Key points are not available for this paper at this time.
The National Park Service manages over 400 units across the United States, each with unique resource protection needs. Many parks have natural resource conservation priorities that include the protection of natural soundscapes and the sound-producing species who contribute to them. It is challenging, however, to translate terabytes of raw audio into applied information for the diverse research and management questions posed by individual parks. Recent advances in terrestrial bioacoustics are making this process more tractable, particularly as BirdNET has emerged as an extensible off-the-shelf tool equipped to detect diverse assemblages of sound-producing species across varied park ecosystems. We illustrate the applied utility of using and adapting this emerging tool in combination with open code and visualizations to answer multi-taxa questions about occurrence, phenology, and wildlife responses to management. This work demonstrates both the iterative nature and challenges inherent to keeping a federal agency astride the leading edge in a rapidly progressing, data-heavy field, amid a backdrop of global change.
Cathleen Balantic (Fri,) studied this question.