Towards a better understanding of avian collision in wind energy facilities using automatic detection systems

Abstract The rapid expansion of wind power energy has direct negative impacts on biodiversity, notably on avifauna through collisions with turbines. A better understanding of the collision causes is key to improving mitigation efforts. Collisions are the result of a combination of environmental factors that increase bird sensitivity and/or exposure to collisions. To date, potential risk factors have mostly been assessed individually, in few species of interest and/or at small spatiotemporal scales, despite the multifaceted nature of collision risk. To fill this gap, we used for the first time data from automatic detection systems (optic systems that automatically detect and monitor birds in the vicinity of wind turbines) to simultaneously assess the effects of behavioural and environmental factors on bird sensitivity (here, estimated as the bird presence in the risk zone) and exposure (here, estimated as the frequency with which birds use the zone). We analysed 205,867 bird trajectories recorded between 2018 and 2023 in 11 French wind energy facilities. We obtained results similar to previous studies relying on other methods (GPS, direct observations). Results suggest that bird sensitivity was higher during periods of high bird activity (first hours of daylight and migrations). They also suggest that sensitivity and exposure may increase in conditions that reduce the birds' visual perception of turbines (high nebulosity, low visibility and low rotor speeds) and in conditions that may influence birds' flight height (high temperatures and high wind speeds). We found a nonsynchronicity of exposure and sensitivity peaks, highlighting the importance of considering both drivers of risk when investigating the collision risk. However, our results show a high variability between species, flight behaviours and sites that should be addressed in the future to clarify the relationships between collision risk, birds' visual perception and behaviour. Policy implications . Data from automatic detection systems are a promising non‐invasive approach that requires few human and logistical resources to develop a more comprehensive bird collision risk mitigation strategy. They can be valuable to biodiversity stakeholders to highlight environmental factors that locally increase bird exposure and sensitivity, to avoid installing turbines in sensitive areas or to adjust detection system settings.