Abstract

1. Wind energy development poses collision risks to birds, yet species-specific vulnerability assessments incorporating measurement uncertainty, flight behaviour, and environmental conditions remain scarce for shorebirds.
2. We conducted the first collision vulnerability assessment for any Patagonian shorebird using GPS telemetry and a novel probabilistic approach incorporating GPS altitude measurement uncertainty. We analysed 237 flight observations from 15 Two-banded Plovers (Anarhynchus falklandicus) at Bah & iacute;a Lomas, Chilean Patagonia, assessing collision vulnerability as the proportion of flights within the rotor swept zone under four turbine configurations. We used beta regression models incorporating GPS altitude uncertainty, flight behaviour, and meteorological conditions (wind speed and direction from ERA5 reanalysis) to identify factors influencing vulnerability.
3. Collision vulnerability ranged from 42.6% (low ground clearance: 17.34 m) to 24.6% (high clearance: 30.00 m). The best-supported models included flight speed, solar elevation, and wind direction, but not wind speed. Vulnerability varied with wind direction, peaking during winds from the northwest quadrant where median flight altitudes (similar to 30 m) coincided with the lower rotor swept zone, and was lowest during prevailing southwesterly winds when flights remained at low altitudes (median similar to 7 m). Incorporating GPS altitude uncertainty revealed that traditional binary classification underestimates collision vulnerability by 22-44%.
4. Synthesis and application. Our probabilistic framework incorporating measurement uncertainty and meteorological covariates provides a template for collision vulnerability assessment applicable across taxa. Increasing turbine ground clearance reduced vulnerability by 42%. The influence of wind direction on flight altitude-and thus collision vulnerability-highlights the importance of incorporating meteorological data in collision assessments, particularly in regions with strong prevailing winds.