Habitat loss is the main driver of the current biodiversity crisis, a landscape-scale process that affects the survival of spatially-structured populations. Although it is well-established that species responses to habitat loss can be abrupt, the existence of a biodiversity threshold is still the cause of much controversy in the literature and would require that most species respond similarly to the loss of native vegetation. Here we test the existence of a biodiversity threshold, i.e. an abrupt decline in species richness, with habitat loss. We draw on a spatially-replicated dataset on Atlantic forest small mammals, consisting of 16 sampling sites divided between forests and matrix habitats in each of five 3600-ha landscapes (varying from 5% to 45% forest cover), and on an a priori classification of species into habitat requirement categories (forest specialists, habitat generalists and open-area specialists). Forest specialists declined abruptly below 30% of forest cover, and spillover to the matrix occurred only in more forested landscapes. Generalists responded positively to landscape heterogeneity, peaking at intermediary levels of forest cover. Open area specialists dominated the matrix and did not spillover to forests. As a result of these distinct responses, we observed a biodiversity threshold for the small mammal community below 30% forest cover, and a peak in species richness just above this threshold. Our results highlight that cross habitat spillover may be asymmetrical and contingent on landscape context, occurring mainly from forests to the matrix and only in more forested landscapes. Moreover, they indicate the potential for biodiversity thresholds in human-modified landscapes, and the importance of landscape heterogeneity to biodiversity. Since forest loss affected not only the conservation value of forest patches, but also the potential for biodiversity-mediated services in anthropogenic habitats, our work indicates the importance of proactive measures to avoid human-modified landscapes to cross this threshold.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.