial resolution. However, in contrast to data gathered by conventional methods, citizen-reported data is biased toward certain groups of anglers. To predict “ground-truth” angler behavior in freshwater bodies across Canada we apply several machine-learning models. The models were trained and compared on their prediction performance using citizen-reported fishing trips, auxiliary variables on the environment, and conventional angler surveys. We illustrate how machine-learning methods can be used to support fisheries management and conservation in a rapidly changing environment.