Data from: Long-term monitoring of Ziphius cavirostris behavior using 3D tracking from fixed hydrophone arrays off Southern California
Data files
Dec 05, 2025 version files 62.19 MB
-
DRYAD_upload_10212025.zip
62.18 MB
-
README.md
5.19 KB
Abstract
Goose-beaked whales (Ziphius cavirostris) are a deep-diving toothed whale species and top predators in deep sea ecosystems. Much is yet to be learned about their social and foraging strategies due to their elusive behavior, but this information is increasingly relevant given their demonstrated behavioral changes in association with anthropogenic sound. This study used direction-of-arrival (DOA) localization to track the position of goose-beaked whales from echolocation clicks recorded on seafloor-mounted hydrophone arrays offshore Southern California. Overall, 2738 tracks of diving goose-beaked whales were processed from acoustic recordings collected at four long-term monitoring sites between 2018 and 2023. Results highlight distinct spatial use patterns driven by bathymetric features at each site, with whales foraging closer to the seafloor at sites with complex bathymetry and showing a preference for certain bathymetric features. Group sizes at depth ranged from 1 to 9 individuals with a mean of 2.34 and exhibited site-specific seasonal variability as well as a strong diel trend at one site. During many of these encounters, individuals exhibited highly coordinated behaviors. This study demonstrates the value of long-term passive acoustic tracking for studying elusive, deep-diving species and provides significant advancements in understanding goose-beaked whale behavior at depth over long time scales.
Dataset DOI: 10.5061/dryad.280gb5n1r
Description of the data and file structure
Diving whales were localized using the MATLAB-based software package Where's Whaledo (Snyder et al., 2024; https://github.com/MarineBioAcousticsRC/Wheres-Whaledo). Acoustic data were recorded on small-aperture 4-hydrophone arrays at four sites offshore Southern California. This acoustic data, for times when goose-beaked whales were present, was high-pass filtered with a cutoff frequency of 20 kHz to detect click candidates with a peak-to-peak received level ≥ 112 dB re 1 μPa^2. The acoustic data in a 50μs window around the click candidate were cross-correlated, and the time delay between each hydrophone pair was used to estimate the direction of arrival. These DOAs were represented as azimuth and elevation angles, manually inspected and cleaned, and assigned labels to individual whale tracks. 3D whale positions relative to the receivers were calculated from the intersection of the cleaned DOAs from both 4-hydrophone arrays.
Goose-beaked whale encounters were identified using a customized machine learning workflow (Frasier et al., 2017; Frasier et al., 2021). Group size was modeled in relation to temporal covariates.
Files and variables
File: DRYAD_upload_10212025.zip
Description:
Ziphius_cavirostris_localized_whales: Tracked Ziphius cavirostris. Divided by site, then into encounters. Each encounter contains an .nc file with the following information:
- deployment_id: Naming schema is region (SOCAL), site (W, H, E, or N), and deployment number
- author: Corresponding author name (Lauren Baggett)
- author_contact: Email address for corresponding author (lbaggett@ucsd.edu)
- doi: DOI Permalink to this dataset (https://doi.org/10.5061/dryad.280gb5n1r)
- effort_start: Start time of tracking effort for this encounter, in UTC. Encounters calculated from single-channel presence/absence detections, see Ziphius cavirostris encounter tables for all encounters per deployment.
- effort_end: End time of tracking effort for this encounter, in UTC. Encounters calculated from single-channel presence/absence detections, see Ziphius cavirostris encounter tables for all encounters per deployment.
- species: Species tracked in this encounter (all are Ziphius cavirostris)
- reference_latitude: Reference latitude (°N) for localized positions (both localized and smoothed)
- reference_longitude: Reference longitude (°E) for localized positions (both localized and smoothed)
- reference_depth_m: Reference depth (in meters) for localized positions (both localized and smoothed)
- instrument_type: Type of instrument used to generate these tracks (4-hydrophone HARPs)
- file_write_date: Date that the .nc file was generated (UTC)
- time: Timestamps of localized points (UTC, in POSIX format)
- position: XYZ position of localized points in meters relative to reference location (see reference lat/lon/depth above)
- smoothed_position: XYZ position of smoothed localized points in meters relative to reference location (see reference lat/lon/depth above)
- confidence_interval: Upper and lower limits (x upper, x lower, y upper, y lower, z upper, z lower in meters) for the calculated confidence interval of the localized XYZ positions
- whale_number: Integer identification number corresponding to a localized individual whale in the encounter. Each entry corresponds to a single timestamp.
Ziphius_cavirostris_binned_presence_GAMs: Summed group size for non-autocorrelated time bins. 'dayNight' is normalized time of day where sunrise is at 0, and sunset is at -1/1 (positive numbers are daytime, negative are nighttime). 'jd' is Julian date. 'presBins' are the summed number of 5-minute bins with Zc presence from the single hydrophone data.
Ziphius_cavirostris_encounter_tables: Zc encounters from the single hydrophone data at each deployment. 'startEnc' and 'endEnc' are the start and end times of the encounter, and 'startSnipped' and 'endSnipped' are padded by one hour. These are the time windows over which tracking software was run. 'encDur' is the encounter duration, 'encN' is the encounter number assigned to that encounter, and 'startNumEnc' and 'endNumEnc' are the start and end times in MATLAB datenums. Encounters that do not have a matching localized whale file did not yield any tracks.
Any questions regarding this data should be sent via email to lbaggett@ucsd.edu
Code/software
The Where’sWhaledo software package can be found on GitHub (https://github.com/MarineBioAcousticsRC/Wheres-Whaledo). Code related to track analysis can also be found on GitHub (https://github.com/laurenbaggett/Zc-Tracking).
