Passive acoustic surveys and the BirdNET algorithm reveal detailed spatiotemporal variation in the vocal activity of two anurans

• Published in: Bioacoustics (Berkhamsted) Vol. 32; no. 5; pp. 532 - 543
• Main Authors: Wood, Connor M., Kahl, Stefan, Barnes, Stephanie, Van Horne, Rachel, Brown, Cathy
• Format: Journal Article
• Language: English
• Published: Taylor & Francis 03.09.2023
• Subjects: Artificial intelligence; machine learning; Pacific chorus frog; passive acoustic monitoring; Yosemite toad
• ISSN: 0952-4622; 2165-0586
• DOI: 10.1080/09524622.2023.2211544

Passive acoustic monitoring has proven effective for broad-scale population surveys of acoustically active species, making it a valuable tool for conserving threatened species. However, successful automated classification of anuran vocalisations in large audio datasets has been limited. We deployed five autonomous recording units at three known breeding areas of the Yosemite toad (Anaxyrus canorus), which is threatened and relatively uncommon, and the sympatric Pacific chorus frog (Pseudacris regilla), which is widespread and more common, to test the viability of bioacoustics as a means of supplementing ongoing, human survey efforts. We analysed the audio data with the BirdNET algorithm, which was originally developed for birds but has been expanded to include both species. We achieved efficient and accurate identification of both species in 2,756 h of audio, which yielded high-resolution phenological data about seasonal and daily vocal activity as well as daily detection counts. These findings demonstrate that a newly expanded machine learning detector, BirdNET, can effectively process passive acoustic surveys for these species. Further exploration of how passive acoustic monitoring may complement existing survey techniques for these and other Anurans is warranted.