Skip to main content

The soundscape of swarming: Proof of concept for a non-invasive acoustic species identification of swarming Myotis bats

Cite this dataset

Bergmann, Anja et al. (2022). The soundscape of swarming: Proof of concept for a non-invasive acoustic species identification of swarming Myotis bats [Dataset]. Dryad.


Bats emit echolocation calls to orientate in their predominantly dark environment. Recording of species-specific calls can facilitate species identification, especially when mist-netting is not feasible. However, some taxa, such as Myotis bats are hard to distinguish acoustically. In crowded situations where calls of many individuals overlap the subtle differences between species are additionally attenuated. Here we sought to non-invasively study the phenology of Myotis bats during autumn swarming at a prominent hibernaculum. To do so we recorded sequences of overlapping echolocation calls (N=564) during nights of high swarming activity and extracted spectral parameters (peak frequency, start frequency, spectral centroid) and Linear Frequency Cepstral Coefficients (LFCCs) which additionally encompass the timbre (vocal ‘colour’) of calls. We used this parameter combination in a stepwise discriminant function analysis (DFA) to classify the call sequences to species level. A set of previously identified call sequences of single flying Myotis daubentonii and Myotis nattereri, the most common species at our study site, functioned as a training set for the DFA. 90.2% of the call sequences could be assigned to either M. daubentonii or M. nattereri, indicating the predominantly swarming species at the time of recording. We verified our results by correctly classifying a second set of previously identified call sequences with an accuracy of 100%. In addition, our acoustic species classification corresponds well to the existing knowledge on swarming phenology at the hibernaculum. Moreover, we successfully classified call sequences from a different hibernaculum to species level and verified our classification results by capturing swarming bats while we recorded them. Our findings provide the basis for a new non-invasive acoustic monitoring technique that analyses “swarming soundscapes” by combining classical acoustic parameters and LFCCs, instead of analysing single calls. Our approach for species identification is especially beneficial in situations with multiple calling individuals, such as autumn swarming.


We employed a total of three data sets consisting of echolocation call sequences for the analyses. The first data set (test data A and B) contained recordings of overlapping echolocation call sequences of swarming bats in front of the Kalkberg cave (A) and a second site in Northern Germany (B). Our goal was to identify the predominantly echolocating species in these recordings. Therefore, the reference data set was used as training data in a discriminant function analysis to classify recordings from the test data and the control data. 

Test data A: Sound recordings of the test data set A were conducted on nights with high swarming activity of Myotis bats during the autumn swarming seasons in 2018 and 2019 at both entrances of the Kalkberg cave in Bad Segeberg, Germany. Recordings were made whenever a high number of bats was swarming simultaneously using a high-quality ultrasonic microphone (Avisoft USG 116Hm with condenser microphone CM16; frequency range 1‑200 kHz) connected to a small computer (Dell Venue 8) running the software Avisoft Recorder (v4.2.05, R. Specht, Avisoft Bioacoustics, Glienicke, Germany). For the subsequent acoustic analysis, 564 echolocation call sequences (mean: 11.3 sequences per night; range: 1-29) with a length of four seconds each were selected based on the quality of the sound recordings and the presence of a high number of echolocation calls without interfering social vocalizations.

Test data B: The swarming bats within this data set were recorded at another autumn swarming site in Northern Germany (Lüneburg) during one night (22.09.2021). Recordings were made and selected as described above. Bats of this data set were identified by simultaneous mist-netting. 

Reference data: To classify the recorded echolocation call sequences from the swarming situation, identified echolocation call sequences of M. daubentonii and M. nattereri were used as a reference (i.e. as training set in a discriminant function analysis). These echolocation call sequences came from singly flying individuals and were recorded at ten underground sites with a Batcorder (ecoObs GmbH, Nürnberg, Germany) using a sampling rate of 500 kHz and a trigger threshold of -36 dB (quality 26-28). The calling species were identified via photos from synchronized camera trap images (Wimmer and Kugelschaft (2015): Akustische Erfassung von Fledermäusen in unterirdischen Quartieren: GRIN Verlag). 

Control data: To validate our statistical classification of the test data sets A and B, we classified an additional data set as a control using the same reference data. The echolocation call sequences in the control data set were recorded using a Petterson D980 (Pettersson Elektronik AB, Sweden) in time expansion mode (Skiba (2009): Europäische Fledermäuse: Kennzeichen, Echoortung und Detektoranwendung, 2nd ed.: Westarp Wissenschaften-Verlagsgesellschaft mbH). 


Elsa-Neumann Foundation