Rana sierrae annotated aquatic soundscapes (2022)
Data files
Nov 20, 2023 version files 40.97 MB
Abstract
This dataset is associated with the following manuscript, which contains details in the methodology of data collection and annotation:
Lapp, S., Smith, T. C., Wilhelm, A, Knapp, R., Kitzes, J. In press. Aquatic soundscape recordings reveal diverse vocalizations and nocturnal activity of an endangered frog. The American Naturalist.
Rana sierrae (the Sierra Nevada yellow-legged frog) is an endangered species residing in high-elevation lakes in the Sierra Nevada mountains. The species is highly aquatic and, unlike most amphibians, primarily vocalizes while underwater. As a result, its vocalizations have rarely been recorded and its vocal repertoire is not well studied.
This dataset contains an annotated set of underwater soundscape recordings containing 1236 annotations of R. sierrae vocalizations. We annotated five distinct vocalization types of R. sierrae, only two of which have been previously documented for this species. Besides the calls of R. sierrae, these audio recordings also contain stridulation sounds (not annotated), which were most likely produced by members of the family Corixidae or other aquatic invertebrates that stridulate underwater.
README: Rana sierrae annotated soundscape recordings
Audio files and Raven annotations for Rana sierrae (Sierra Nevada Yellow-legged frog) vocalization types, from aquatic soundscape recordings.
Name: rana_sierrae_2022
Version: 1.0
This dataset is associated with the following manuscript, which provides details on the methodology of data collection an annotation:
Lapp, S., Smith, T. C., Wilhelm, A, Knapp, R., Kitzes, J. In press. Aquatic soundscape recordings reveal diverse vocalizations and nocturnal activity of an endangered frog. The American Naturalist.
Data and file structure
This dataset contains audio files annotated in Raven Pro for Rana sierrae vocalization types. Frequency-time boxes were drawn around all sounds identifiable as vocalizations of Rana sierrae. Five vocalization types are annotated, and are described in the associated manuscript. In total, the dataset contains 1236 annotations of Rana sierrae vocalizations. Other sounds in the aquatic soundscape include stridulations most likely produced by members of the family Corixidae or other aquatic invertebrates.
Audio:
Audio files are provided in mp3 format in the mp3 subdirectory.
The dataset contains 672 10-second files. These are the first 10 seconds of each audio file recorded during the week of June 20-26 2022 on one device (Device 3 in the associated paper; North corner of lake; this device had the highest activity level). The underwater AudioMoth 1.2.0 recorder in underwater case recorded 1 minute starting every 15 minutes 24 hours per day, resulting in (24*4*7)=672 audio recordings.
Annotations:
Files were annotated by Sam Lapp using Raven Pro with closed-back headphones while viewing spectrogram. Only calls that could both be heard and seen on spectrogram were annotated. Multiple vocalizations were included in a single annotation box if they were separated by greater than 1 second of intervening time without vocalizations of the same type.
Labels correspond to the associated manuscript:
A primary vocalization
B stuttered vocalization described in Vredenburg et al
C chuck, double/triple chuck calls
D short downward single note
E frequency-modulated call
X: could not determine if sound is R. sierrae or not; these are excluded from training and validation of the CNN
Raven annotation files
The subdirectory raven_selection_tables contains one Raven-formatted text file (tab separated values format) for each audio file. The file audio_and_raven_files.csv is a table that lists each audio file and the corresponding raven annotation file. The filename for each annotation file matches the corresponding audio file (for instance, sine2022a_MSD-0558_20220620_000000_0-10s.mp3 and sine2022a_MSD-0558_20220620_000000_0-10s.Table.1.selections.txt).
One-hot labels
This dataset also includes the file (labels_2s.csv), which contains one-hot labels (0/1 per class per audio clip) for 2-second segments of audio. To generate these labels, we considered R. sierrae vocalizations to be present in a 2-second sample if any R. sierrae annotation overlapped with the sample by at least 0.2 seconds or if greater than 50% of an annotation box overlapped in time with the sample.
A notebook in the associated GitHub repository documents how the Raven annotations were converted to one-hot labels.
Sharing/Access information
This data is publicly available here. Associated scripts for data analysis are located in a GitHub repository.
-Sam Lapp November 2023
Methods
The audio in this dataset is a set of 672 10-second audio files recorded at a spacing of 15 minutes over the course of 7 days on a single underwater audio recorder. The recorder, an AudioMoth version 1.2.0 in an underwater case, was deployed approximately 0.5 m from the shoreline, on the bottom of a lake in the Sierra Nevada in which R. sierrae breed and overwinter.
The annotations of the five call types correspond to the descriptions in the associated manuscript:
A primary vocalization "meow" described in Vredenburg et al 2007
B stuttered vocalization, also described in Vredenburg et al 2007
C chuck, double/triple chuck calls
D short downward single note
E frequency-modulated call
X: could not determine if sound is R. sierrae or not; these were excluded from training and validation of the CNN in the manuscript
Files were annotated by Sam Lapp using Raven Pro with closed-back headphones while viewing spectrogram. Only calls that could both be heard and seen on spectrogram were annotated. This dataset also contains one-hot labels (0/1 per class per audio clip) for 2-second segments of audio. To generate these labels, we considered R. sierrae vocalizations to be present in a 2-second sample if any R. sierrae annotation overlapped with the sample by at least 0.2 seconds or if greater than 50% of an annotation box overlapped in time with the sample. A notebook in the associated GitHub repository demonstrates how the Raven annotations were converted to one-hot labels.
Works Cited
Usage notes
Audio files are provided in .mp3 format and are compatible with any program that supports this format. For instance, audio can be played and viewed as a spectrogram in the free Audacity software. Annotation files are provided in two formats:
(1) Raven annotation tables, which can be opened along with audio files in Raven Lite (free) or Raven Pro (paid) software to view, create, and manipulate annotation boxes on spectrograms. Note that even without Raven Lite / Pro, the annotation files can be loaded, explored, and manipulated using the free and open-source python package OpenSoundscape.
(2) A .csv file containing a table of 0/1 (present/absent, also known as "one-hot") labels for each call type for 2-second audio segments. This file can be opened with text or table editors, or loaded into Python using the pandas package.