Strategies to identify dangerous electricity pylons for birds

Bird electrocution rates in Secanos de Lerida, an important bird area in central Catalonia (northeast Spain), were estimated based on 804 visits to 507 electric pylons between 1995 and 1999. Electrocution caused a minimum of 160 victims on 67 pylons. Victims were corvids (36%), diurnal birds of prey (60%) and owls (4%). Electrocution rates ranged between <0.01 birds pylon(-1).year(-1), in the less risky areas, and 0.20 birds.pylon(-1).year(-1), in the areas with higher risk. The number of electrocution victims per killing pylon ranged from 1 to 12. Casualties occurred in a fairly contagious pattern, since 50% of deaths took place on only 2% of the visited pylons, and 50% of the killing pylons accumulated 80% of the victims. Univariate analysis revealed that the technical design of the pylons was very important in determining the potential danger of electrocution (metal crossbows were found to be the most dangerous design, followed by earthed flat pylons and vaults). Most mortality (97%) could be eliminated if all technically dangerous pylons were modified, but these would entail 67% of the pylons in the study area. Modelling the presence of carcasses under the pylons was used to identify the pylons which concentrate casualties, so that a 'preferred pylon' approach could be used to allocate mitigation resources. A single under-line inspection identified 78% of the pylons causing bird casualties in the area, which were responsible for 91% of the deaths. Information collected during this single inspection was used to build a logistic model which allowed the correct classification of most of the dangerous pylons missed in the first inspection. This approach revealed that geographical location and habitat setting were as important as technical design in determining the actual risk of electrocution. In that way, up to 99% of mortality can be eliminated by modifiying only 23% of the pylons in the area.