A research team from Indiana University Bloomington used the rare opportunity of the April 8, 2024 total solar eclipse to study how sudden changes in light affect bird behavior. The project was led by Ph.D. student Liz Aguilar in associate professor Kimberly Rosvall’s lab, with support from collaborators at the Luddy School of Informatics, Computing, and Engineering.
The study, recently published in Science, involved both traditional field monitoring and a public science initiative using a custom smartphone app called SolarBird. The app was developed to allow members of the public to observe and record bird behaviors before, during, and after the eclipse. “Scientists can’t be in a thousand places at once,” Aguilar said. “The app gets around this problem by leveraging the public as scientists. It also encourages people to look around and listen, adding to the show in the sky.”
SolarBird users were instructed to find a bird and document its activity for 30 seconds at different stages of the eclipse by selecting from options such as singing or flying within the app interface. Paul Macklin, professor at IU’s Luddy School who helped develop SolarBird, explained: “A main theme was how we could make data collection easy for newcomers, and minimize any distractions from the day’s event. We used the smartphone’s GPS coordinates and known formulas to determine the eclipse phase, which is the percentage of sun obscured by the moon.”
The initiative collected nearly 11,000 observations from more than 1,700 participants along the path of totality across North America. “It was wonderful,” Macklin said. “The app worked. And as we looked at the database that night, we saw the community had also worked its magic.”
Alongside this effort, Rosvall’s team deployed automated recording units in Bloomington and nearby rural areas during the week leading up to and including eclipse day. These devices captured extensive audio data on bird vocalizations. Dustin Reichard—an alumnus now teaching at Ohio Wesleyan University—used an artificial neural network called BirdNET to process almost 100,000 recordings quickly.
Focusing on 52 common species locally present during spring migration or breeding season, researchers found that 29 showed clear shifts in vocalization patterns tied directly to changes in light caused by totality. Reichard noted: “Even a very brief disruption in light – just over four minutes in this case – can cause large changes in behavior, especially for species that naturally produce a burst of song around dawn.”
Rosvall commented on broader implications: “It’s crazy that you can turn off the Sun, even briefly, and birds’ physiology is so tuned to those changes that they act like it’s morning. This has important implications on the impact of urbanization or artificial light at night, which are much more widespread.”
She added: “A lot had to align for this to work – that the eclipse would peak in this area at this time of year; that IU is a center of animal behavior research known internationally; that we had the people, the tools, and the public’s interest,” according to Reichard.
Rosvall hopes their success will inspire further collaborations between scientists and communities using AI-driven tools: “It was clear that engaging in this research project enhanced people’s joy in this experience,” she said. “It did for me and those who reached out.” She concluded: “It shows that sometimes a creative idea and a willingness to go all in is what you need to accomplish high-impact work.”
Funding for this project came from grants provided by both Indiana Space Grant Consortium and National Science Foundation.
