���Ϲ���

On May 21, Ryma Benayed was walking to a subway entrance near Central Park when she spotted 14 Cedar Waxwings lying yellow-belly-up by a building. “It looked so strange that at first I thought it was a magic trick, so I got closer to see,” she says. “That’s when I understood that they were dead.” Later that day, Benayed, a technical director at Memorial Sloan Kettering Cancer Center, Googled around for the story. She couldn’t find a thing. “All of these birds die and no one talks about it?” she says. “I was shocked.”

What Benayed’s Googling did unearth was developed by New York City ���Ϲ���. She submitted a photo of the birds and the date, time, and location where she found them. “I’ve never reported anything before, but this time I felt like I had to,” she says.

Benayed’s report was one of 91 submitted to D-Bird during spring migration, an especially deadly time for birds. It’s estimated that 250,000 birds die by colliding with man-made structures in New York City each year; in North America, that figure falls somewhere between 300 million and 1 billion. Because reports are so scattered, a more precise estimate is difficult to pin down. “We have very little idea of what the actual number is,” says Darren Klein, a program and advocacy manager at New York City ���Ϲ���.

And so in 2014, Klein—then a graduate student in environmental policy and an ���Ϲ��� intern—decided to code a website to compile citizen sightings. Each entry to d-bird.org, which can be accessed via desktop or mobile device, becomes a data point on a map. NYC ���Ϲ���’s conservation biologist, Debra Kriensky, checks the map each day. If she sees that someone found an injured bird, she taps ���Ϲ���’s network of transporters to go to the scene and take it to a rehabilitation center. But the dead birds provide useful information, too. More than 550 collision reports have been submitted since the site launched.

“Eventually, we’ll be able to determine the hot spots,” says Susan Elbin, NYC ���Ϲ���’s director of conservation and science. Studying areas where many collisions occur could reveal what exactly makes one building or city corner so deadly. The data has also helped fill in gaps to the group’s , which deploys volunteers to track bird collisions along specific routes during spring and fall migration. For example, the Project Safe Flight surveys didn’t record a single American Woodcock during spring of 2015, yet woodcocks were the most reported species in D-Bird that same year.

Such a data-gathering tool could be useful all along migratory flyways—anywhere birds encounter human infrastructure. “This is an issue that almost every state and local chapter has to deal with,” Klein says. So this spring he adapted D-Bird for use by ���Ϲ��� Texas, ���Ϲ��� Minnesota, and the Atlanta ���Ϲ��� Society; a chapter in Silicon Valley will experiment with it this fall.

Adam Betuel, the Atlanta ���Ϲ��� Society’s director of conservation, notes that half of building collisions actually involve low-rise, or residential buildings, as opposed to the glassy skyscrapers that loom over Manhattan. “And that’s where D-Bird can really make a difference: individual cases outside of people’s homes,” he says. “We’re just getting started, but data is already trickling in that we otherwise would have missed.”

For her part, Benayed hopes the waxwing data will help other birds avoid the same fate. She’d like to see buildings become more bird-friendly, something New York City ���Ϲ��� has been actively working on. Last year, it partnered with the American Bird Conservancy to for local architects and engineers, educating them on the importance of installing patterned or angled glass on buildings and adding decals to existing windows. Of course, each building poses a unique threat that requires a tailored solution, and this is what makes D-Bird’s data essential.

“D-Bird bubbled up from one chapter to focus on a local issue, but it has the potential to spread widely,” says the ���Ϲ���’s director of community conservation, John Rowden. “It can lead to better, more systematic monitoring of particularly troublesome sites in each local landscape. Each ���Ϲ��� chapter will decide how they want to implement this tool and thus have a greater effect on the whole network.”