Flight behaviour of Red Kites within their breeding area in relation to local weather variables: Conclusions with regard to wind turbine collision mitigation.

Birds and bats are prone to collisions with wind turbines. To reduce the number of bat collisions, weather variables are commonly used to shut down wind turbines when a certain constellation of weather variables occurs. Such a general approach might also be interesting to mitigate raptor collisions. Studies on the relationship between flight behaviour and weather variables are needed. To investigate the flight behaviour of raptors within their breeding area in relation to local weather variables, we used high resolution data of flight tracks of Red Kites collected on a wind energy test site (Germany). Birds were tracked with a laser range finder (LRF) or with Global Positioning System (GPS) transmitters. Weather variables were continuously registered on site. We used generalised linear mixed models to analyse the influence of weather variables and of the measurement method on different flight parameters. Furthermore, we investigated the probability of flying within a virtual rotor height range defined by three hub heights (84, 94 and 140 m; diameter: 112 m). The median flight altitude measured by LRF (52.5 m, 95% CI: 44.9–61.0, N = 2511) was on average 25 m higher than the corrected one resulting from GPS (27.8 m, 95% CI: 24.7–31.2, N = 6792). Flight speed also differed between methods (GPS: 29.2 km/h, 95% CI: 28.2–30.3 km/h; LRF: 25.1 km/h, 95% CI: 24.0–26.3 km/h). The effects of the weather variables were weak. Birds tended to fly less and lower during wet (humid, rainy or foggy) than dry weather, and lower during strong than weak winds. Probabilities of flying within a height range of virtual rotors increased with decreasing hub height, and hence ground clearance. Synthesis and applications: Flight behaviour was highly variable. Flights occurred during all weather conditions at different altitudes throughout the day over the entire season. Further research into the relationship between flight behaviour, weather variables, collisions and other factors is needed as a basis for developing shutdown regimes generally suitable for raptors. The mean flight altitude and speed differed between the measurement methods. Any values resulting from studies should be interpreted in the context of the method. [ABSTRACT FROM AUTHOR]