Disentangling drivers of power line use by vultures: Potential to reduce electrocutions

Abstract

Accidents on power lines are the leading cause of mortality for many raptor species. In order to prioritise corrective measures, much effort has been focused on identifying the factors associated with collision and electrocution risk. However, most studies lack of precise data about the use of pylons and its underlying driving factors, often relying on biased information based on recorded fatalities. Here, we used multiple years of high-resolution data from 49-GPS tagged Canarian Egyptian Vultures (Neophron percnopterus majorensis) to overcome these typical biases. Birds of our target population use electric pylons extensively for perching (diurnal) and roosting (nocturnal), so accidents with these infrastructures are nowadays the main cause of mortality. Predictive models of pylon intensity of use were fitted for diurnal and nocturnal behaviour, accounting for power line, environmental, and individual vulture's features. Using these measures as a proxy for mortality risk, our model predictions were validated with out-of-sample data of actual mortality recorded during 17 years. Vultures used more pylons during daytime, but those chosen at night were used more intensively. In both time periods, the intensity of use of pylons was determined by similar drivers: vultures avoided pylons close to roads and territories of conspecifics, preferentially used pylons located in areas with higher abundance of food resources, and spread their use during the breeding season. Individuals used pylons unevenly according to their sex, age, and territorial status, indicating that site-specific mitigation measures may affect different fractions of the population. Our modelling procedures predicted actual mortality reasonably well, showing that prioritising mitigation measures on relatively few pylons (6%) could drastically reduce accidents (50%). Our findings demonstrate that combining knowledge on fine-scale individual behaviour and pylon type and distribution is key to target cost-effective conservation actions aimed at effectively reducing avian mortality on power lines.