USFWS Great Lakes and Upper Midwest Acoustic Bat Dataset

This dataset contains results from acoustic monitoring for bats at 276 sites across the Great Lakes basin and Upper Midwest From 2010 - 2018. These data provide information on the magnitude, timing, and species composition of bat activity across the region. These data were collected by the US Fish and Wildlife Service Region 3 Avian Radar Team (https://www.fws.gov/radar) funded by the Great Lakes Restoration Initiative, as well as USFWS Field Offices, the University of Minnesota, and various collaborators including Bat Conservation International, Texas Christian University, US Geological Survey, Western Ecosystems Technology Inc., and organizations listed here: https://www.fws.gov/radar/links/index.html

METHODS: All detectors (Wildlife Acoustics SM2Bat+) were deployed at approximately 1 m above ground level on fence posts or existing structures. PVC cups were placed around microphones to block extraneous ground noise but allow acoustic signals from overhead. Detectors were not placed beneath completely closed canopy where possible, to provide partially or fully open sky above. Detectors ran autonomously for 10 - 14 days on external batteries. Recordings were made via triggering operation under default settings with a 1 s trigger window, recording to 15-minute WAC compressed files. A low-pass filter excluding sounds below 16 kHz was applied. Data were recorded to SD cards and transferred to headquarters by field personnel. WAC files were decompressed using WAC2WAV utility, with split triggers ON. Kaleidoscope Pro v. 4.3.2 (-1 liberal setting) was used for species identification, with species classifiers (the group of species included in analysis as potential species to be identified) selected for each site using IUCN range maps, plus a 50 km buffer. Automated species identifications have not been manually vetted for accuracy, and may contain false identifications. Nightly Pass Counts represent only nights when 8 or more WAC files (indicating 2 hours or more of detector operation) were recorded. Nights with fewer than 8 WAC files, or with data that indicated detector malfunction were removed from Nightly Pass Counts, but may be represented in Pass Lists. For more information on data collection, please refer to: Heist KW. Assessing bat and bird fatality risk at wind farm sites using acoustic detectors. PhD Dissertation, University of Minnesota. 2014.