Meyer, C.F.J. (2015) Methodological challenges in monitoring bat population - and assemblage-level changes for anthropogenic impact assessment.Mammalian Biology, 80, 159-169. DOI:10.1016/j.mambio.2014.11.002 (IF2015 1,595; Q2 Zoology)
Recent years have seen increased attention to bats as an effective bioindicator group for assessing responses to drivers of global change, which concurrently has led to a revived interest in establishing a global bat monitoring network. To be effective and efficient, global-scale monitoring of bats will largely have to rely on integrating data collected as part of a network of regional monitoring schemes. Herein, I highlight and discuss some of the principal challenges faced in the monitoring of population- and assemblage-level changes of bats, focusing mainly on methodological and statistical issues and the selection of suitable state variables for quantifying regional trends in bat biodiversity. Particularly in the tropics, where detailed single-species monitoring is challenging due to high species richness, I recommend that monitoring programs focus on tracking changes in species turnover and composition as more informative measures of anthropogenic impact than species richness. Imperfect species detection is an important source of variation and uncertainty associated with animal count data. Bat monitoring programs need to correct for this, most importantly through the use of sampling protocols that rely on strictly standardized approaches and a well-balanced design, or a posteriori by using appropriate statistical models so as to avoid the detection of spurious trends. Multi-species occupancy models that allow for simultaneous assemblage- and species-level inference about occurrence and detection probabilities provide a suitable analysis framework for monitoring data, and are a comparatively low-cost approach that should prove useful especially in the regional monitoring of bats in the tropics. To ensure robust inference about temporal and spatial trend estimates in the state variables of interest, the efficacy of sampling designs should be carefully gauged at the design stage to ensure sufficient statistical power, and data should be collected according to a formal randomized design to allow for regional-scale inference. I stress the importance for long-term bat monitoring programs to have sustained funding, the need to establish trigger points for the application of appropriate mitigation measures, and for monitoring to be adaptive so as to maximize effectiveness and efficiency based on the data collected. Finally, I argue that to overcome the challenges associated with initiating monitoring networks in tropical countries – a major step towards the realization of global-scale bat monitoring – reliance on citizen scientists and participatory monitoring will be key.