As Birds Decline, High-Precision eBird Models Offer a Vision of Hope

Five years after the 3 Billion Birds Lost research, a follow-up study published in the journal Science showed how eBird Trends can reveal the patterns behind America's bird declines—and pinpoint areas for future recoveries.

From the Summer 2025 issue of Living Bird magazine. Subscribe now.

Five years ago, most ornithologists had a general sense that birds were declining.

The shocking aspect of the research published in the journal Science in September 2019, which made news headlines around the world, was the cumulative loss when all 500-plus species trends across the U.S. and Canada were tallied up—more than 3 billion birds lost in the past 50 years.

“I think the magnitude was pretty striking. Especially because I was born in 1970,” says Amanda Rodewald, the senior director of the Center for Avian Population Studies at the Cornell Lab of Ornithology. “So in my lifetime alone, almost one in three birds lost.”

In May 2025 a new expansive study of North American bird population trends (also published in Science) showed that bird declines have continued, and in some cases, accelerated in recent years.

But this study also did what the previous research couldn’t. Fueled by tens of millions of eBird checklists and sophisticated big-data-crunching technology, it added a geospatial element that provides a deeper look into the dynamics of how these bird declines are playing out. And importantly, it provides clues about where to focus conservation efforts most effectively to turn declines around.

Alison Johnston, the lead author on the 2025 Science study, says there’s an action-oriented component to this eBird Trends–based research.

“The 3 billion birds [research] was telling us, ‘There’s an emergency,’” says Johnston. “And now we have the information for an emergency response plan.”

an “MRI Scan” for Conservation

The May 2025 Science study relied on eBird Trends models that were built using more than 6 million hours on a National Science Foundation–funded supercomputing network to process 36 million eBird observations from 2007 to 2021.

“Without the massive amount of data available from eBird, we would not have been able to complete this study,” says study coauthor Daniel Fink, a Cornell Lab senior research associate and statistician. Fink has been leading research on eBird computer modeling for two decades.

This study produced individual geospatial assessments for 495 bird species across North America, and found that 75% of bird species were showing overall declining trends—with 67% of species showing widespread losses across more than half of their ranges.

“We’re seeing a fundamental shakeup and reorganization of the natural world,” says Johnston, who started this research as a Cornell Lab postdoctoral researcher and is now an ecological statistician at the University of St. Andrews in Scotland. “We’re viewing [these declines] through birds, but likely it’s also happening at the level of insects and plants and many other parts of nature.”

Rodewald, who is also a coauthor on the 2025 Science study, agrees that the results are concerning for bigger reasons than just birds: “Birds are excellent indicators of environmental change. … If environments aren’t healthy enough to support birds, they’re not likely to be healthy for human populations either.”

Rodewald and Johnston also stress that this research isn’t just concerning, but also exciting—because the fine-scale resolution of the eBird Trends models allows scientists to drill down into bird declines at 27 km x 27 km pixels (or about 17 miles x 17 miles), the smallest parcels of land ever used for such an analysis across an entire continent. They say it’s like the difference between seeing that your knee is swollen, and getting an MRI that tells you exactly which ligaments are injured, thus providing the information to go about healing it.

“This is the first time we have had fine-scale information that … allows us to understand the drivers [of declines] much better than we’ve ever been able to before,” says Rodewald.

“We’re really hoping that this marks a new era of conservation.”

Downward Trends in the Heart of Great Blue Heron Country

The epicenter of North America’s Great Blue Heron breeding population is along the Mississippi River Valley, where local populations of herons have declined by upwards of 20% to 40% over just the past decade, according to eBird Trends data. Explore the full eBird Trends map for Great Blue Heron.

The Strongholds Are Shrinking

When Johnston, Rodewald, Fink, and the other 13 authors of the 2025 Science study (all of whom work for or are affiliated with the Cornell Lab) used the fine-scale resolutions in eBird Trends to geographically analyze population trends for hundreds of North American bird species, they didn’t believe their eyes.

“We thought, this is weird. This doesn’t quite make sense,” recalls Johnston. “We went about doing further analyses, double-checking whether this could have been caused by any statistical artifact or anything else in the data that might have led to this.

“We ruled out all the other options. … This is a real ecological process happening here.”

That process, as stated in the abstract of the 2025 Science study: Bird “populations tended to decline most steeply in strongholds where species were most abundant.”

Put another way, populations of birds are plummeting in the very places that have historically been their population centers—such as the Mississippi River Valley for Great Blue Herons, and the Great Plains for Burrowing Owls. If America’s bird populations were a block of ice, they aren’t melting around the edges. They’re dissolving at the core.

“The center of abundance [is] where it should be best,” says Rodewald. “Those strongholds in the places where conditions have historically been best … that’s where we’re seeing these declines happening that are the steepest.”

“We’re not just seeing small little shifts happening,” says Johnston. “We’re actually seeing these populations declining where they were once really abundant. Where the habitat and the climate was once really suitable for them, it no longer is.”

The study authors posit a few possible explanations for this pattern of declines in population strongholds, which was observed in 83% of the 495 bird species in the analysis. For example, they write that the historically high-quality habitat sites could be some of the places most affected by stressors such as climate change, land-use changes, and pollution.

Eliot Miller—one of the study coauthors who started working with eBird data as a Cornell Lab postdoctoral researcher in 2015, and is now manager of the American Bird Conservancy’s BirdsPlus Index biodiversity conservation project—sees trouble for the source populations of some species when he looks at the eBird Trends maps.

According to the Cornell Lab’s Handbook of Bird Biology academic textbook, a source population for birds is a population that produces more offspring than the area can hold, leading some young birds to disperse to new areas. Miller says the big, bright red dots on some eBird Trends maps— meaning big declines in areas of high abundance—are warning lights for the source populations of some species, even for birds that are seemingly common today.

“If the source is dying off,” Miller says, “I’m sure that’s more or less what happened with the Passenger Pigeon.”

But like Rodewald, Miller also sees hope for the future in these eBird Trends maps.

“The coolest part for me about the [eBird Trends] data—nothing’s cool about the pattern, but about the data—is that now we’ve got place-and-time-based population estimates,” Miller says.

“We can study specific drivers [of population declines]. So now we can actually figure out what the hell is happening.”

For Burrowing Owls, Modest Increases Are Peppered in Among the Declines

The population center for Burrowing Owls is in the Great Plains, where eBird Trends data shows that local populations plummeted by as much as 30% to 50% between 2012 and 2022. Explore the full eBird Trends map for Burrowing Owl.

A Silver Lining

A silver lining in the results of the 2025 Science study could be the “heterogeneity” of the population trends, as the authors put it.

Of the 495 bird species in the analysis, 97% of species showed a mix of areas of decrease and areas of increase across their distributions. That is, while three-quarters of bird species were declining overall, almost all of those species had at least some places on the continent where local populations were increasing.

In the study there are hundreds of species “increasing where they’re at low abundance, even though they’re decreasing where they’re at high abundance,” says Johnston. “That shows us most of the species in those [low-abundance] communities do have potential for recovery.”

Rodewald says there may be more than one reason for the pattern of heterogeneity.

“Populations in marginal habitats may be occupied by individuals that are more tolerant to disturbance, or able to bounce back more quickly,” she says. “In some cases, those may be the populations that hold the most promise for recovery.”

Additionally, the places hosting increasing bird populations could be morphing into more suitable habitat of the future.

“For some species, we do see signs of shifting ranges … a shift in where species are most at home,” says Johnston. “The areas where they previously thrived are now areas where they can’t do so well, due to shifting changes in climate and weather and food resources. And the areas that previously might have been too cold, or limited for other reasons, are places where they’re moving into.”

Corina Newsome, a conservation scientist at the National Wildlife Federation, says that this fine-scale data on where birds are increasing is very helpful to NWF efforts to advance land-management decisions that benefit wildlife and people as both respond to climate change.

She says that NWF recently released a report entitled Innovation in Climate Adaptation that guides natural resource managers in adaptive management for biodiversity and ecosystems.

“Being able to see this heterogeneity in trends through this data gives us the tools to be adaptive,” Newsome says.

She also says that the eBird Trends data in urban areas can help NWF’s efforts to identify conservation opportunities in large metro areas that haven’t been prioritized as wildlife habitat in the past.

“Cities can actually be some of our hidden jewels when it comes to providing spaces for wildlife,” Newsome says. “Birds move. They move through cities even if they spend most of their time in the tropics in the middle of the rainforest. They might stop through Atlanta. They might stop through Chicago. And so we need to make sure that these [urban] landscapes are ones in which they can thrive.

“Data like this helps [NWF] be very measured and very specific in where we’re investing in the collection of more data and investing in habitat restoration. … It strengthens the case that we’re making for ensuring that vulnerable people and vulnerable wildlife [in cities] get the investments that they need to thrive in this changing climate.”

Now We Can Look for Lessons

For Amanda Rodewald, the ability to locate areas of local population increases among bird species is the finding in the 2025 Science study that makes it actionable science.

“Where we have these smaller pockets of increase, that would be a place where we can look for lessons as to what might be working on the ground,” she says. “What we need to do is smarter conservation … more strategic, more precise, and more accommodating of human activities when we make decisions. And the reality is it takes a ton of information.”

Andrew Stillman is a quantitative ecologist at the Cornell Lab who is on the front lines of getting that information out of eBird Trends models and into the hands of decision-makers at government agencies.

A New Way to View Declines—and Increases

Hundreds of bird species in North America have overall declining trends but smaller areas with increasing trends. For example, eBird Trends models show increasing local populations for Cerulean Warblers in the Ozark National Forest region, despite declines in the core of the bird’s range. Likewise, Red-cockaded Woodpeckers are increasing in the North Carolina Sandhills, at the northern edge of the bird’s range.

“Where we have these smaller pockets of increase, that would be a place where we can look for lessons as to what might be working on the ground,” says study coauthor Amanda Rodewald.

According to Stillman, who was also a coauthor on the 2025 Science study, natural resource agencies in the U.S. want to know the relationship between bird population trends and the things that they can control, such as land management. eBird Trends, says Stillman, is “a public tool that allows land managers to look at different scenarios, pinpoint the areas of opportunity, and see where an investment in conservation actions could benefit birds the most.”

Earlier this year, Stillman was part of a team of scientists at the Cornell Lab who put together customized eBird Trends reports for each of the two dozen Migratory Bird Joint Ventures organizations across North America. Joint Ventures are regional partnerships of federal agencies (such as U.S. Fish and Wildlife Service and U.S. Forest Service), state and provincial agencies, and local nonprofit groups such as Audubon or Nature Conservancy chapters. When Ashley Peele, the science integration coordinator for the Appalachian Mountains Joint Venture, saw the eBird Trends map for Eastern Whip-poor-will in her work area, she immediately saw opportunities that weren’t apparent before.

“The spatial resolution of it is, frankly, a game changer. … For a long time, the only data we had was for the entire geography. … And there’s a ton of nuance with what is going on for any given species across this massive geography,” Peele says, referring to the AMJV’s work area spanning from New York State down through Georgia and Alabama. “You don’t have to dig hard at the trends map for whip-poor-wills. It’s immediately apparent that there’s some very interesting things going on with whips in the central Appalachian region.”

She’s referring to the cluster of blue dots from eastern Kentucky through West Virginia to Virginia in the eBird Trends map for Eastern Whip-poor-will. “To me, I see a lot of opportunity,” Peele says. “Our joint venture has very strong partnerships in the areas where these [whip-poor-will] populations are doing better.”

According to Peele, the AMJV has long-standing relationships with state wildlife agencies, private industrial timberland owners, and landowners in the region’s coal fields.

“One of our key partners, West Virginia DNR, has been doing a tremendous job for many years now of acquiring these old mine sites and doing some tremendous reclamation work and habitat management work,” Peele says. “One of the things that’s a little tantalizing about this is, whip-poor-wills need a varied matrix of habitats on the landscape. … These reclaimed mine sites—or even some of the industrial timberlands, because of the way they do rotational harvest and cutting—actually create a matrix. In theory, that looks pretty appealing for what we know whip-poor-wills prefer in terms of their breeding habitat.”

Eastern Whip-poor-will was designated as a Tipping Point species in the 2025 State of the Birds report, which means it has lost half its entire population since 1970. eBird Trends indicates that whip-poor-wills declined 15% across their entire range in about the last 10 years. Peele’s hope is that central Appalachia could incubate a rebounding Eastern Whip-poor-will population to help act against the bird’s rapid declines in population centers in the Great Lakes region.

And she says, the AMJV can move fast.

“When we look at that map, we don’t just see the data,” says Peele. “We have this background knowledge and history on the partnership networks that exist there. We know what’s going on in that landscape in terms of land use. We can look at these maps and immediately start seeing opportunity and ideas.

“We can take this information and really do a lot with it quickly.”

To Andrew Stillman, the Cornell Lab quantitative ecologist, that’s the beauty of eBird Trends: “The information that sounded the alarm is the very same information we can turn around and use to solve the problems.”

Top banner photo credits (via Macaulay Library):

Top row: Black-bellied Whistling-Duck by Matthew Addicks; Northern Cardinal by Zachary Vaughan; Lawrence’s Goldfinch by Aidan Brubaker; Bald Eagle by Christian Hagenlocher; Franklin’s Gull by Jonathan Casanova.

Middle row: Least Tern by Alex Eberts; Eastern Towhee by Amber Miller; Red-cockaded Woodpecker by Craig Brelsford; Great Blue Heron by Gaetan Dupont; Cerulean Warbler by Fernando Burgalin Sequeria.

Bottom row: Florida Scrub-Jay by Davey Walters; Eastern Whip-poor-will by Daniel Hart; Eastern Bluebird by Micah Hale; Chestnut-sided Warbler by Frederik Gustavsson; American Robin by Ezra J. Campanelli.

Back to top