Where is the risk? Highlighting differences in vulture behaviour between dangerous and safe wind turbines

Abstract Spain is the second European producer of wind energy and hosts approx. 30.000 breeding pairs of griffon vultures (Gyps fulvus). Each autumn 4000-5000 juveniles cross the Strait of Gibraltar to reach their wintering quarters in Africa, flying through an area with a high density of wind farms and exposing them to collide with turbines. The aim of this study is to determine whether flying behavior of vultures in the nearby surroundings of the turbines is affecting collision risk. From the 8th of October 2012 to the 23rd of November 2012 two observers carried out 123 hours of simultaneous surveys at 10 safe and 10 dangerous turbines in La Janda, Cádiz (southern Spain). Dangerous turbines were selected on the basis of the number of killed griffons per month, selecting those with the highest ratio, while safe turbines had never killed a griffon from their setting up. Observers recorded the number of flying birds, their flight direction, the type of flight (flapping or soaring), if they were at risk of collision with turbines or not, wind direction and intensity. Pearson’s correlation analysis revealed a positive correlation between the number of griffons watched at dangerous turbines and historical record of vultures killed in these turbines in previous years and a positive correlation between the number of risky situations and the historical record of vultures killed in these turbines. According to Wilcoxon’s test, more vultures were recorded at dangerous turbines at noon and in the afternoon. However, significant differences in the number of risky situations have been found only in the afternoon, probably due to small sample size. The number of observed vultures in every wind turbine was strongly correlated with the number of risky situations, both in dangerous and in safe turbines. Although Fisher’s exact test showed significant differences in the number of flapping birds that crossed the turbines with eastern and western wind, neither significant differences were found in the proportion of birds at risk in the two groups, nor in the number of risky situations with respect of wind direction. We conclude that the higher number of risky situations recorded at dangerous turbines seems to be more a density-dependent phenomenon rather than something linked to bird behavior. Analysis have been hampered by the low number of dangerous situations recorded and the short sampling period, hence further observations are needed. However the achieved results may provide appropriate management tools for managing mortality in wind farms by optimizing the allocation of mitigation measures with surveys, and may also help to the suitable design of future wind farms