Scorching heat waves, severe cold snaps, and other extreme weather events are limiting where species can exist — and affecting biodiversity estimates — for more than 500 North American bird species, according to a new, Yale-led study.
The issue is particularly acute in the Great Plains and southwestern regions of the United States, where species ranges and total species counts are significantly overestimated when variable and extreme weather are not considered, the study says.
The findings were published in the journal Global Change Biology.
“We all know extreme weather is on the rise with climate change,” said Jeremy Cohen, the study’s lead author. He is an associate research scientist in Yale’s Department of Ecology & Evolutionary Biology, a member of the lab of co-author and Yale professor Walter Jetz, and part of the Yale Center for Biodiversity and Global Change. “This is the first study to look at how increasingly variable and extreme weather affects avian biodiversity at a continental scale.”
Cohen, whose specialty is macroecology, uses large ecological datasets, remote sensing data, and species distribution models to study how climate change affects wildlife at continental and sometimes global scales.
For the study, the researchers created species distribution models that linked satellite and ground-based environmental data with observational data from eBird, a citizen-science database in which members of the public submit bird sightings. These models enabled the researchers to determine which factors drive the occurrence of each species in particular locations — and where each species was predicted to exist.
The researchers then compared these new models with existing models that did not include any environmental variability (i.e., weather averages) and models that included seasonal weather variability.
They found that the models factoring in temperature and drought extremes were more accurate — with habitat ranges for individual species that were much smaller than those based on models that didn’t take variable or extreme weather into account. The differences were most pronounced when predicting species distribution in the winter months, during which weather in North America is more variable.
The habitat ranges in models without extreme weather events were, on average, 10% too large for winter months and 6% too large for summer months, Cohen said.
The discrepancies for individual species were even more dramatic, in some cases.
Without weather variability or extreme weather included, range predictions for the hooded oriole (a songbird that breeds in the southwestern U.S.), for example, were about 30% higher than what is actually observed; range predictions for the eastern phoebe (a small bird species that winters in the southeastern U.S.) were about 20% higher; predictions for the varied thrush (a songbird that breeds in the Pacific Northwest and is susceptible to drought conditions) were about 10% higher.
Such detailed information about the risk of severe weather conditions has not been widely available until recently, the researchers explained.
“Unfortunately, extreme weather events are becoming ever more frequent and forcing both us humans and animals to adapt,” said Jetz, a professor of ecology and evolutionary biology in Yale’s Faculty of Arts and Sciences. He is also director of the Yale Center for Biodiversity and Global Change.
“Our work shows that accounting for this irregular part of climate is now central to understanding the distribution of biodiversity and its change,” Jetz said.
Having precise models for bird ranges is essential for providing the best information for conservation efforts and preserving biodiversity, Jetz and Cohen said. Although certain U.S. localities have a wealth of good data about species ranges and populations, there are large expanses throughout the country where data is sparse. For those areas, accurate bird range models are crucial for habitat conservation programs, they said.
“Extreme conditions define the boundaries where organisms can exist,” Cohen said. “Seasonal weather averages alone may not tell the full story.”
Co-authors of the study are Shubhi Sharma, who earned her doctorate at Yale earlier this year; Frank La Sorte, a senior scientist at the Yale Center for Biodiversity and Global Change; and Diego Ellis-Soto, a 2024 Yale doctoral graduate now at the University of California at Berkeley.
Support for the research came from NASA grants and the E.O. Wilson Biodiversity Foundation in furtherance of the Half-Earth Project.
Continental-Scale Biodiversity Predictions Are Influenced by Climatic Variability and Extreme Weather
8-Aug-2026