Data from: Large-scale bioacoustic monitoring to elucidate the distribution of a non-native katydid
Data files
Feb 01, 2024 version files 345.50 KB
-
README.md
-
Recognizer_data_DRYAD_ABMI_Hits.csv
-
Recognizer_data_DRYAD_Control_Hits.csv
-
Recognizer_data_DRYAD_Recorder_overview.csv
Abstract
For animals that produce species-specific audible sounds, environmental recordings combined with automated acoustic monitoring software (passive acoustic monitoring, PAM) may be an effective monitoring tool because it allows audio data from many, widely distributed autonomous recording units (ARUs) to be processed in a relatively short period of time. Males of many insect species produce loud, species-specific mating songs, yet acoustic insects have received less attention from PAM relative to vertebrates.
We evaluated the use of PAM to monitor, Roeseliana roeselii (Orthoptera, Tettigoniidae), an acoustic insect that has expanded its range to Alberta, Canada far outside its naturalized North American range. We analyzed environmental recordings from ARUs: 1) at two control sites known to be occupied by R. roeselii, and 2) across Alberta established by the Alberta Biodiversity Monitoring Institute (ABMI) to search for new populations.
PAM successfully detected R. roeselii at the two control sites, but not at any of the 74 ABMI sites that we analyzed. Despite the failure to detect new locations of R. roeselii, our analysis of ABMI environmental recordings detected several other species of acoustic insects, including Orchelimum gladiator, Gryllus sp. and Allonemobius spp.
Our results add to the growing body of work showing the feasibility of using PAM for acoustic insects. We make suggestions for how to maximize the effectiveness of this monitoring tool for the conservation and management of singing insects in North America.
README: Bioacoustic monitoring for Roeseliana roeselii using recognizers from the Alberta Bioacoustic Monitoring Institute
The dataset contains 3 CSV files that include an overview of ABMI recorders that were analysed (Recognizer_data_DRYAD_Recorder_Overview.csv), every individual hit from the recognizer for the ABMI environmental recorders (Recognizer_data_DRYAD_ABMI_Hits.csv),
and every individual hit taken from the control (Recognizer_data_DRYAD_Control_Hits.csv).
The Recognizer_data_DRYAD_Recorder_Overview.csv file is a summary of the number of hits collected at each environmental recording.
Column Header | Description |
---|---|
ARU | Automated Recording Unit name/identifier. (ARUs denoted with an * are condensed in the ABMI_hits.csv file to include all the recorders within the subcategory.) |
Year | Year that ARU was deployed in the field. |
Deployment | Date that ARU was deployed in the field. |
Retrieval | Date that ARU was retrieved from the field. |
Duration (days) | Number of days that ARU was in the field recording |
Latitude | Latitude of ARU location. |
Longitude | Longitude of ARU location. |
Distance (km) | Distance of ARU from nearest field collection location. |
R. roeselii | Number of hits classified as Roesel's katydid. |
Tettigoniidae | Number of hits classified as Tettigoniidae. |
Grylloidea | Number of hits classified as Grylloidea. |
Acrididae | Number of hits classified as Acrididae. |
Unidentified Insects | Number of hits classified as Unidentified Insects. |
Other Acoustic Insects | Sum of the number of hits classified as insects other than Roesel's katydid. |
Other Sounds | Number of hits classified as Other Sounds. |
Total Hits | Sum of the number of hits. |
The Recognizer_data_DRYAD_ABMI_Hits.csv and Recognizer_data_DRYAD_Control_Hits.csv files are raw output's of hits from the recognizers analyzing environmental recordings in Song Scope.
A description of the column identifiers for Recognizer_data_DRYAD_ABMI_Hits.csv can be seen below:
Column Header | Description |
---|---|
Filename | Unique identifier used by the ABMI. |
Timestart (H:MM:SS) | Time that the hit was detected in the audio clip in hours: minutes: seconds. |
Duration (sec) | Length of the hit that the recognizer identified. |
Quality | Indication of a likelihood of a match to the candidate vocalizations. |
Score | Indication of a likelihood of a match to the candidate vocalizations. |
Recognizer | Training recognizer connected to the detected hit. |
Classification | Classification based on human listener to recorded call. |
A description of the column identifiers for Recognizer_data_DRYAD_Control_Hits.csv can be seen below:
Column Header | Description |
---|---|
Filename | Unique identifier for each recorder. |
Quality | Indication of a likelihood of a match to the candidate vocalizations. |
Score | Indication of a likelihood of a match to the candidate vocalizations. |
Recognizer | Training recognizer used: created from the lab or field recordings |
Classification | Classification based on human listener to recorded call. |
We trained two recognizers(algorithms meant to detect similar songs in recordings) in Song Scope to the song of a non-native species of Katydid in North America called Roeseliana roeselii. We used these recognizers to analyze recordings made in environmental recordings taken by the Alberta Biodiversity Monitoring Institute.
We were unable to find R. roeselii in environmental recordings but found a large number of other Orthoptera species calls.
Data and file structure
The file structure represent individual locations where recording has taken place in Recognizer_data_DRYAD_Recorder_Overview.csv. For each location we provide the physical location (latitude and longitude), date of deployment, date of retrieval, the distance from the nearest R. roeselii detection, the number of recognizer hits (represented a match to the recognizer sound), and if Orthoptera were detected in the file.
The file structures of Recognizer_data_DRYAD_ABMI_Hits.csv and Recognizer_data_DRYAD_Control_Hits.csv represent each hit and a description of what was found. For each hit we provide the quality and score (metrics of similarity between the recording and the recognizer), the recognizer that detected it and a description of what was detected (within the bins of Noise or birds, Othoptera, or Roeseliana roeselii).