Computer scientists at the University of Massachusetts Amherst, in collaboration with biologists at the Cornell Lab of Ornithology, recently reported in the journal Methods in Ecology and Evolution a new, predictive model that can accurately predict where a migratory bird will go next — one of the most difficult tasks in biology . The model is called BirdFlow, and while it’s still being refined, it should be available to scientists within a year and eventually to the general public.
“People have been trying to figure out bird migration for a really long time,” says Dan Sheldon, a professor of information and computer science at UMass Amherst, the paper’s lead author and an avid amateur bird watcher. “But,” adds Miguel Fuentes, lead author of the paper and a graduate student in computer science at UMass Amherst, “it’s incredibly difficult to get accurate, real-time information about which birds are where, let alone where exactly they’re flying.”
Many efforts, both past and ongoing, have been made to tag and track individual birds, yielding invaluable insights. However, it is difficult to physically tag birds in large enough numbers – not to mention the cost of such an undertaking – to create a complete enough picture to predict bird movements. “It’s really hard to understand how an entire species moves across a continent using tracking approaches,” says Sheldon, “because they tell you what routes some birds caught in specific places took, but not how birds in completely different places might move. “
In recent years, there has been an explosion in the number of citizen scientists monitoring and reporting sightings of migratory birds. Birders from around the world contribute to more than 200 million bird sightings annually through eBird, a project managed by the Cornell Lab of Ornithology and international partners. It is one of the largest biodiversity-related science projects in existence and has hundreds of thousands of users, enabling state-of-the-art modeling of species distribution through the eBird Status & Trends lab project. “The eBird data is amazing because it shows every week where birds of a species are moving throughout their range,” says Sheldon, “but it doesn’t track individuals, so we have to infer what routes individual birds take to best explain the species. -level patterns.”
BirdFlow pulls from eBird’s Status & Trends database and its estimates of relative bird abundance, then runs that information through a probabilistic machine learning model. This model is tuned using real-time GPS and satellite tracking data so it can “learn” to predict where individual birds will move next during migration.
The researchers tested BirdFlow on 11 species of North American birds—including the American Woodcock, Wood Thrush, and Swainson’s Hawk—and found that BirdFlow not only outperformed other bird migration tracking models, but could accurately predict migration flows without real-time GPS and satellite tracking data, making is BirdFlow a valuable tool for tracking species that can literally fly under the radar.
“Birds today are experiencing rapid environmental change, and many species are declining,” says Benjamin Van Doren, a postdoctoral researcher at the Cornell Lab of Ornithology and co-author of the study. “With BirdFlow, we can unify different data sources and paint a more complete picture of bird movements,” adds Van Doren, “with exciting applications for leading conservation actions.”
Read Now:<strong>Canadian Foreign Minister Melanie Joly will visit India next week for trade deal</strong>
