COS 7-1 - Marine soundscapes indicate kelp forest condition

Monday, August 12, 2019: 1:30 PM
L011/012, Kentucky International Convention Center
Benjamin L. Gottesman1, David T. Savage1, Kristen Bellisario1, Joshua Sprague2, David J. Kushner2, Megan F. McKenna3, Mary Jo Barkaszi4, Michele B. Halvorsen4, Eva M. DiDonato5, David L. Conlin6 and Bryan C. Pijanowski1, (1)Department of Forestry and Natural Resources, Purdue University, West Lafayette, IN, (2)Channel Islands National Park, National Park Service, Ventura, CA, (3)Natural Sounds and Night Skies Division, National Park Service, (4)CSA Ocean Sciences, Stuart, FL, (5)Ocean and Coastal Resources, National Park Service, Ft. Collins, CO, (6)Submerged Resources Center, National Park Service, Lakewood, CO
Background/Question/Methods

Different marine habitats produce distinct soundscapes. Assessing these soundscapes has demonstrated potential to provide scalable and continuous information on habitat condition. However, there is still a sizable knowledge gap concerning which soundscape features reflect different ecological properties and the mechanisms underlying these relationships.

In this study, we tested the utility of soundscape recording and analysis in kelp forest habitats off the coast of California, USA, in Channel Islands National Park. Specifically, we investigated 1) if soundscape features correlated with ecological variables of interest and 2) whether soundscape features differed inside and outside of marine protected areas.

We collected a total of 6,523 hours of acoustic recordings between Nov. 2017–Sept. 2018 from five kelp forest sites. Three sites were in marine protected areas with high kelp abundance (7.7–34.8 sqm-1) and low sea urchin density (0.4–3.6 sqm-1) and two were in adjacent, unprotected habitats with low kelp abundance (0–0.1 sqm-1) and high sea urchin density (10.5–26.8 sqm-1). To analyze the acoustic data, we calculated a set of soundscape features using sound detection algorithms and acoustic indices. These soundscape features were then correlated with historical data on ecological condition collected annually at each site since at least 2005.

Results/Conclusions

Three main drivers of regime shifts in kelp forests—the density of sea urchins, the density of kelp, and the diversity of fish—significantly related to different soundscape features. Sea urchin density positively correlated (p=0.005) and kelp density negatively correlated (p=0.022) with the number of shrimp snaps per minute. Fish species richness (p=0.014) positively correlated with sound pressure levels from 50-750 Hz, the frequency range containing most fish vocalizations. We found few significant differences between the soundscape features of protected and unprotected habitats, though this could have resulted from insufficient replicates. To investigate mechanisms underlying the relationship between kelp forest soundscapes and habitat condition, we introduce the concept of a “sound web”, which refers to evaluating trophic structure and regime shifts through the relative abundance of sounds made by predators (fish, lobster) and prey (sea urchins, shrimp).

This study demonstrates that marine soundscapes indicate the condition of kelp forests, which are increasingly vulnerable to destruction from sea urchin overgrazing. If snapping shrimp activity is indeed a reliable indicator of sea urchin and kelp density, then managers have a tool to rapidly assess habitat condition at far greater spatial scales than is currently possible with diver surveys.