Human density as an influence on species/area relationships: double jeopardy for small African reserves?

Small reserves are especially likely to lose species. Is that because the reserves are small, or because small reserves are located in especially adverse landscapes? It seems that the question has rarely, if ever, been asked. Data on reserve size and location in Africa, and calculations of local (within 50 km) mean human densities from available census records per province per country were the database here used to answer the question. IUCN grade I and II reserves in Africa are located across the range of human densities per country, including in regions of higher than average density. Furthermore reserve size correlates with local human density, such that small reserves are indeed significantly more likely than are large reserves to be located in regions of high human density (n = 169; P < 0.0001). However, while local human density correlates significantly with human-caused mortality of carnivores (the only taxon for which we had data), it does not correlate with detected extinctions in reserves in east Africa (the only region with available data). Rather, area of reserve is the main predictor. Nevertheless, abundant other evidence of the adverse effects of high human density on persistence of species and wilderness indicates that we need to take as a warning the findings reported here that small reserves occur in regions of high human density, and that human density correlates with human-caused mortality. They indicate that small reserves might face the double jeopardy of both their small size, and also their situation in especially hostile surroundings. In effect, small reserves are more isolated in more adverse habitat than current analyses in conservation biology, landscape ecology, or metapopulation analysis usually indicate.