Habitat, human pressure, and social behavior:: Partialling out factors affecting large-scale territory extinction in an endangered vulture

Extinctions are often the result of multiple factors that are difficult to disentangle and so methods for identifying simple and combined types of extinctions are valuable for both basic and applied ecology. We applied a modification of variance partitioning and a hierarchical partitioning analysis to test several hypotheses that attempt to explain the recent large-scale disappearance of Egyptian Vulture breeding territories in Spain. Our aim was to identify and then separate the simple (or pure) from the combined effects of habitat features, human pressure, and the social behavior of the species on the risk of extinction from a territory while controlling for spatial autocorrelation. Deviance partitioning showed that a complex mix of factors is significantly related to the disappearance of more than 400 territories throughout Spain. Abandoned territories were located in areas that are isolated from other conspecific territories and far from communal roost sites. In addition, these territories were found in places where there is a lack of natural habitats, high habitat fragmentation and reduced habitat diversity, and where food availability seems to be low and the illegal use of poison to control predators is a common practice. Deviance partitioning also showed an important spatial component in the probability of extinction. Abandoned territories were not randomly distributed; rather, they were aggregated in extinction 'hotspots', mainly related to food availability and human pressure. Deviance partitioning turned out to be an useful tool for identifying the relative contribution of a variety of factors - and their combined effects - associated with an extinction process. The deviance explained by each factor must be interpreted, however, in the context of a good knowledge of the life history of the species. Hierarchical partitioning can help rank conservation priorities and, by using as an objective criterion the relative weight of each independent variable that could be effectively managed for conservation, may provide wildlife managers with a means of saving funds and optimizing action plans. (C) 2006 Elsevier Ltd. All rights reserved.