The impact of vehicular noise on the effectiveness of acoustic indices

Published Dec 14, 2023 on Dryad. https://doi.org/10.5061/dryad.sxksn038p

Abstract

Passive acoustic monitoring (PAM) is a sampling technique that has gained increasing popularity in the field of wildlife monitoring and research. The technique involves the deployment of programmable autonomous recording units, allowing for non-invasive and cost-effective collection of acoustic information across a breadth of temporal and spatial scales. Retrieving biological information from PAM recordings can often involve time-consuming sound annotation methodologies, but the advent of acoustic indices has allowed for biodiversity metrics to be estimated quickly and with reasonable accuracy.

While correlations between acoustic indices and species richness have been observed in a variety of ecological contexts, these relationships tend to faulter in environments with increased vehicular noise. In turn, many researchers will avoid areas with vehicular traffic for PAM with intent on using acoustic indices. Here, we assessed the direct impact of vehicular noise on nine acoustic indices through controlled manipulation of vehicular noise within computer-generated soundscapes. Acoustic index values derived from these soundscapes were then used to fit a suite of linear mixed-effects models from which indicator variable coefficients were derived. These coefficients were used to determine the impact of respective vehicular noise levels on acoustic index values.

Our results demonstrate that recording distance from roadsides and number of passing cars per minute have notable and persistent impacts on acoustic index values, but the magnitude of the effect varies across indices. Four acoustic indices demonstrated greater resilience to vehicular noise interference and may therefore be better suited for PAM in developed areas: Bioacoustic Index (BI), Acoustic Complexity Index (ACI), Acoustic Diversity Index (ADI), and Acoustic Evenness Index (AEI).

By contributing to the collective understanding of acoustic index behaviors under anthropogenic noise pollution, we hope to better inform their ecological application within future PAM efforts in human-developed contexts.

Rebecca L. Ducay & Brent S. Pease

Please contact the first author for questions about the code or data: Rebecca L. Ducay (rebecca.ducay@siu.edu)

Repository Directory

RCode: Contains code for cleaning, processing, and analyzing experimental and supplemental data

101 - Bird Assemblages - EBD_list.R - script used to retrieve a random list of bird species from an eBird checklist. Data used from eBird is not compatible with the CC0 license waiver required by Dryad and thus cannot be posted. The EBD was used in Feb 2023 to retrieve checklists from Illinois from May - July 2022. The EBD data can be downloaded using the following link with a valid eBird log-in: https://science.ebird.org/en/use-ebird-data/download-ebird-data-products.
201 - Acoustic indices - Acoustic Index Loop.R - script used for calculating all acoustic indices except AR from .wav files.
202 - Acoustic indices - AR Code.R - script used for calculating AR from .wav files.
301 - Analysis - Species Effect Code.R - script used for generating Figure 2, depicting the interactive effects of distance and cars at each level of species richness.
302 - Analysis - Vehicular Effect Code.R - script used for generating Figure 3, depicting the interactive effects of distance and cars with 4 vocalizing bird species. Also used to retrieve p-values for all mixed- effects model coefficients
303 - Analysis - Species Equivalence Code.R - script used for generating Figure 4, depicting the added species equivalence of each acoustic index at low, mid, and high traffic levels.
401 - Supplemental - Appendix Figure Code.R - script used for generating Figure A1, depicting the control acoustic index values as predicted by the number of added bird species.

Data: Contains raw data used in scripts

all_species_indices.csv - acoustic index values for all soundscapes containing 2, 8, and 12 species, as well as all soundscapes derived from the first 4 assemblages containing 4 bird species. Variables are as follows:
- assemblage - the index number indicating the randomized bird assemblage used to develop the soundscape, in which soundscapes with the same assemblage contain the same bird species.
- Cars - the number of car drive-by sound clips added to the soundscape.
- Distance - designates which car drive-by sound clips were added to the recording in order to simulate the effort of a deployment configuration at specific distances from the roadside.
- Species - the number of vocalizing bird species added to the soundscape.
- BI - Bioacoustic Index first described in Boelman et al. (2007).
- H - Acoustic Entropy Index first described in Suer et al. (2008).
- ACI - Acoustic Complexity Index first described in Pieretti et al. (2011).
- ADI - Acoustic Diversity Index first described in Villanueva-Rivera et al. (2011).
- AEI - Acoustic Evenness Index first described in Villanueva-Rivera et al. (2011).
- M - Amplitude Index first described in Depraetere et al. (2012).
- NDSI - Normalized Difference Soundscape Index first described in Kasten et al. (2012).
- NP - Number of Frequency Peaks first described Gasc et al. (2013).
- AR - Acoustic Richness Index first described in Depraetere et al. (2012).
four_species_indices.csv - acoustic index values for all soundscapes containing 4 bird species. Variables are the same as described above.

Audio_Files: contains original and modified car driveby soundclips, as well as .wav files for all generated soundscapes

Car driveby.wav - original drive-by sound clip retrieved from the Youtube Audio Library
Car_50m.wav - drive-by sound clip edited to reflect the frequency and amplitude at 50 meters from the roadside
Car_150m.wav - drive-by sound clip edited to reflect the frequency and amplitude at 150 meters from the roadside
Car_250m.wav - drive-by sound clip edited to reflect the frequency and amplitude at 150 meters from the roadside
All other audio files are generated soundscapes. File names provide the assemblage index number, the number of added car drive-by sound clips, and the distance that the added car drive-by clips reflect (e.g., assemblage23_5c_250m.wav). Control soundscapes only contain bird vocalizations and pink noise with no added car sounds.