CAN YOU TRUST THE RESULTS OF YOUR AUTOMATED BAT CALL CLASSIFIER?
Bats are an integral part of our ecosystem and provide numerous ecosystem services such as pollination, seed dispersal, and in our region insect control. Sampling the bat community as a means to assess ecosystem health is typically accomplished through physical capture and the collection of acoustic data. In many instances, surveys are acoustic only in nature because it allows researchers to noninvasively sample the community as the ability to assign species identification to each call has improved over the last few decades. In fact, regulatory agencies currently use acoustics to confirm presence or absence of endangered species in a location. Unfortunately, identifying species based on calls is extremely challenging and, often plagued by high inaccuracies. Our objective for this project was to compare two distinct classification programs that are currently approved by the Fish and Wildlife Service as automated bat-call classifiers. To accomplish this objective, we used calls collected with two Anabat Swift detectors in a forest landscape between July and August 2018 in Lumpkin Co., GA and classified the dataset using both classification programs. The dataset contained 148,563 files with a total of 805 calls identified between the two detectors. Automated Bat Call Identification (BCID) software found 10 distinct species and Echoclass found 7 distinct species. We found very little agreement between the two classification methods, consequently, we recommend caution when relying solely on the software programs to determine abundance of species in a particular area. In many instances, indices of activity may be a more appropriate response variable of interest given the high degree of uncertainty that remains when attempting to ID calls to species. "
Robertson*, Sara Alisha and Bender, Michael
"CAN YOU TRUST THE RESULTS OF YOUR AUTOMATED BAT CALL CLASSIFIER?,"
Georgia Journal of Science, Vol. 77, No. 1, Article 74.
Available at: https://digitalcommons.gaacademy.org/gjs/vol77/iss1/74