Using trait-based filtering as a predictive framework for conservation: A case study of bats on farms in southeastern Australia

Abstract

1.With world-wide changes in human land use, an important challenge for conservation biologists is to develop frameworks to predict how species will respond to landscape change. Environmental filtering, where different environments favour different species' traits, has the potential to be a useful predictive framework. Therefore, it is important to advance our understanding of how species with different traits respond to environmental variables. 2.We investigated the distribution of microbats in a 1000000ha agricultural region in southeastern Australia, with specific emphasis on the effects of tree density on bat species characterized by different sizes, wing shapes and echolocation frequencies. The study area is substantially cleared, and trees are continuing to decline because grazing inhibits tree regeneration. We monitored bat activity acoustically at 80 sites spanning a wide range of tree densities. We used regression modelling to quantify the response of bats to tree density and other ecological covariates, and RLQ analysis to assess how different traits correlated with various environmental gradients. 3.Total bat activity and species richness peaked at intermediate tree densities. Species composition was explained by tree density and the traits of individual species. Sites with low tree cover were dominated by large, fast-flying species, whereas sites with dense tree cover were dominated by smaller, highly manoeuvrable species. These findings are consistent with recent findings from other locations around the world. 4.Synthesis and applications. Trait-based predictive frameworks enable landscape managers to assess how different management strategies and landscape modifications are likely to affect different species. Here, we propose a framework to derive general predictions of how bats respond to landscape modification, based on tree density and species traits. We apply this framework to a current conservation issue of tree decline in our study area and derive management priorities including: (i) maintaining a range of tree densities throughout the region; (ii) ensuring the persistence of locations with intermediate tree densities; and (iii) using environmentally sensitive grazing practices, for example, by incorporating long rest periods. Trait-based predictive frameworks enable landscape managers to assess how different management strategies and landscape modifications are likely to affect different species. Here, we propose a framework to derive general predictions of how bats respond to landscape modification, based on tree density and species traits. We apply this framework to a current conservation issue of tree decline in our study area and derive management priorities including: (i) maintaining a range of tree densities throughout the region; (ii) ensuring the persistence of locations with intermediate tree densities; and (iii) using environmentally sensitive grazing practices, for example, by incorporating long rest periods.