Unravelling vulture avoidance tactic of wind turbines combining empirical and simulation data

The increase of wind turbine installations to limit climate change may affect bird populations because of collisions with rotor blades. Birds may respond to wind turbine presence along a gradient of behavioural changes: avoiding the wind farm (macro-scale) or only the wind turbines either by anticipating wind turbine locations (meso-scale) or engaging into last-minute flee attempts after late perception (micro-scale). We investigated the flight response at these three spatial scales of 25 adult griffon vultures (Gyps fulvus) equipped with GPS tags over three years when flying in an area including ten wind farms in the Causses, France. At macro-scale, the population foraging range and habitat use revealed that vultures did not avoid wind farms. To investigate avoidance at meso- and micro-scales we focused on the four mostly visited wind farms. We compared vulture flights to null movement models, based on a method allowing us to keep the correlation between flights and topography while creating movement independent of wind turbine locations. At most sites, vultures did not show avoidance behaviour. Yet, simulations from our agent-based model highlighted that the avoidance pattern detected at one wind farm matched with an anticipated avoidance of turbines, probably linked to the presence of a ridge nearby. Overall, our results suggest wind farm-specific responses by soaring birds as a function of the landscape topography. Thus, stakeholders should carefully consider the wind farm location for siting and designing preventive measures (e.g. improve detection of species not able to avoid turbines in switching off on-demand technologies) to reduce collision risk of soaring birds.