Using species distribution modelling to predict bat fatality risk at wind farms

Bats are often killed in large numbers at wind farms over a wide geographical range. We aimed to predict which areas conferred higher fatality risks to bats at wind farms. In an innovative approach, we combined species distribution modelling with mortality data and the ecological conditions at wind farms located in Portugal. We then generated predictive models to determine areas of probable mortality and which environmental factors were promoting it. Mortality data for four bat species, Hypsugo savii, Nyctalus leisleri, Pipistrellus kuhlii and Pipistrellus pipistrellus, were used. These experienced the highest levels of fatalities at wind farms in Portugal, comprising 290 of the 466 fatalities recorded from 2003 to 2011. The mortality risk models showed robust performances. Wind farms sited at humid areas with mild temperatures, closer than 5 km to forested areas and within 600 m of steep slopes showed higher probabilities of mortality. High mortality risk areas also overlapped highly with the potential distribution of N. leisleri in Portugal, suggesting that populations of this species may be at high risk due to wind farm fatalities. Moreover, a large extent of the area predicted to be a hotspot for mortality (i.e. areas likely to confer high mortality risk for four species) overlaps with sites highly suitable for wind farm construction. In summary, the approach used in this study could be paradigmatic for the development of important pre-emptive conservation measures for bat populations by identifying mortality risk in areas prior to wind farm installation and determining which conditions promote mortality. (C) 2012 Elsevier Ltd. All rights reserved.