Home

Welcome to the Odontocete Repertoire Project wiki!

The collection of all possible sounds an animal can produce is called vocal repertoire. Only a tiny portion of vocal repertoires have been characterized, with those of songbirds and non-human primates among the most studied. In these two systems, a growing number of studies show support for an evolutionary increase in vocal repertoire size with increasing social complexity.  However, for one of the most acoustically specialized lineages of vertebrates, the odontocetes, the evolutionary basis of their vocal repertoire remains largely unknown.

Odontoceti is an excellent lineage to test the hypothesis that evolutionary changes in vocal repertoire diversity are linked to key components of social living. These species have vocal repertoires of different sizes and complexity and live in groups that vary in size and types of social bonds. Quantifying the vocal repertoire of open learner species is undoubtedly challenging, yet such characterization represents an important first step in understanding how these animals organize socially, and the evolutionary basis of their communication.

The goal of this global initiative is two-fold:

1. Create an open-source pipeline (using ML tools) to generate comparable measures of vocal repertoires including size, diversity, composition, and acoustic structure for various species across the tree of life of Odontoceti, that vary in their social organization.

2. To use these metrics to estimate the rate of evolutionary change of vocal repertoires, and test for evolutionary correlations and test hypothesis about the factors shaping vocal repertoires in toothed whales (i.e., social living, environment).

Together this will facilitate the generation of species-level vocal repertoire manuscripts, expanding our understanding of present and past factors shaping their communication.

Purpose of This Wiki

This Wiki is designed to:

• Serve as a centralized, open source for data across the odontocete phylogeny to be collected from and used by collaborators.
• Provide step-by-step installation guides for the bioacoustic software packages being used.
• Offer detailed user guides on how to input, process, and analyze vocalizations using these tools.
• Share scripts, code snippets, and troubleshooting tips for common challenges when analyzing bioacoustic data.

Software Tools Overview

Below is a brief description of each tool reviewed in our study. Click on the links for detailed user guides and installation instructions:

Luscinia: A bioacoustics analysis tool well-suited for manual contour detection and tracing. It features robust tools for extracting frequency and temporal features, and built-in multivariate analyses for repertoire similarity.

Beluga: A semi-automated software specializing in the detection and categorization of acoustic contours. Its machine learning features strike a balance between automation and manual oversight.

ARTwarp: Focuses on fully automated contour similarity-based categorization. It offers flexible settings for defining similarity thresholds, making it a powerful tool for acoustic categorization.

DeepSqueak: A machine-learning-driven package for semi-automated contour detection and classification. It excels in speed and ease of use while providing accurate contour categorization across large datasets.

PAMGuard: Designed primarily for passive acoustic monitoring, this tool automates call detection but may require extensive post-processing for contour extraction and categorization.

SASLab: A comprehensive bioacoustic analysis package that includes fully automated detection and contour extraction tools but requires additional customization for optimal performance in vocal repertoire studies.

Publications using this methodology

1. Austin, M., Oswald, J. N., Rege-Colt, M, Gagne, E., Ramos, E. A., De Weerdt, J., Ransome, N. and L. J. May-Collado.2025. A computational framework to characterize and compare the tonal repertoires of toothed whales. Methods in Ecology and Evolution. 00:1–19.https://doi.org/10.1111/2041-210X.70065

2. Rege-Colt, M., Oswald, J.N., De Weerdt, J. et al. Whistle repertoire and structure reflect ecotype distinction of pantropical spotted dolphins in the Eastern Tropical Pacific. Sci Rep 13, 13449 (2023). https://doi.org/10.1038/s41598-023-40691-8.

3. Gagne E, Perez-Ortega B, Hendry AP, Melo-Santos G, Walmsley SF, Rege-Colt M, Austin M and May-Collado LJ (2022) Dolphin communication during widespread systematic noise reduction-a natural experiment amid COVID-19 lockdowns. Front. Remote Sens. 3:934608. doi: 10.3389/frsen.2022.934608.

4. Antichi, S., Rege-Colt, M., Austin, M., May-Collado, L.J., Carlon-Beltran, O., Urban, J., Martinez-Aguilar, S., and L. Viloria-Gomora. 2024. Whistles contour variation between two sympatric dolphin species in the Gulf of California. Hydrobiology. https://doi.org/10.1007/s10750-024-05675-3.

5. Antichi, S., Austin, M., May-Collado, L.J., Urban, J. R., Martinez-Aguilar, S., and L. Viloria-Gomora. 2023. Differences in the whistles of two ecotypes of bottlenose dolphins in the Gulf of California. Journal of the Acoustical Society of America- Express Letters. https://doi.org/10.1121/10.0019502.

6. Melos-Santos, G., Figueiredo Rodrigues, A. L., Tardin, R. H., de Sa Maciel, I., Marmontel, M., Da Silva, M. L., and L. J. May-Collado. 2019.The newly described Araguaian river dolphins, Inia araguaiaensis (Cetartyodactyla, Iniidae), produce a diverse repertoire of acoustic signals. Peer J. https://peerj.com/articles/6670/

Collaboration

We are actively seeking collaborators who can provide acoustic data from species across the phylogeny, so that we can have exhaustive and comprehensive representations across their tree of life! We request contributors to allow for contours to be open access on the website and the metadata. Raw recordings will require permission from owners before we share them.

For more information, please contact Dr. Laura J May Collado (lmaycoll@uvm.edu) at the University of Vermont.