Population-level migration modeling of North American birds through data integration with BirdFlow

Accurate information on the population-level movements of migratory animals is highly valuable for migration research and critical for designing effective conservation strategies for a changing world. However, population-wide movement information is lacking for most migratory species due to the effort and expense needed to collect data across species ranges. BirdFlow is a probabilistic modeling framework that infers population-level movements from weekly species distribution maps produced by the participatory science project eBird. Producing accurate species-specific BirdFlow models has required model tuning using high-resolution individual tracking data, which is not available for most migratory species. Here, we introduce a new model tuning framework that eliminates this reliance on tracking data and generalizes BirdFlow to hundreds of migratory species. This framework allows us to tune and validate BirdFlow models using a combination of data sources, including GPS tracks, banding recoveries, and radio telemetry data from the Motus Wildlife Tracking System. We investigate the generalizability of this approach by (1) investigating predictive performance compared to null models; (2) validating the biological plausibility of BirdFlow models by comparing movement properties such as route straightness, number of stopovers, and migration speed between model-generated routes and real movement tracks; and (3) comparing the performance of models tuned on species-specific movement data to models tuned using hyperparameters transferred from other species. Our results show that the BirdFlow modeling framework achieves biologically realistic performance, even for prediction horizons of thousands of kilometers and several months. When species-specific data are unavailable, models can still be tuned using data from other phylogenetically adjacent species to achieve improved performance. Alongside this study, we release 153 tuned BirdFlow models, representing the first collection of large-scale population-level movement forecasting models and offering a foundation for more accurate predictions for applications in conservation, disease surveillance, aviation, and public outreach.