Multi-scale models link biodiversity and habitat loss

A recent explosion in global biodiversity data is changing scientists' understanding of how biodiversity and global climate are linked. This has resulted in new tools and theories to help understand the mechanisms underlying the emerging patterns of biodiversity.

The WORLDIVERSITY (Linking global species richness and beta diversity to individual species distributions at multiple phylogenetic and spatial scales) project used newly available data and advances in statistics to develop tools for improving biodiversity maps and estimates of biodiversity loss through habitat loss.

Researchers used available data sets on the global distribution of birds, mammals and amphibians, and integrated them into a single database. This was used to devise a theory that connects species distributions to habitats and species extinctions to habitat loss. The novelty of the theory is that it explicitly links processes at local, regional and global scales.

Several methods, including software tools, were developed to estimate biodiversity loss along with habitat loss. Some of the tools are also able to predict geographic locations of the origin of a species, or improve predictions of species with high-resolution maps.

WORLDIVERSITY used these techniques to show the effect of habitat destruction on biodiversity loss and also revealed that pollinator diversity in Europe is no longer declining. The methods developed were also used to map malaria vectors in southern Africa. These high-profile results will help to steer the attention of the conservation community towards other aspects of biodiversity, such as phylogenetic or functional diversity.

The methods and theories developed through the project will benefit global biodiversity conservation by improving knowledge of species distributions and enabling better estimates of actual extinction rates where habitat loss occurs. Moreover, the developed tools can be applied to epidemiology, conservation biology, climatology, image processing and remote sensing.