The Pettersson D500x detector was mounted permanently next to a permanent water on McNary National Wildlife Refuge. Microphone was 3 meters above ground level and about 2 meters from the waters edge.
Detector was set to record from sunset to 2 hours before sunrise.
Detector settings were: Frequency = 500; Pretrigger = OFF, File Length = 5 seconds; Input Gain = 80; Trigger Level = 120; Interval = 0.
Files were processed with SonoBat 3.1. They were scrubbed with the high-grade scrubber, then batch analyzed using the Washington East Filter. Default settings include: Max # of calls to consider per file = 8, acceptable call quality = 0.80, acceptable quality to tally passes = 0.20,
decision threshold = 0.90. Filter settings were 5 kHz and autofilter.
A small numer of files were vetted by hand. Vetting was conducted to verify species presence for each calendar month, and to double-check identifcation of species suspected to be out of range or uncommon species in the dataset.
For calls that were not vetted, the CONSENSUS column from SonoBat was used for species identification. If the consensus column was blank, calls were grouped by low or high frequency, based on those columns in the output file of SonoBat.
Data summarizes presented are raw output from SonoBat, with little hand-vetting of results. Approximately 3% of the calls files were vetted.
The numbers included in the dataset represent number of call files collected per species per night.
I am a zone biologist with the USFWS, Region 1, Inventory and Monitoring Initiative