Canadian Wildfire Smoke May Hush Songbirds, Skew Annual Counts

When Canadian wildfire smoke turned the sky over New York orange in June 2023, people noticed the color. Birdwatchers noticed something stranger. Warblers stopped turning up. Not in one park or one county, but across a state’s worth of checklists, the small forest songbirds that birders find almost entirely by ear went missing from the record.

An analysis of nearly 99,000 birdwatching lists suggests those birds may have gone quiet. Researchers at the University at Buffalo and the Cornell Lab of Ornithology matched three breeding seasons of amateur sightings against daily air pollution readings from government monitors across New York State. Of the 84 species common enough to analyze, 55, roughly 65%, showed a statistically detectable link between smoke pollution and how likely a birder was to record them. Forty species got harder to find. Fifteen got easier.

Whether those 40 species actually left, or simply stopped announcing themselves, is a question the study cannot answer. Either way, it lands on a problem with real teeth: North America tracks its bird populations by counting what people see and hear in spring and summer, which is now exactly when the smoke arrives.

Wildfire Smoke and the Missing Warblers

Among species with a clear link, the losers skewed heavily toward the woods. Blackpoll, Black-throated Green, Black-and-white, Black-throated Blue, Magnolia, and Chestnut-sided Warblers all became less likely to be logged as fine particle pollution rose. So did Wood Thrush, Veery, Swainson’s Thrush, Hermit Thrush, Blue-headed Vireo, and Ruby-crowned Kinglet. Sorted by habitat, forest and shrubland birds skewed negative almost across the board.

What unites that list is not size or diet but acoustics. Small, drab, and hidden in dense summer foliage, those birds are located by song, not sight. A warbler that stops singing is effectively invisible. Prior research has found that animals do quiet down in smoke: gibbons in Borneo, orangutans, and grassland birds in New York have all shown reduced vocal activity during and after smoke events. Earlier soundscape research the authors cite found that smoke-driven changes in the air can mask bird calls and lower acoustic detection rates, and cut visibility on top of that.

Birds that hunt in open air moved the opposite direction. Barn Swallow and Chimney Swift trended positive. Osprey showed the strongest positive association of any species in the study. Herons, egrets, cormorants, gulls, and Rock Pigeon nudged upward too. Those species get spotted visually, so a scrambled soundscape costs them nothing, and the authors point to earlier work suggesting fire and smoke can alter insect activity, potentially increasing airborne prey and, with it, visible hunting.

Twenty-nine species showed no dependable link at all, including American Robin, Northern Cardinal, Mourning Dove, and Common Grackle. Roughly a third of the birds, in other words, appeared to carry on regardless.

Reading 99,000 Birdwatching Checklists

Wildfires are rare in New York. Roughly 2,100 acres burn in a typical year, a rounding error next to the 7.2 million acres that burn annually across North America. What New York gets instead is other people’s smoke, carried down from boreal forests in Canada. During June and July of 2023, that imported haze pushed daily average fine particle pollution past 120 micrograms per cubic meter in parts of the state, a level the paper describes as “eight times the value deemed unsafe for humans,” or eight times the World Health Organization guideline of 15 micrograms per cubic meter over 24 hours, and the dirtiest air recorded in New York in more than half a century.

Fine particle pollution, written as PM2.5, means airborne specks 2.5 micrometers across or smaller, small enough to slip deep into lungs. It serves as a standard stand-in for wildfire smoke because smoke is loaded with it, and because the Environmental Protection Agency already measures it hour by hour nationwide.

Bird data came from eBird, the Cornell-run program where hobbyists log what they see and hear. Lead author Festus O. Adegbola and colleagues pulled 98,960 complete checklists submitted by 9,838 observers during the May-through-August breeding seasons of 2021, 2022, and 2023. Each list was matched to the nearest air quality monitor, and lists logged more than 25 kilometers from a monitor were dropped, which is why the data cluster around cities and populated corridors rather than blanketing the state.

Casual checklists are not scientific surveys, and the team worked hard to compensate. Their models adjusted for habitat, temperature, humidity, time of day, day of year, year, how long the birder searched, how far they walked, and the identity of the birder, since a veteran with a good ear finds more birds than a beginner. Only species recorded on at least 600 checklists in 2021 made the cut, leaving 84.

Wildfire Smoke Could Be Skewing Bird Population Counts

Detection probability sounds like a technical footnote. It is the whole ballgame for conservation. Long-running efforts such as the Breeding Bird Survey and the Monitoring Avian Productivity and Survivorship program run in spring and summer, and they measure bird populations by how often observers encounter birds. If smoke quiets warblers for three weeks, a smoky June looks statistically identical to a June when those warblers are simply gone.

Adegbola’s team put the risk plainly, writing that failure to account for particle pollution “could lead to spurious apparent declines in forest songbird detections during smoke years, or to apparent increases in aerial insectivore detections that may reflect behavioral shifts rather than genuine population changes.” Their fix is unglamorous: treat air quality as a routine survey variable, the same way ornithologists already adjust for rain, wind, and hour of the morning.

Honesty about a competing explanation runs through the paper. Birders are also animals that respond to smoke. On a bad air day, a person may skip the four-hour forest hike and scan a wetland from the car window instead, which would inflate detections of open-country species without a single bird changing its behavior. Adegbola and colleagues flag that some positive associations “may partly reflect changes in birder behavior rather than bird behavior,” and note that controls for observer identity and search effort cannot fully account for where people choose to go.

Nothing here proves smoke harms birds. Detection measures whether a person recorded a bird, not whether it was healthy, or breeding, or even present. Separate studies have linked smoke exposure to body mass loss in wild birds and to altered migration routes in geese, so the physiological case is building elsewhere. This study measures something narrower and more unsettling: how thoroughly a few smoky weeks can distort the record itself.