Skip to main content
Dryad

Pacific white-sided dolphin hourly binned echolocation clicks

Cite this dataset

Alksne, Michaela N. (2024). Pacific white-sided dolphin hourly binned echolocation clicks [Dataset]. Dryad. https://doi.org/10.5061/dryad.95x69p8rj

Abstract

This study investigates the biogeographic patterns of Pacific white-sided dolphins (Lagenorhynchus obliquidens) in the Eastern North Pacific based on long-term passive acoustic records (2005-2021). We aim to elucidate the ecological and behavioral significance of distinct echolocation click types and their implications for population delineation, geographic distribution, environmental adaptation, and management. Over 50 cumulative years of Passive Acoustic Monitoring (PAM) data from 14 locations were analyzed using a deep neural network to classify two distinct Pacific white-sided dolphin echolocation click types. The study assessed spatial, diel, seasonal, and interannual patterns of the two click types, correlating them with major environmental drivers such as the El Niño Southern Oscillation and the North Pacific Gyre Oscillation, and modeling long-term spatial-seasonal patterns. Distinct spatial, seasonal, and diel patterns were observed for each click type. Significant biogeographical shifts in presence were observed following the 2014-2016 marine heatwave event. Distinct spatial distributions of the two click types support the hypothesis that Pacific white-sided dolphins produce population-specific echolocation clicks. Seasonal and diel patterns suggest spatio-temporal niche partitioning between the populations in Southern California. Lastly, interannual changes, notably initiated during the 2014-2016 marine heatwave, indicate climate-driven range expansions and contractions related to the gradual tropicalization of the Southern California Bight.

README: Pacific white-sided dolphin hourly binned echolocation clicks

https://doi.org/10.5061/dryad.95x69p8rj

Each CSV file contains the hourly acoustic presence of Pacific white-sided dolphin echolocation clicks. The files are formatted such that the click type and location are stored in the file header. For instance, "SCB_LoA.csv" represents the hourly presence of the LoA click type at recording station SCB_M. 

The recording effort has been included here as a CSV file titled "PWD_effort.csv". Therefore, a user can cross-reference the recording location, recording effort, and time series name. If the one-click type was not detected at a given recording site, then that time series was not included.

Each of the datasets contains two columns:

  • The first column is the time bin in hours (deploymentHour)
  • The second is the number of click-positive minutes in that hour (Click_pos_min_per_hour), with a maximum of 60 positive minutes. If any number of clicks were present in a given minute, that minute would be considered "positive". Thus, the number of click-positive minutes in an hour could be summed up for each hour of recording effort.
  • The hourly time bins begin at the start time of recording the first deployment for a given location. During periods of no recording effort, no time bins are present.

Acronyms:

  • LoA - Type A click type produced by Pacific white-sided dolphins with spectral peaks at 22, 27.5, and 39 kHz
  • LoB - Type B click type produced by Pacific white-sided dolphins with spectral peaks at 22, 26, and 37 kHz
  • GofAK - Gulf of Alaska recording station
  • OCNMS - Olympic Coast National Marine Sanctuary recording station
  • PS - Point Sur recording station
  • CINMS - Channel Islands National Marine Sanctuary recording station
  • SCB_M - Southern California Bight recording station M
  • SCB_H -  Southern California Bight recording station H
  • SCB_N -  Southern California Bight recording station N
  • SCB_T -  Southern California Bight recording station T
  • SCB_P -  Southern California Bight recording station P
  • SCB_U -  Southern California Bight recording station U
  • Baja_GI - Baja California Guadalupe Island recording station
  • GofCA - Gulf of California recording station

Methods

Raw acoustic data was passed through a click detector which returned all acoustic signals within an expected frequency range and duration of odontocete echolocation clicks. An unsupervised clustering algorithm was run on the detections to group them into 5-minute bin-level averages. Cluster bins were then labeled as one of six categories by a trained neural network. Clusters labeled as either one of two Pacific white-sided dolphin click types were extracted and manually verified. Verified pacific white-sided dolphin detections were then binned into 'click-positive minutes per hour', where a click positive minute was a minute that contained any number of clicks. The timeseries of click-positive minutes per hour, for each click type, at multiple long-term recording locations, is included here. 

Funding

United States Department of Defense, National Science and Engineering Graduate Student Fellowship