Novel approach to estimating avian mortality from vehicle–bird collisions on U.S. roads

Bird population declines are an important ecosystem issue, and mortality from human-derived sources—including collisions with vehicles—is a contributing factor. Current knowledge of bird collisions with vehicles is based on detections of roadside casualties by on-the-ground searchers. We present an estimate of annual vehicle–bird collisions (hereafter collisions) based on counts of collisions by drivers. From 30 December 2020 to 12 January 2022, drivers recorded 125 collisions and 2,265,126 driven km. Nationwide annual collision rates in the USA were estimated by multiplying our modeled per-km collision rates with the U.S. Department of Transportation’s annual vehicle mileage data. The resulting estimated annual collision rate for unpaved rural two-lane and paved rural two-lane roads combined was notably higher than for other categories: 60 collisions per million km driven (95% confidence interval [CI] 42–85) compared with paved rural multi-lane (12 [95% CI 6–25]) and urban/suburban (14 [95% CI 4–44]) roads. The predicted annual collision rate for U.S. Fish and Wildlife Service regions was highest in the Pacific (95 collisions per million km driven [95% CI 47–195]) and lowest in the Pacific Southwest (eight collisions per million km driven [95% CI 1–61]; no collisions were recorded in the Southeast [limited sampling]). We estimated total number of vehicle–bird collisions in the continental USA to be 96 million birds per year (95% CI 59–200). Our results are based only on light-duty trucks and sedans, and so could be refined with a more representative sample of vehicle types. Nevertheless, our nationwide estimates are useful to help prioritize broad-scale efforts to slow or reverse population declines. Our work suggests vehicle–bird collision risk is highest on two-lane roads and for small birds, and research to identify mitigation measures to reduce vehicle–bird collision rates might be most useful on two-lane roads.