Low migratory flight altitudes may explain increased collision risk for Scolopax minor (American Woodcock)

Understanding bird migration at low altitudes is critical to evaluating risk of collision with obstacles. Recent advances in satellite tracking technologies allow quantifying use of low altitudes by small migrating birds with a high level of precision, allowing species-level inference into potential collision risk based on flight altitude. The American Woodcock (Scolopax minor) is suspected to be a low altitude migrant due to its frequent collisions with buildings, and subsequent mortality during migration may be contributing to population declines. We investigated migratory flight altitudes using satellite transmitters deployed on woodcock in 2020–2024 and examined how flight altitudes compare to the heights of common airspace obstacles. Each transmitter recorded a nocturnal GPS location with an altitude reading every 1–3 days during fall and spring migrations. We implemented a Bayesian hierarchical mixture model to identify whether locations were recorded on the ground or during flight, isolate measurement error, and describe the distribution of flight altitudes. We found that migrating woodcock fly at mean altitudes of 379 m above ground level, flying higher during spring (mean: 444 m, 95% credible interval: 333–578 m) than fall (338 m, 95% CRI: 267–423 m). Woodcock flight altitudes were frequently lower than could be observed using weather radar (27% of observations), and 57% of observations fell within the altitude range of ≥1 airspace obstacle. Our results suggest that woodcock fly at altitudes lower than reported for most nocturnal migrants, which likely contributes to their vulnerability to obstacle collisions. Woodcock provide an example of how vulnerability to obstacle collisions during nocturnal migratory flights are often species-specific, and mitigation efforts should incorporate measures aimed at reducing collisions during both diurnal stopovers and nocturnal migratory flights to effectively reduce bird collision mortality.

Methods
Data Collection and Preprocessing
We collected woodcock locations with altitude readings from 2020 to 2024 using GPS transmitters as a part of a larger collaborative effort by the Eastern Woodcock Migration Research Cooperative (Blomberg et al. 2023, Clements et al. 2024, Fish et al. 2024). We captured woodcock at 100 sites across the eastern portion of their range, including Alabama, Florida, Georgia, Louisiana, Maine, Maryland, New Jersey, New York, North Carolina, Nova Scotia, Ontario, Pennsylvania, Québec, Rhode Island, South Carolina, Vermont, Virginia, West Virginia, and Wisconsin. We caught woodcock using a combination of spotlighting and mist netting (McAuley et al. 1993). We aged and sexed birds upon capture, where we classified birds undertaking their first fall and spring migrations as juveniles, and all other birds as adults. We then attached 4–7 g PinPoint transmitters (Lotek Wireless Inc., Newmarket, Ontario, CA) using a rump-mounted leg loop harness (Fish et al. 2024).

We programmed transmitters to collect locations every 1–3 days during migration, with locations alternating between diurnal (1300–1500 hours Eastern Time) and nocturnal (0000–0100 hours) times. Transmitters recorded time, latitude, longitude, and GPS-derived altitude above the WGS84 ellipsoid, and transmitted data back to the ARGOS satellite constellation after every third location. We subset these locations to include only those within the migratory classification dataset produced by Berigan (2024). This dataset classified individual locations as migratory or non-migratory based on the assumption that migration starts after the first ≥16.1 km movement and ends after the final ≥16.1 km movement of the season. We used ArcGIS Pro 3.2.1 (ESRI 2024a) to calculate the difference between the altitude and orthometric elevation recorded for each location (ESRI composite elevation layer; ESRI 2024b), providing a measurement of altitude above ground level for each point.

We classified data for our models based on prior descriptions of woodcock activity patterns. Woodcock are ground-feeding birds that rarely fly outside of crepuscular hours (Rabe et al. 1983). When rare diurnal flights do occur, these are generally brief, comprising 1–3% of diurnal time budgets, and close to the ground (McAuley et al. 2020). We therefore made a modeling assumption that all diurnal locations could be treated as though they were known to be recorded on the ground (hereinafter “known ground locations”). As woodcock are nocturnal migrants, we define potential flight locations as all points that were nocturnal, occurred during migration based on the classification in Berigan (2024), and were preceded and followed by >6.68 km steps (defined as lines connecting consecutive locations). The 6.68 km threshold was based on the 99th percentile of step lengths recorded within a stopover site (Berigan 2024). Ensuring that the preceding and following steps were >6.68 km increased the likelihood that the bird had moved away from a stopover site before the point was recorded.

References
Berigan, L.A. (2024). Full annual cycle analysis of American Woodcock (Scolopax minor) distribution, habitat use, and migration ecology. PhD dissertation. University of Maine, Orono, Maine.

Blomberg, E. J., A. C. Fish, L. A. Berigan, A. M. Roth, R. Rau, S. J. Clements, G. Balkcom, B. Carpenter, G. Costanzo, J. Duguay, C. L. Graham, et al. (2023). The American Woodcock Singing Ground Survey largely conforms to the phenology of male woodcock migration. Journal of Wildlife Management 87:e22488.

Clements, S. J., L. A. Berigan, A. C. Fish, R. L. Darling, A. M. Roth, G. Balkcom, B. Carpenter, G. Costanzo, J. Duguay, and K. Filkins (2024). Satellite tracking of American Woodcock reveals a gradient of migration strategies. Ornithology141:ukae008.

ESRI (2024a). ArcGIS Pro. Redlands, CA, USA.

ESRI (2024b). Terrain. [Online.] Available at https://www.arcgis.com/home/item.html?id=58a541efc59545e6b7137f961d7de883.

Fish, A. C., A. M. Roth, G. Balkcom, L. Berigan, K. Brunette, S. Clements, G. Costanzo, C. L. Graham, W. F. Harvey, M. Hook, D. L. Howell, et al. (2024). American woodcock migration phenology in eastern North America: implications for hunting season timing. Journal of Wildlife Management 88:e22565.

McAuley, D. G., D. M. Keppie, and R. M. Whiting Jr. (2020). American Woodcock (Scolopax minor), version 1.0. In Birds of the World (A. F. Poole, Editor). Cornell Lab of Ornithology, Ithaca, NY, USA.

McAuley, D. G., J. R. Longcore, and G. F. Sepik (1993). Techniques for Research into Woodcocks: Experiences and Recommendations. Proceedings of the eighth American woodcock symposium. U.S. Fish and Wildlife Service, p. 5.

Rabe, D. L., H. H. Prince, and E. D. Goodman (1983). The effect of weather on bioenergetics of breeding American woodcock. Journal of Wildlife Management 47:762–771.