Acoustic classification of dolphins in the California Current using whistles, echolocation clicks, and burst pulses

Abstract

Passive acoustic monitoring of dolphins is limited by our ability to classify calls to species. Significant overlap in call characteristics among many species, combined with a wide range of call types and acoustic behavior, makes classification of calls to species challenging. Here, we introduce BANTER, a compound acoustic classification method for dolphins that utilizes information from all call types produced by dolphins rather than a single call type, as has been typical for acoustic classifiers. Output from the passive acoustic monitoring software, PAMGuard, was used to create independent classifiers for whistles, echolocation clicks, and burst pulses, which were then merged into a final, compound classifier for each species. Classifiers for five species found in the California Current ecosystem were trained and tested using 153 single‐species acoustic events recorded during a 4.5 mo combined visual and acoustic shipboard cetacean survey off the west coast of the United States. Correct classification scores for individual species ranged from 71% to 92%, with an overall correct classification score of 84% for all five species. The conceptual framework of this approach easily lends itself to other species and study areas as well as to noncetacean taxa.

Whistle Classification of Sympatric False Killer Whale Populations in Hawaiian Waters Yields Low Accuracy Rates

“…We applied the RF machine learning classification method to analyze whistle characteristics of the three false killer whale populations because of its high performance with diverse variables, including prior work differentiating dolphin species based on their whistle characteristics (Pal, 2005;Cutler et al, 2007;Oswald, 2013;Keen et al, 2014;Li et al, 2016;Rankin et al, 2017). Overall, RF classification models poorly differentiated the three populations as is evident from the low correct classification rates and low kappa coefficients for each model.…”

“…Additional whistle data for all populations may increase classification performance to differentiate the populations with more confidence and allow further investigation into social and population structure as well as how the populations remain demographically independent. Future analyses may also incorporate characteristics of echolocation clicks to improve classification, hybrid versions of important variables (Rankin et al, 2017) or incorporate additional population or behavior variables (social cluster, group size, etc.) to better capture variability in whistle context and therefore whistle characteristics.…”

“…Acoustic-based detection and classification methods continue to improve for many cetacean species (Charif and Clark, 2009;Delarue et al, 2009;Roch et al, 2011;Baumann-Pickering et al, 2013;Rankin et al, 2017). Many dolphin species can be identified based on characteristics of their whistle and click vocalizations, and in some cases, population-level differences are evident (Rendell et al, 1999;Oswald et al, 2007;Soldevilla et al, 2008;Gannier et al, 2010;Azzolin et al, 2014;Baumann-Pickering et al, 2015).…”

Description and classification of echolocation clicks of Indian Ocean humpback (Sousa plumbea) and Indo-Pacific bottlenose (Tursiops aduncus) dolphins from Menai Bay, Zanzibar, East Africa

“…For example, clicks of melon-headed whales (Peponocephala electra), common bottlenose (Tursiops truncatus) and Gray's spinner (Stenellla longirostris longirostris) dolphins were separated using spectral parameters and discriminant function analysis providing 93%, 75% and 54% correct classification rates for the three delphinid species, respectively [14]. Furthermore, clicks of seven delphinid species, striped dolphin (Stenella coeruleoalba), long-beaked common dolphin (Delphinus capensis), short-beaked common dolphin (Delphinus delphis), Risso's dolphin (Grampus griseus), Pacific white-sided dolphin (Lagenorhynchus obliquidens), pilot whale (Globicephala macrorhynchus) and killer whale (Orcinus orca), off the coasts of Washington, Oregon and California were classified using the Random Forest classification model with overall correct classification score of 49%, which was significantly greater than that expected by chance for the seven species (14%) [12].…”

“…echolocation clicks for navigation, orientation and prey detection [7], and burst pulses for communication [8]. Whistles are highly variable at an individual level [9] whereas echolocation clicks (here on referred to as "clicks"), are more consistent and can be used for species classification [10][11][12][13]. However, some sympatric species of odontocetes produce similar clicks which can limit the effectiveness of PAM for species-specific studies, as acoustic species classification can be challenging [14].…”