Assessing how habitat loss restricts the geographic range of Neotropical anurans. Issue 6 (29th September 2016)

• Record Type: Journal Article
• Authors: Vasconcelos, Tiago S.; Doro, João Lucas P.
• Is Part Of: Ecological research. Volume 31:Issue 6(2016)
• Page Start: 913
• Page End: 921
• Publication Date: 2016-09-29
• Subjects: Anura -- Atlantic forest -- Cerrado -- Habitat loss -- Species distribution modeling; Ecology -- Periodicals; Ecology -- Japan -- Periodicals; Écologie; Japon; Ecology; Japan; Ressource Internet (Descripteur de forme); Périodique électronique (Descripteur de forme); Periodicals; 577.05
• DOI: 10.1007/s11284-016-1401-8
• Languages: English
• ISSNs: 0912-3814

Abstract

Abstract: Habitat loss and fragmentation exert unquestionable negative effects in a wide range of taxa on both regional and local scales. However, there is a debate over whether habitat change impacts geographic species distribution. We assess how habitat loss restricts large‐scale species distribution on a geographic scale for four South American anurans that are known to occur in well conserved habitats, yet which are absent in others that are close by and more degraded. We used occurrence records of each species in Brazil and performed different modeling algorithms to compare ensemble distribution models generated by two different sets of predictors: a climate‐only versus a climate‐habitat procedure. We found that the distribution area predicted by the climate‐only procedure was larger than that of the climate‐habitat procedure for all species. The areas not predicted by the climate‐habitat but predicted by the climate‐only procedure for all species are commonly located in inland areas in southeastern Brazil, which coincides with areas that have suffered the most from habitat loss in the country. Plotting the predictions against well‐surveyed areas where the species have not been recorded, we found evidence that habitat loss may have restricted the current geographic ranges of Hypsiboas faber and Rhinella ornata . Finally, modeling approaches incorporating habitat landscape metrics, particularly for habitat‐specialist species, may be a helpful tool for identifying areas that harbored these species before deforestation took place.