In 2022–2024 breeding bird surveys, buteo hawks and vultures fell below trend, but waterfowl and gulls did not, and H5N1 detections did not predict which species or routes declined

Question: does the 2021–2024 H5N1 HPAI panzootic show in the standard breeding bird surveys that measure common bird trends? Data: North American Breeding Bird Survey 1990–2024 (USA 420 species, Canada 255) and UK Breeding Bird Survey 1994–2024 (139 species); exposure from USDA APHIS HPAI-positive dead or sick wild birds (2022–2024) and Defra positive findings (2021–2024). Method: for each species, a Poisson annual index with route-and-observer effects; the 2022–2024 index was compared with the linear 2011–2021 trend and calibrated against 18 placebo break years (2000–2017), with route bootstrap intervals. Results: across species, the number of positive birds did not predict the anomaly (USA Spearman rho = -0.07, p = 0.17; UK rho = +0.03, p = 0.75). Geese, swans, ducks and gulls, which had most detections, stayed within their placebo ranges (Canada Goose USA +1%, 95% interval -11% to +15%; Mute Swan UK -5%), but the surveys can only detect declines of about 20–40% for these groups. Open-country buteos and vultures fell below trend in all surveys: Red-tailed Hawk -15% (USA) and -30% (Canada), Swainson's Hawk -23% and -27%, Common Buzzard -12% (UK), Turkey Vulture -12% (USA). US raptors and vultures had lower placebo-calibrated anomalies than other species (mean z -0.53 vs +0.12, permutation p = 0.006). But within species, routes within 50 km of detections did not decline more than routes over 100 km away (mean difference +1.1% +/- 2.9%, 19 species); the Red-tailed Hawk drop was largest in prairie, western and northern regions and absent in the Northeast, where detections per bird were highest; and drier routes (PDSI 2021 to mid-2023) declined more for most species (median -6%) and for Red-tailed Hawk (-10%). A West Nile virus positive control (American Crow, -14% after 2002) gave only a moderate placebo-calibrated signal. Conclusion: the surveys show a real 2022–2024 drop in some hawks and vultures, but its pattern across species and places does not support assigning it to H5N1; drought explains part of it. Limitations: observational design with one common break year, detection data that reflect surveillance effort, no Canadian exposure data, sensitive UK species absent, and raptor-group and drought tests chosen after first results.

highly pathogenic avian influenza · H5N1 · Breeding Bird Survey · population trends · raptors · Red-tailed Hawk · Common Buzzard · vultures · waterfowl · placebo test · difference-in-differences · drought · West Nile virus · wildlife disease