This dataset contains fish chorusing event logs used in the study "Fish chorusing patterns in California National Marine Sanctuaries", published in Marine Ecology Progress Series (Kim et al. 2025). The logs include the start and end times of fish chorusing events detected across nine passive acoustic monitoring sites located in the Monterey Bay, Channel Islands, and Chumash Heritage National Marine Sanctuaries from 2018–2022. Chorus detections are categorized by species, including bocaccio rockfish (Sebastes paucispinis), plainfin midshipman (Porichthys notatus), white seabass (Atractoscion nobilis), and two unidentified fish types. 

Through the Sanctuary Soundscape Monitoring project, we collected passive acoustic data across 9 sites, each recording for ~2 years for a cumulative 18.3 recording-years. We manually analyzed acoustic data (logging start and end times of each chorus type) for fish chorusing in Monterey Bay, Chumash Heritage, and Channel Islands National Marine Sanctuaries.

We collected passive acoustic data using SoundTraps (Soundtrap ST500 and Soundtrap ST600, Ocean Instruments, NZ) at three sites in MBNMS, one in originally proposed CHNMS, and five sites in CINMS over multiple deployments from 2018-2023. Sites were named after location, with Channel Islands abbreviated to “CI,” Chumash Heritage to “CH,” and Monterey Bay to “MB,” and numbers allowing for further differentiation between recording sites, and then further number indicating deployment numbers.  Raw WAV files underwent compression for long-term spectral averages (LTSAs), employing 1000-point fast Fourier transforms (FFTs), Hanning window, no overlap, and 5-s and 1-Hz resolution. LTSAs were generated using Triton, a custom MATLAB (Mathworks, Natick, MA) software (Triton Github, 2023; Wiggins and Hildebrand, 2007). We defined a fish chorus as sustained acoustic activity that exceeded ambient noise levels by at least 3 dB re 1 mPa2/Hz for an extended period (minutes to hours), within species-specific frequency bands between 10 and 1000 Hz (Cato 1978, Parsons et al. 2016, Pagniello et al. 2019). Choruses were classified as originating from fish by analyzing similarity of acoustic features to other reported fish choruses and calls, including frequency content, timing, call duration, and the number of repeated pulses when calls were present (Parsons et al. 2016, Pagniello et al. 2019). In Triton, we manually scanned LTSAs for fish chorusing presence. We scanned LTSAs from 1 to 1000 Hz in species-specific frequency bands, using a viewing length of 24 h and 40% brightness and 120% contrast. We utilized ‘Logger,’ a Triton software package, to log start and end times of choruses, when sound levels were at least 3 dB re 1 mPa2/Hz above ambient noise (Wirth & Warren 2020). Selecting a 3 dB re 1 mPa2/Hz threshold, or a doubling of acoustic intensity, captured a biologically meaningful rise in acoustic activity well above ambient noise, aligning with fish calling studies that highlight periods of activity significantly above ambient levels (Cato 1978, Luczkovich et al. 1999, Wirth & Warren 2020). In cases of uncertainty, we scanned 30 s spectrograms (700 pt FFT, 90% overlap, 40% brightness, 120% contrast) to identify individual fish calls and clarify chorus presence. We logged choruses as separate events if there was an interval of at least 20 min between them in the LTSA. We used a 20 min cutoff between chorusing events to reflect natural breaks we observed, and to accurately capture independence of acoustic events, consistent with practices in other bioacoustics studies (e.g. Rice et al. 2021). Choruses that have been previously identified to species level were documented. Choruses that had no prior species identification, but contained calls previously identified as belonging to a particular species (during time periods without other choruses present), were attributed to that species. Choruses without species assignments, and no visible or identifiable calls, were labeled as ‘unidentified fish’ choruses. Unidentified fish choruses were labeled ‘UF#,’ with numbers referring to peak frequencies of calls or dominant frequencies of choruses when no calls were visible. For more details see Kim et al. 2025.