The alarming increase in global temperature observed over the last hundred years, driven by the use of fossil fuels, has prompted a shift towards â€śgreenerâ€ť energy production. An extensive expansion of wind power exploitation is expected in the coming years, which makes its effect on vulnerable species an issue of growing conservation concern. Among the wildlife affected by wind turbines, vultures are probably the most vulnerable avian ecological guild. They have experienced a sharp decline during the last decades and their survival in many areas is the result of targeted recovery and conservation actions. The bearded vulture (Gypaetus barbatus) represents an emblematic example. After having been extirpated from the European Alps, the species once again inhabits its former habitat, thanks to the massive long-lasting effort of a dedicated reintroduction programme. There are concerns, however, that the sprawl of wind turbines in the Alpine massif will jeopardise this successful population recovery. The main goal of this PhD thesis was therefore to predict areas in the Swiss Alps where conflicts between bearded vulture conservation and wind energy development are likely to occur, thus allowing for a more biodiversity-friendly spatial planning of wind turbines. Using a spatially explicit modelling framework with combined information of casual observations and GPS data, I predicted speciesâ€™ potential distribution as well as its flight behaviour in relation to landscape, wind, and foraging conditions. First, I investigated the species ecological requirements in relation to season and age and translated these into distribution maps covering the whole Swiss Alpine arc. Here the focus was on evaluating the ability of the models to predict the possible future expansion of the species, a crucial point for anticipating potential conflicts arising from the spread of wind energy. During this process, I secondly had to delve into methodological challenges, especially with regard to taking objective decisions for model tuning. Based on the example of modelling the distribution of the bearded vulture, I introduced a new genetic algorithm for hyperparameters tuning, which drastically reduces computation time while achieving a model performance comparable or equal to that obtained with standard methods. Moreover, I generalised the developed routines so as to make them applicable to the most common species distribution modelling techniques and compiled the solutions in an R package now available to the scientific community. Thirdly, I explored the flight height patterns of bearded vultures to identify key factors driving low-height flight activity and delineated areas where the species is likely to fly within the critical height range that is typically swept by the blades of modern wind turbines.
Overall, I found that food availability is an important driver of both distribution and low-height flight activity of bearded vultures. Habitat selection differed between seasons and between age classes during the cold season. While food availability and geological substrates were the main drivers of the distribution during the warm season, I observed a shift in the requirement of adult birds in the cold season, where habitat selection was mainly influenced by climatic conditions. This suggests that adult birds may be constrained by favourable winter conditions for the selection of breeding territories. Combining the ecological requirements of both age classes and seasons I found that 40% of the Swiss Alps offers suitable habitat for the species. The model trained with species data collected between 2004 and 2014 was able to accurately predict new breeding territories established in 2015 â€“ 2019, and thus adequately delineated areas where the spreading population will likely to occur in the future and where conflicts with wind energy development might arise. The flight-height analysis of the GPS-tagged birds revealed that bearded vultures mainly fly within the critical height range swept by the turbine blades (77.5% of GPS locations), which poses the species at high risk of collision. Flying at low heights most frequently occurred along south exposed mountainsides and in areas with a high probability of ibex (Capra ibex) presence, a key food source for bearded vulture. Synthesising the information on bearded vulture distribution with the flight height behaviour allowed identifying and mapping areas where the species is likely to fly at risky height within its habitat. This high resolution, spatially explicit information represents a valuable tool for planners involved in wind energy development as well as a first basis for detailed impact assessments, while the methodological framework I developed represents a transferable approach for scientists studying potential conflicts between the development of aerial infrastructure and other target organisms.