Biodiversity gradients in the Alps: The overriding importance of elevation

Abstract

Land abandonment is causing woodland expansion and loss of open habitats in the Alps, coupled with a shift in forestry practices from coppice management to high forest. Despite such rapid large-scale changes, there has been very little investigation of the environmental predictors of biodiversity in the Alpine landscape. We assessed the richness of amphibians, reptiles and breeding birds (n = 189 species), used as a surrogate of biodiversity, in 58 quadrats of 100 km2, located within a well surveyed area of the province of Trento (central-eastern Italian Alps). The surrogates were then related to a series of environmental variables by means of stepwise multiple regression. Depending on the surrogate analysed, species richness declined linearly or quadratically with elevation, and increased with habitat heterogeneity and the availability of grassland and arid-rocky habitats. The same results were obtained when incorporating a measure of species threat into the biodiversity estimates. Different surrogates were positively inter-correlated, probably because of a common response to the same factor, namely elevation, which was the only variable to enter all models. Such elevational gradient produced a clear biodiversity peak in low-elevation areas, generating potential conflict between efficient biodiversity conservation and economic interests linked to human development, a scenario which probably applies to many mountain regions worldwide. The current network of protected areas was quite satisfactory in terms of area covered but biased towards high-elevation areas, of high scenic beauty but relatively low in animal biodiversity value. Low-elevation reserves were small and isolated. Proposed conservation targets include the establishment of corridors increasing the connectivity of low-elevation reserves and the promotion of incentives for the extensive management of grassland, an agro-ecosystem of high historical and biological value. © 2007 Springer Science+Business Media B.V.