Community Dynamics and Occupancy for Bat Species within the Great Lakes Region Following White-Nose Syndrome Exposure

Abstract

North American bats are threatened by several factors including timber harvest, wind energy development, and white-nose syndrome (WNS). We used 5 long-term bat monitoring stations (LTBMS) to monitor the presence of bats in Wisconsin and determine changes in occupancy and detection probabilities from 2007 to 2017. We automatically classified calls to species using Kaleidoscope Pro (Wildlife Acoustics Inc., Maynard, MA), performed manual vetting, and generated an occurrence dataset for each year. We used multi-species occupancy models in a Bayesian framework to compare community dynamics of bats temporally by year before and after the arrival of WNS in the state. Occupancy probability for all species of bats excluding hoary bats (Lasirurus cinereus) declined over the survey period (i.e. 95% CI were <0 for all estimate coefficients) and following the detection of WNS. Big brown bats (Eptesicus fuscus) declined the least in occupancy probability overall. We observed larger declines in occupancy probabilities for the remaining cave-dwelling species, (little brown bats (Myotis lucifugus), tri colored bats (Perimyotis subflavus), and northern long-eared bats (Myotis septentrionalis)) than other species of bats. Occupancy probabilities of eastern red bats (Lasirurus borealis) and silver-haired bats (Lasionycteris noctivagans) declined over the 11-year period. Detection probability fluctuated across all species and years with cave-dwelling species ranging in the lowest values overall. Our results indicate that LTBMS are successful in documenting long-term changes in the occupancy probabilities of bat species. Managers with similar objectives could use our methodology or results as a means of comparison or prediction to changes in the bat community across North America. Several species of bats in North American are of conservation concern due to multiple factors including habitat degradation, incidental take, and the devastating fungal disease white-nose syndrome. The impact of these factors varies across species, resulting in the need for bat research on the community scale. The Western Great Lakes Region (WGLR; Minnesota, Wisconsin and the northern peninsula of Michigan) is an area where all of these threats are present. During the summers of 2015-2017, we surveyed 14 sites in the WGLR using an array of 20 SMZC echolocation recorders (Wildlife Acoustics Inc. Maynard, MA) randomly stratified between riparian and upland habitats. We used Kaleidoscope Pro (Wildlife Acoustics) and Echoclass (Eric Britzke, U.S. Army Engineer Research and Development Center, Vicksburg, MS, USA) software to identify recorded bat calls to species. Over 3 years and 2,624 recorder nights, 207,171 audio recordings were collected and from these, 74,267 bat calls were identified. We used multi-species occupancy modeling to determine the influence of covariates including, the co-detection of other bat species, distance to road, distance to water, elevation, canopy stage, and year. Our models identified important co-detection relationships across the community; big brown bats (Eptesicus fuscus) influenced the occupancy of eastern red bats (Lasirurus borealis), and northern long-eared bats (Myotis septentrionalis); the co-detection of little brown bats (Myotis lucifugus) influenced the occupancy of big brown bats; and the co-detection of northern long-eared bats influenced the occupancy of little brown bats. Year of survey was the only variable that influenced the occupancy of hoary bats (Lasirurus cinereus), silver-haired bats (Lasionycteris noctivagans), and tri-colored bats (Perimyotis subflavus). Our results indicate that the analytical method, multi-species occupancy modeling is a logical application for bat echolocation acoustic data and can help address research needs on community dynamics and species interactions. Managers can use our results throughout the WGLR to guide research on species interactions and niche ecology across the bat community on a fine-scale to determine temporal niche partitioning.