When towering saguaro cactuses in the Sonoran Desert erupt spectacularly into springtime bloom, winged flashes of red and yellow dart among them. Gilded Flickers flit between flowers like overgrown hummingbirds, lapping up ants and nectar with long, sticky tongues. These large woodpeckers’ lives are wrapped up in those of the saguaro; not only do they dine on insects living in the succulent flesh, but they also hammer out cavities that protect their families from the harsh sun. When they vacate the hollows, Elf Owls, Purple Martins, and Brown-crested Flycatchers move in.

But the Sonoran’s avian real estate developers could go extinct by 2080 if the region grows hotter and drier, creating prime conditions for wildfires that incinerate the flickers’ homes. It’s just one of the sobering findings from Audubon’s report, Survival by Degrees: 389 Bird Species on the Brink, published in October. Using advanced climate models, the report gives an unparalleled view into which birds, and which places, are at greatest risk, providing our best field guide to the future of North American birds.

Under the worst-case scenario, the changes to that future field guide are stark. The shorebird section would be considerably slimmer, since Arctic breeders like Sanderlings and American Golden-Plovers would have largely disappeared, unable to reproduce once temperatures climb too high. Pages of grassland birds such as Baird’s Sparrows and Long-billed Curlews, and woodland specialists like Acorn Woodpeckers and Clark’s Nutcrackers, would be removed as the Great Plains dry out and western forests decline due to drought and wildfire. The artwork would be less colorful with the loss of Blackburnian and Magnolia Warblers and other neotropical migrants that have for millennia flocked to the boreal forest each spring. 

In this speculative book, the range maps of species distribution would morph considerably. Some would balloon: Barn Swallows, European Starlings, Downy Woodpeckers, and Northern Mockingbirds would blanket far more of the continent, while Great Egrets and Great Blue Herons would expand north into areas previously too cold. But the maps on many pages would cover substantially less ground than they do today. 

Of the 604 species modeled, 389 are vulnerable to extinction, meaning that as soon as 2080 more than half of their current range would become inhospitable and they wouldn’t gain new ground. If that comes to pass, that future guide would be half the thickness of the one now on your shelf. Ninety-nine percent of birds could have to cope with more frequent extreme-weather events, like intense spring heat and heavy rainfall; at the same time, sea-level rise and urbanization could consume much-needed habitat. Most birds will likely experience multiple, compounding threats—unless we curb emissions and prioritize conserving the areas, identified by the models, that will be critical to climate-threatened birds. 

The models, the culmination of five years of research, are “cutting-edge,” says Josh Lawler, an ecologist at the University of Washington in Seattle who uses similar models to predict how wildlife might respond to climate change; he was not involved in the study. “Not only did they use new climate data, but they did things in about as sophisticated a way as you can for this kind of modeling.” Such geographically broad and species-rich studies are needed to help us help wildlife survive climate change, he says. “There is time to identify the places to protect, the places to restore, the species that are going to need the most help.” 

The report tells us that the time to act is now and points at the strides that could make the difference between survival and extinction for Gilded Flickers and myriad other species. 

A Model for These Times

The new study comes five years after Audubon published its first climate report, an investigation that was itself seven years in the making. The 2014 study found that of the 588 bird species investigated, 314 could potentially face extinction by 2080. 

At the time, it was the broadest and most detailed study of its kind. The scientists reached their startling conclusion by first combing 44,000 records from the Audubon Christmas Bird Count and North American Breeding Bird Survey to determine where birds live. Then they pinpointed the range of temperatures, amount of rainfall, and other climate characteristics of each species’ habitat and plugged the information into computer projections of the global climate. Out came potential future ranges with suitable climate conditions for each species, mapped to a resolution of 10 square kilometers.

Since then, huge advances have occurred in computing and access to international databases has grown, allowing scientists to generate even more detailed forecasts. “We are learning every day how to improve models,” says Terry L. Root, a retired biologist who pioneered computerized avian distribution models in the 1980s; she sits on the National Audubon Society board. “And we are just getting better.”

To more precisely define the climate conditions species inhabit today, Audubon scientists incorporated a whopping 140 million bird observations from 70 datasets. (If you submit sightings to eBird, you likely contributed to the new study.) They included data from Mexico and Canada that weren’t used in 2014. They also went beyond simply considering climate conditions by factoring in vegetation, agriculture, surface water, and other variables in order to ensure that birds would have fitting habitat, too. “We want to get at the areas that are not only suitable based on climate, but also have the appropriate habitat—they’re not converted to agriculture or a city or something that’s not appropriate for that particular species,” says Brooke Bateman, Audubon’s senior climate scientist who led the research. 

Bateman and colleagues fed those variables into climate models to determine whe