Bat diversity in Cerrado landscapes: is there a fragmentation threshold?

Abstract

The knowledge about which factors determine biodiversity on the actual conditions of habitat loss and fragmentation are essential to the development of strategies for species conservation. The order Chiroptera, second most diverse among mammals, is a promising group for studies about answers to fragmentation and habitat disruption. Due to the high mobility of several bats and low number of studies, there is no consensus about the response to landscape effects in the bats diversity and distribution, mainly on Brazilian savannah. To guide landscape restoration programs, one might consider native vegetation thresholds in relation to biodiversity. When a threshold response occurs, trhough a critical point of vegetal cover loss, the species loss Will be higher than expected by a linear relation.. In this study, we will test if there are response thresholds on bat's diversity in function of landscape structure. The study will be conducted in a set of 15 landscapes (5x5 km) distributed on a vegetal cover gradient (from 5 up to 60%) on the São Paulo State, southeastern Brazil, specifically in formations, on a context with interface with semideciduous forest formations. We will sample bat diversity through mist nets on the 15 landscapes, on fragments and its respective matrices. To verify thresholds in the bats species composition we will use a multiple concurrent hypothesis model selection approach based on Akaike Information´s Theory (AIC). The following metrics will be used to compose the competitor models: remaining native and cerrado habitat percent, patch size, average patch isolation, structural and functional connectivity and type of matrix. We will analyze the response in different scales and for the ensembles and functional groups. We prospect to identify structural thresholds, which can guide conservation strategies and identify areas with more restoration potential, considering bats as allies on related projects.