Where's Whaledo: A software toolkit for array localization of animal vocalizations.

Where's Whaledo is a software toolkit that uses a combination of automated processes and user interfaces to greatly accelerate the process of reconstructing animal tracks from arrays of passive acoustic recording devices. Passive acoustic localization is a non-invasive yet powerful way to contribute to species conservation. By tracking animals through their acoustic signals, important information on diving patterns, movement behavior, habitat use, and feeding dynamics can be obtained. This method is useful for helping to understand habitat use, observe behavioral responses to noise, and develop potential mitigation strategies. Animal tracking using passive acoustic localization requires an acoustic array to detect signals of interest, associate detections on various receivers, and estimate the most likely source location by using the time difference of arrival (TDOA) of sounds on multiple receivers. Where's Whaledo combines data from two small-aperture volumetric arrays and a variable number of individual receivers. In a case study conducted in the Tanner Basin off Southern California, we demonstrate the effectiveness of Where's Whaledo in localizing groups of Ziphius cavirostris. We reconstruct the tracks of six individual animals vocalizing concurrently and identify Ziphius cavirostris tracks despite being obscured by a large pod of vocalizing dolphins.

Bayesian detection and tracking of odontocetes in 3-D from their echolocation clicks.

Localization and tracking of marine animals can reveal key insights into their behaviors underwater that would otherwise remain unexplored. A promising nonintrusive approach to obtaining location information of marine animals is to process their bioacoustic signals, which are passively recorded using multiple hydrophones. In this paper, a data processing chain that automatically detects and tracks multiple odontocetes (toothed whales) in three dimensions (3-D) from their echolocation clicks recorded with volumetric hydrophone arrays is proposed. First, the time-difference-of-arrival (TDOA) measurements are extracted with a generalized cross-correlation that whitens the received acoustic signals based on the instrument noise statistics. Subsequently, odontocetes are tracked in the TDOA domain using a graph-based multi-target tracking (MTT) method to reject false TDOA measurements and close gaps of missed detections. The resulting TDOA estimates are then used by another graph-based MTT stage that estimates odontocete tracks in 3-D. The tracking capability of the proposed data processing chain is demonstrated on real acoustic data provided by two volumetric hydrophone arrays that recorded echolocation clicks from Cuvier's beaked whales (Ziphius cavirostris). Simulation results show that the presented MTT method using 3-D can outperform an existing approach that relies on manual annotation.

Changes in sea ice and range expansion of sperm whales in the eclipse sound region of Baffin Bay, Canada.

Sperm whales (Physeter macrocephalus) are a cosmopolitan species but are only found in ice-free regions of the ocean. It is unknown how their distribution might change in regions undergoing rapid loss of sea ice and ocean warming like Baffin Bay in the eastern Canadian Arctic. In 2014 and 2018, sperm whales were sighted near Eclipse Sound, Baffin Bay: the first recorded uses of this region by sperm whales. In this study, we investigate the spatiotemporal distribution of sperm whales near Eclipse Sound using visual and acoustic data. We combine several published open-source, data sets to create a map of historical sperm whale presence in the region. We use passive acoustic data from two recording sites between 2015 and 2019 to investigate more recent presence in the region. We also analyze regional trends in sea ice concentration (SIC) dating back to 1901 and relate acoustic presence of sperm whales to the mean SIC near the recording sites. We found no records of sperm whale sightings near Eclipse Sound outside of the 2014/2018 observations. Our acoustic data told a different story, with sperm whales recorded yearly from 2015 to 2019 with presence in the late summer and fall months. Sperm whale acoustic presence increased over the 5-year study duration and was closely related to the minimum SIC each year. Sperm whales, like other cetaceans, are ecosystem sentinels, or indicators of ecosystem change. Increasing number of days with sperm whale presence in the Eclipse Sound region could indicate range expansion of sperm whales as a result of changes in sea ice. Monitoring climate change-induced range expansion in this region is important to understand how increasing presence of a top-predator might impact the Arctic food web.

Performance metrics for marine mammal signal detection and classification.

Automatic algorithms for the detection and classification of sound are essential to the analysis of acoustic datasets with long duration. Metrics are needed to assess the performance characteristics of these algorithms. Four metrics for performance evaluation are discussed here: receiver-operating-characteristic (ROC) curves, detection-error-trade-off (DET) curves, precision-recall (PR) curves, and cost curves. These metrics were applied to the generalized power law detector for blue whale D calls [Helble, Ierley, D'Spain, Roch, and Hildebrand (2012). J. Acoust. Soc. Am. 131(4), 2682-2699] and the click-clustering neural-net algorithm for Cuvier's beaked whale echolocation click detection [Frasier, Roch, Soldevilla, Wiggins, Garrison, and Hildebrand (2017). PLoS Comp. Biol. 13(12), e1005823] using data prepared for the 2015 Detection, Classification, Localization and Density Estimation Workshop. Detection class imbalance, particularly the situation of rare occurrence, is common for long-term passive acoustic monitoring datasets and is a factor in the performance of ROC and DET curves with regard to the impact of false positive detections. PR curves overcome this shortcoming when calculated for individual detections and do not rely on the reporting of true negatives. Cost curves provide additional insight on the effective operating range for the detector based on the a priori probability of occurrence. Use of more than a single metric is helpful in understanding the performance of a detection algorithm.

Echolocation click discrimination for three killer whale ecotypes in the Northeastern Pacific.

Three killer whale ecotypes are found in the Northeastern Pacific: residents, transients, and offshores. These ecotypes can be discriminated in passive acoustic data based on distinct pulsed call repertoires. Killer whale acoustic encounters for which ecotypes were assigned based on pulsed call matching were used to characterize the ecotype-specific echolocation clicks. Recordings were made using seafloor-mounted sensors at shallow (∼120 m) and deep (∼1400 m) monitoring locations off the coast of Washington state. All ecotypes' echolocation clicks were characterized by energy peaks between 12 and 19 kHz, however, resident clicks featured sub peaks at 13.7 and 18.8 kHz, while offshore clicks had a single peak at 14.3 kHz. Transient clicks were rare and were characterized by lower peak frequencies (12.8 kHz). Modal inter-click intervals (ICIs) were consistent but indistinguishable for resident and offshore killer whale encounters at the shallow site (0.21-0.22 s). Offshore ICIs were longer and more variable at the deep site, and no modal ICI was apparent for the transient ecotype. Resident and offshore killer whale ecotype may be identified and distinguished in large passive acoustic datasets based on properties of their echolocation clicks, however, transient echolocation may be unsuitable in isolation as a cue for monitoring applications.

Acoustic localization, validation, and characterization of Rice's whale calls.

The recently named Rice's whale in the Gulf of Mexico is one of the most endangered whales in the world, and improved knowledge of spatiotemporal occurrence patterns is needed to support their recovery and conservation. Passive acoustic monitoring methods for determining spatiotemporal occurrence patterns require identifying the species' call repertoire. Rice's whale call repertoire remains unvalidated though several potential call types have been identified. This study uses sonobuoys and passive acoustic tagging to validate the source of potential call types and to characterize Rice's whale calls. During concurrent visual and acoustic surveys, acoustic-directed approaches were conducted to obtain visual verifications of sources of localized sounds. Of 28 acoustic-directed approaches, 79% led to sightings of balaenopterid whales, of which 10 could be positively identified to species as Rice's whales. Long-moan calls, downsweep sequences, and tonal-sequences are attributed to Rice's whales based on these matches, while anthropogenic sources are ruled out. A potential new call type, the low-frequency downsweep sequence, is characterized from tagged Rice's whale recordings. The validation and characterization of the Rice's whale call repertoire provides foundational information needed to use passive acoustic monitoring for better understanding and conservation of these critically endangered whales.

DetEdit: A graphical user interface for annotating and editing events detected in long-term acoustic monitoring data.

Passive acoustic monitoring has become an important data collection method, yielding massive datasets replete with biological, environmental and anthropogenic information. Automated signal detectors and classifiers are needed to identify events within these datasets, such as the presence of species-specific sounds or anthropogenic noise. These automated methods, however, are rarely a complete substitute for expert analyst review. The ability to visualize and annotate acoustic events efficiently can enhance scientific insights from large, previously intractable datasets. A MATLAB-based graphical user interface, called DetEdit, was developed to accelerate the editing and annotating of automated detections from extensive acoustic datasets. This tool is highly-configurable and multipurpose, with uses ranging from annotation and classification of individual signals or signal-clusters and evaluation of signal properties, to identification of false detections and false positive rate estimation. DetEdit allows users to step through acoustic events, displaying a range of signal features, including time series of received levels, long-term spectral averages, time intervals between detections, and scatter plots of peak frequency, RMS, and peak-to-peak received levels. Additionally, it displays either individual, or averaged sound pressure waveforms, and power spectra within each acoustic event. These views simultaneously provide analysts with signal-level detail and encounter-level context. DetEdit creates datasets of signal labels for further analyses, such as training classifiers and quantifying occurrence, abundances, or trends. Although designed for evaluating underwater-recorded odontocete echolocation click detections, DetEdit can be adapted to almost any stereotyped impulsive signal. Our software package complements available tools for the bioacoustic community and is provided open source at https://github.com/MarineBioAcousticsRC/DetEdit.

An empirical model for wind-generated ocean noise.

An empirical model for wind-generated underwater noise is presented that was developed using an extensive dataset of acoustic field recordings and a global wind model. These data encompass more than one hundred years of recording-time and capture high wind events, and were collected both on shallow continental shelves and in open ocean deep-water settings. The model aims to explicitly separate noise generated by wind-related sources from noise produced by anthropogenic sources. Two key wind-related sound-generating mechanisms considered are: surface wave and turbulence interactions, and bubble and bubble cloud oscillations. The model for wind-generated noise shows small frequency dependence (5 dB/decade) at low frequencies (10-100 Hz), and larger frequency dependence (∼15 dB/decade) at higher frequencies (400 Hz-20 kHz). The relationship between noise level and wind speed is linear for low wind speeds (<3.3 m/s) and increases to a higher power law (two or three) at higher wind speeds, suggesting a transition between surface wave/turbulence and bubble source mechanisms. At the highest wind speeds (>15 m/s), noise levels begin to decrease at high frequencies (>10 kHz), likely due to interaction between bubbles and screening of noise radiation in the presence of high-density bubble clouds.

Seal bomb explosion sound source characterization.

Small explosive charges, called seal bombs, used by commercial fisheries to deter marine mammals from depredation and accidental bycatch during fishing operations, produce high level sounds that may negatively impact nearby animals. Seal bombs were exploded underwater and recorded at various ranges with a calibrated hydrophone to characterize the pulse waveforms and to provide appropriate propagation loss models for source level (SL) estimates. Waveform refraction became important at about 1500 m slant range with approximately spherical spreading losses observed at shorter ranges. The SL for seal bombs was estimated to be 233 dB re 1 µPa m; however, for impulses such as explosions, better metrics integrate over the pulse duration, accounting for the total energy in the pulse, including source pressure impulse, estimated as 193 Pa m s, and sound exposure source level, estimated as 197 dB re 1 µPa2 m2 s over a 2 ms window. Accounting for the whole 100 ms waveform, including the bubble pulses and sea surface reflections, sound exposure source level was 203 dB re 1 µPa2 m2 s. Furthermore, integrating the energy over an entire event period of multiple explosions (i.e., cumulative sound exposure level) should be considered when evaluating impact.

Using context to train time-domain echolocation click detectors.

This work demonstrates the effectiveness of using humans in the loop processes for constructing large training sets for machine learning tasks. A corpus of over 57 000 toothed whale echolocation clicks was developed by using a permissive energy-based echolocation detector followed by a machine-assisted quality control process that exploits contextual cues. Subsets of these data were used to train feed forward neural networks that detected over 850 000 echolocation clicks that were validated using the same quality control process. It is shown that this network architecture performs well in a variety of contexts and is evaluated against a withheld data set that was collected nearly five years apart from the development data at a location over 600 km distant. The system was capable of finding echolocation bouts that were missed by human analysts, and the patterns of error in the classifier consist primarily of anthropogenic sources that were not included as counter-training examples. In the absence of such events, typical false positive rates are under ten events per hour even at low thresholds.

Automated classification of dolphin echolocation click types from the Gulf of Mexico.

Delphinids produce large numbers of short duration, broadband echolocation clicks which may be useful for species classification in passive acoustic monitoring efforts. A challenge in echolocation click classification is to overcome the many sources of variability to recognize underlying patterns across many detections. An automated unsupervised network-based classification method was developed to simulate the approach a human analyst uses when categorizing click types: Clusters of similar clicks were identified by incorporating multiple click characteristics (spectral shape and inter-click interval distributions) to distinguish within-type from between-type variation, and identify distinct, persistent click types. Once click types were established, an algorithm for classifying novel detections using existing clusters was tested. The automated classification method was applied to a dataset of 52 million clicks detected across five monitoring sites over two years in the Gulf of Mexico (GOM). Seven distinct click types were identified, one of which is known to be associated with an acoustically identifiable delphinid (Risso's dolphin) and six of which are not yet identified. All types occurred at multiple monitoring locations, but the relative occurrence of types varied, particularly between continental shelf and slope locations. Automatically-identified click types from autonomous seafloor recorders without verifiable species identification were compared with clicks detected on sea-surface towed hydrophone arrays in the presence of visually identified delphinid species. These comparisons suggest potential species identities for the animals producing some echolocation click types. The network-based classification method presented here is effective for rapid, unsupervised delphinid click classification across large datasets in which the click types may not be known a priori.

Evolution of herbivory in Drosophilidae linked to loss of behaviors, antennal responses, odorant receptors, and ancestral diet.

Herbivory is a key innovation in insects, yet has only evolved in one-third of living orders. The evolution of herbivory likely involves major behavioral changes mediated by remodeling of canonical chemosensory modules. Herbivorous flies in the genus Scaptomyza (Drosophilidae) are compelling species in which to study the genomic architecture linked to the transition to herbivory because they recently evolved from microbe-feeding ancestors and are closely related to Drosophila melanogaster. We found that Scaptomyza flava, a leaf-mining specialist on plants in the family (Brassicaceae), was not attracted to yeast volatiles in a four-field olfactometer assay, whereas D. melanogaster was strongly attracted to these volatiles. Yeast-associated volatiles, especially short-chain aliphatic esters, elicited strong antennal responses in D. melanogaster, but weak antennal responses in electroantennographic recordings from S. flava. We sequenced the genome of S. flava and characterized this species' odorant receptor repertoire. Orthologs of odorant receptors, which detect yeast volatiles in D. melanogaster and mediate critical host-choice behavior, were deleted or pseudogenized in the genome of S. flava. These genes were lost step-wise during the evolution of Scaptomyza. Additionally, Scaptomyza has experienced gene duplication and likely positive selection in paralogs of Or67b in D. melanogaster. Olfactory sensory neurons expressing Or67b are sensitive to green-leaf volatiles. Major trophic shifts in insects are associated with chemoreceptor gene loss as recently evolved ecologies shape sensory repertoires.

Odorant receptors and antennal lobe morphology offer a new approach to understanding olfaction in the Asian longhorned beetle.

The Asian longhorned beetle Anoplophora glabripennis (Motchulsky) is an exotic forest pest that has repeatedly invaded North America and Europe from Asia, and has the potential to kill millions of trees and cause billions of dollars in damage. Traps baited with an attractive mixture of volatile organic compounds from hosts have been of limited success in monitoring invasion sites. We propose that lures might be improved through studying the olfactory system of adult beetles, especially the gene family of odorant receptors (ORs) and the structure of the antennal lobes of the brain. Here, we report identification of 132 ORs in the genome of A. glabripennis (inclusive of one Orco gene and 11 pseudogenes), some of which are orthologous to known pheromone receptors of other cerambycid beetles. We also identified three ORs that are strongly biased toward expression in the female transcriptome, and a single OR strongly biased toward males. Three-dimensional reconstruction of the antennal lobes of adults suggested a male-specific macroglomerulus and several enlarged glomeruli in females. We predict that functional characterization of ORs and glomeruli will lead to identification of key odorants in the life history of A. glabripennis that may aid in monitoring and controlling future invasions.

Deep-water measurements of container ship radiated noise signatures and directionality.

Underwater radiated noise from merchant ships was measured opportunistically from multiple spatial aspects to estimate signature source levels and directionality. Transiting ships were tracked via the Automatic Identification System in a shipping lane while acoustic pressure was measured at the ships' keel and beam aspects. Port and starboard beam aspects were 15°, 30°, and 45° in compliance with ship noise measurements standards [ANSI/ASA S12.64 (2009) and ISO 17208-1 (2016)]. Additional recordings were made at a 10° starboard aspect. Source levels were derived with a spherical propagation (surface-affected) or a modified Lloyd's mirror model to account for interference from surface reflections (surface-corrected). Ship source depths were estimated from spectral differences between measurements at different beam aspects. Results were exemplified with a 4870 and a 10 036 twenty-foot equivalent unit container ship at 40%-56% and 87% of service speeds, respectively. For the larger ship, opportunistic ANSI/ISO broadband levels were 195 (surface-affected) and 209 (surface-corrected) dB re 1 µPa2 1 m. Directionality at a propeller blade rate of 8 Hz exhibited asymmetries in stern-bow (<6 dB) and port-starboard (<9 dB) direction. Previously reported broadband levels at 10° aspect from McKenna, Ross, Wiggins, and Hildebrand [(2012b). J. Acoust. Soc. Am. 131, 92-103] may be ∼12 dB lower than respective surface-affected ANSI/ISO standard derived levels.

Ocean ambient sound south of Bermuda and Panama Canal traffic.

Comparisons of current and historic ocean ambient noise levels are rare, especially in the North Atlantic. Recent (2013-2014) monthly patterns in ocean ambient sound south of Bermuda were compared to those recorded at the same location in 1966. Additionally, trends in ocean traffic, in particular, Panama Canal traffic, over this time were also investigated. One year of ocean ambient noise measurements were collected in 1966 using cabled, omnidirectional hydrophones at the U.S. Navy Tudor Hill Laboratory in Bermuda, and repeat measurements were collected at the same location from June 2013-May 2014 using a High-frequency Acoustic Recording Package. Average monthly pressure spectrum levels at 44 Hz increased 2.8 ± 0.8 dB from 1966 to 2013, indicating an average increase of 0.6 dB/decade. This low level of increase may be due to topographic shielding at this site, limiting it to only southern exposure, and the limit in the number of ship transits through the Panama Canal, which did not change substantially during this time. The impending expansion of the Canal, which will enable the transit of larger ships at twice the current rate, is likely to lead to a substantial increase in ocean ambient sound at this location in the near future.

Gulf of Mexico low-frequency ocean soundscape impacted by airguns.

The ocean soundscape of the Gulf of Mexico (GOM) has not been well-studied, although it is an important habitat for marine mammals, including sperm and beaked whales, many dolphin species, and a potentially endangered baleen whale species. The GOM is also home to high levels of hydrocarbon exploration and extraction, heavily used commercial shipping ports, and significant fishery industry activity, all of which are known contributors to oceanic noise. From 2010-2013, the soundscape of three deep and two shallow water sites in the GOM were monitored over 10 - 1000 Hz. Average sound pressure spectrum levels were high, >90 dB re 1 µPa(2)/Hz at <40 Hz for the deep water sites and were associated with noise from seismic exploration airguns. More moderate sound pressure levels, <55 dB re 1 µPa(2)/Hz at >700 Hz, were present at a shallow water site in the northeastern Gulf, removed from the zone of industrial development and bathymetrically shielded from deep water anthropogenic sound sources. During passage of a high wind event (Hurricane Isaac, 2012), sound pressure levels above 200 Hz increased with wind speed, but at low frequencies (<100 Hz) sound pressure levels decreased owing to absence of noise from airguns.

Delphinid echolocation click detection probability on near-seafloor sensors.

The probability of detecting echolocating delphinids on a near-seafloor sensor was estimated using two Monte Carlo simulation methods. One method estimated the probability of detecting a single click (cue counting); the other estimated the probability of detecting a group of delphinids (group counting). Echolocation click beam pattern and source level assumptions strongly influenced detectability predictions by the cue counting model. Group detectability was also influenced by assumptions about group behaviors. Model results were compared to in situ recordings of encounters with Risso's dolphin (Grampus griseus) and presumed pantropical spotted dolphin (Stenella attenuata) from a near-seafloor four-channel tracking sensor deployed in the Gulf of Mexico (25.537°N 84.632°W, depth 1220 m). Horizontal detection range, received level and estimated source level distributions from localized encounters were compared with the model predictions. Agreement between in situ results and model predictions suggests that simulations can be used to estimate detection probabilities when direct distance estimation is not available.

Three-dimensional tracking of Cuvier's beaked whales' echolocation sounds using nested hydrophone arrays.

Cuvier's beaked whales (Ziphius cavirostris) were tracked using two volumetric small-aperture (∼1 m element spacing) hydrophone arrays, embedded into a large-aperture (∼1 km element spacing) seafloor hydrophone array of five nodes. This array design can reduce the minimum number of nodes that are needed to record the arrival of a strongly directional echolocation sound from 5 to 2, while providing enough time-differences of arrivals for a three-dimensional localization without depending on any additional information such as multipath arrivals. To illustrate the capabilities of this technique, six encounters of up to three Cuvier's beaked whales were tracked over a two-month recording period within an area of 20 km(2) in the Southern California Bight. Encounter periods ranged from 11 min to 33 min. Cuvier's beaked whales were found to reduce the time interval between echolocation clicks while alternating between two inter-click-interval regimes during their descent towards the seafloor. Maximum peak-to-peak source levels of 179 and 224 dB re 1 µPa @ 1 m were estimated for buzz sounds and on-axis echolocation clicks (directivity index = 30 dB), respectively. Source energy spectra of the on-axis clicks show significant frequency components between 70 and 90 kHz, in addition to their typically noted FM upsweep at 40-60 kHz.

Description and clustering of echolocation signals of Commerson's dolphins (Cephalorhynchus commersonii) in Bahía San Julián, Argentina.

Commerson's dolphins (Cephalorhynchus commersonii) inhabit coastal waters of Southern South America and Kerguelen Islands. Limited information exists about the acoustic repertoire of this species in the wild. Here, echolocation signals from free-ranging Commerson's dolphins were recorded in Bahía San Julián, Argentina. Signal parameters were calculated and a cluster analysis was made on 3180 regular clicks. Three clusters were obtained based on peak frequency (129, 137, and 173 kHz) and 3 dB bandwidth (8, 6, and 5 kHz). The 428 buzz clicks were analyzed separately. They consisted of clicks emitted with a median inter-click interval of 3.5 ms, peak frequency at 131 kHz, 3 dB bandwidth of 9 kHz, 10 dB bandwidth of 18 kHz, and duration of 56 µs. Buzz clicks were significantly shorter and with a lower peak frequency and a broader bandwidth than most of the regular clicks. This study provided the first description of different echolocation signals, including on- and off-axis signals, recorded from Commerson's dolphins in the wild, most likely as a result of animals at several distances and orientations to the recording device. This information could be useful while doing passive acoustic monitoring.