Vulnerability to climate change of United States marine mammal stocks in the western North Atlantic, Gulf of Mexico, and Caribbean.

Climate change and climate variability are affecting marine mammal species and these impacts are projected to continue in the coming decades. Vulnerability assessments provide a framework for evaluating climate impacts over a broad range of species using currently available information. We conducted a trait-based climate vulnerability assessment using expert elicitation for 108 marine mammal stocks and stock groups in the western North Atlantic, Gulf of Mexico, and Caribbean Sea. Our approach combined the exposure (projected change in environmental conditions) and sensitivity (ability to tolerate and adapt to changing conditions) of marine mammal stocks to estimate vulnerability to climate change, and categorize stocks with a vulnerability index. The climate vulnerability score was very high for 44% (n = 47) of these stocks, high for 29% (n = 31), moderate for 20% (n = 22), and low for 7% (n = 8). The majority of stocks (n = 78; 72%) scored very high exposure, whereas 24% (n = 26) scored high, and 4% (n = 4) scored moderate. The sensitivity score was very high for 33% (n = 36) of these stocks, high for 18% (n = 19), moderate for 34% (n = 37), and low for 15% (n = 16). Vulnerability results were summarized for stocks in five taxonomic groups: pinnipeds (n = 4; 25% high, 75% moderate), mysticetes (n = 7; 29% very high, 57% high, 14% moderate), ziphiids (n = 8; 13% very high, 50% high, 38% moderate), delphinids (n = 84; 52% very high, 23% high, 15% moderate, 10% low), and other odontocetes (n = 5; 60% high, 40% moderate). Factors including temperature, ocean pH, and dissolved oxygen were the primary drivers of high climate exposure, with effects mediated through prey and habitat parameters. We quantified sources of uncertainty by bootstrapping vulnerability scores, conducting leave-one-out analyses of individual attributes and individual scorers, and through scoring data quality for each attribute. These results provide information for researchers, managers, and the public on marine mammal responses to climate change to enhance the development of more effective marine mammal management, restoration, and conservation activities that address current and future environmental variation and biological responses due to climate change.

Combining passive acoustic data from a towed hydrophone array with visual line transect data to estimate abundance and availability bias of sperm whales (Physeter macrocephalus).

Visual line transect (VLT) surveys are central to the monitoring and study of marine mammals. However, for cryptic species such as deep diving cetaceans VLT surveys alone suffer from problems of low sample sizes and availability bias where animals below the surface are not available to be detected. The advent of passive acoustic monitoring (PAM) technology offers important opportunities to observe deep diving cetaceans but statistical challenges remain particularly when trying to integrate VLT and PAM data. Herein, we present a general framework to combine these data streams to estimate abundance when both surveys are conducted simultaneously. Secondarily, our approach can also be used to derive an estimate of availability bias. We outline three methods that vary in complexity and data requirements which are (1) a simple distance sampling (DS) method that treats the two datasets independently (DS-DS Method), (2) a fully integrated approach that applies a capture-mark recapture (CMR) analysis to the PAM data (CMR-DS Method) and (3) a hybrid approach that requires only a subset of the PAM CMR data (Hybrid Method). To evaluate their performance, we use simulations based on known diving and vocalizing behavior of sperm whales (Physeter macrocephalus). As a case study, we applied the Hybrid Method to data from a shipboard survey of sperm whales and compared estimates to a VLT only analysis. Simulation results demonstrated that the CMR-DS Method and Hybrid Method reduced bias by >90% for both abundance and availability bias in comparison to the simpler DS -DS Method. Overall, the CMR-DS Method was the least biased and most precise. For the case study, our application of the Hybrid Method to the sperm whale dataset produced estimates of abundance and availability bias that were comparable to estimates from the VLT only analysis but with considerably higher precision. Integrating multiple sources of data is an important goal with clear benefits. As a step towards that goal we have developed a novel framework. Results from this study are promising although challenges still remain. Future work may focus on applying this method to other deep-diving species and comparing the proposed method to other statistical approaches that aim to combine information from multiple data sources.

Ecological inferences about marine mammals from passive acoustic data.

Monitoring on the basis of sound recordings, or passive acoustic monitoring, can complement or serve as an alternative to real-time visual or aural monitoring of marine mammals and other animals by human observers. Passive acoustic data can support the estimation of common, individual-level ecological metrics, such as presence, detection-weighted occupancy, abundance and density, population viability and structure, and behaviour. Passive acoustic data also can support estimation of some community-level metrics, such as species richness and composition. The feasibility of estimation and certainty of estimates is highly context dependent, and understanding the factors that affect the reliability of measurements is useful for those considering whether to use passive acoustic data. Here, we review basic concepts and methods of passive acoustic sampling in marine systems that often are applicable to marine mammal research and conservation. Our ultimate aim is to facilitate collaboration among ecologists, bioacousticians, and data analysts. Ecological applications of passive acoustics require one to make decisions about sampling design, which in turn requires consideration of sound propagation, sampling of signals, and data storage. One also must make decisions about signal detection and classification and evaluation of the performance of algorithms for these tasks. Investment in the research and development of systems that automate detection and classification, including machine learning, are increasing. Passive acoustic monitoring is more reliable for detection of species presence than for estimation of other species-level metrics. Use of passive acoustic monitoring to distinguish among individual animals remains difficult. However, information about detection probability, vocalisation or cue rate, and relations between vocalisations and the number and behaviour of animals increases the feasibility of estimating abundance or density. Most sensor deployments are fixed in space or are sporadic, making temporal turnover in species composition more tractable to estimate than spatial turnover. Collaborations between acousticians and ecologists are most likely to be successful and rewarding when all partners critically examine and share a fundamental understanding of the target variables, sampling process, and analytical methods.

Sperm whale acoustic abundance and dive behaviour in the western North Atlantic.

Sperm whales are an ideal species to study using passive acoustic technology because they spend the majority of their time underwater and produce echolocation clicks almost continuously while foraging. Passive acoustic line transect data collected between June and August 2016 were used to estimate a depth-corrected acoustic abundance and study the dive behaviour of foraging sperm whales in the western North Atlantic Ocean. 2D localizations (n = 699) were truncated at a slant range of 6500 m and combined with the multipath arrivals of surface reflected echoes to calculate 3D localizations (n = 274). Distance sampling using depth-corrected perpendicular distances resulted in a 10.5% change in the acoustic abundance estimate (2199 whales, CV = 14.6%) compared to uncorrected slant ranges (1969 whales, CV = 14.1%), and a detection function that was a better fit for the data. Sperm whales exhibited multiple foraging strategies, with bottom phases occurring at depths of 400-800, 800-1200, or > 1200 m, accounting for an average 39.2, 49.5, or 44.9% of the total recorded dive time, respectively. These results suggest that estimating 3D localizations using acoustic line transect data improves acoustic abundance estimation and can be used to answer population level questions about foraging ecology.

Speciation in the deep: genomics and morphology reveal a new species of beaked whale Mesoplodon eueu.

The deep sea has been described as the last major ecological frontier, as much of its biodiversity is yet to be discovered and described. Beaked whales (ziphiids) are among the most visible inhabitants of the deep sea, due to their large size and worldwide distribution, and their taxonomic diversity and much about their natural history remain poorly understood. We combine genomic and morphometric analyses to reveal a new Southern Hemisphere ziphiid species, Ramari's beaked whale, Mesoplodon eueu, whose name is linked to the Indigenous peoples of the lands from which the species holotype and paratypes were recovered. Mitogenome and ddRAD-derived phylogenies demonstrate reciprocally monophyletic divergence between M. eueu and True's beaked whale (M. mirus) from the North Atlantic, with which it was previously subsumed. Morphometric analyses of skulls also distinguish the two species. A time-calibrated mitogenome phylogeny and analysis of two nuclear genomes indicate divergence began circa 2 million years ago (Ma), with geneflow ceasing 0.35-0.55 Ma. This is an example of how deep sea biodiversity can be unravelled through increasing international collaboration and genome sequencing of archival specimens. Our consultation and involvement with Indigenous peoples offers a model for broadening the cultural scope of the scientific naming process.

Improve automatic detection of animal call sequences with temporal context.

Many animals rely on long-form communication, in the form of songs, for vital functions such as mate attraction and territorial defence. We explored the prospect of improving automatic recognition performance by using the temporal context inherent in song. The ability to accurately detect sequences of calls has implications for conservation and biological studies. We show that the performance of a convolutional neural network (CNN), designed to detect song notes (calls) in short-duration audio segments, can be improved by combining it with a recurrent network designed to process sequences of learned representations from the CNN on a longer time scale. The combined system of independently trained CNN and long short-term memory (LSTM) network models exploits the temporal patterns between song notes. We demonstrate the technique using recordings of fin whale (Balaenoptera physalus) songs, which comprise patterned sequences of characteristic notes. We evaluated several variants of the CNN + LSTM network. Relative to the baseline CNN model, the CNN + LSTM models reduced performance variance, offering a 9-17% increase in area under the precision-recall curve and a 9-18% increase in peak F1-scores. These results show that the inclusion of temporal information may offer a valuable pathway for improving the automatic recognition and transcription of wildlife recordings.

Publisher Correction: Deep neural networks for automated detection of marine mammal species.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Exploring movement patterns and changing distributions of baleen whales in the western North Atlantic using a decade of passive acoustic data.

Six baleen whale species are found in the temperate western North Atlantic Ocean, with limited information existing on the distribution and movement patterns for most. There is mounting evidence of distributional shifts in many species, including marine mammals, likely because of climate-driven changes in ocean temperature and circulation. Previous acoustic studies examined the occurrence of minke (Balaenoptera acutorostrata) and North Atlantic right whales (NARW; Eubalaena glacialis). This study assesses the acoustic presence of humpback (Megaptera novaeangliae), sei (B. borealis), fin (B. physalus), and blue whales (B. musculus) over a decade, based on daily detections of their vocalizations. Data collected from 2004 to 2014 on 281 bottom-mounted recorders, totaling 35,033 days, were processed using automated detection software and screened for each species' presence. A published study on NARW acoustics revealed significant changes in occurrence patterns between the periods of 2004-2010 and 2011-2014; therefore, these same time periods were examined here. All four species were present from the Southeast United States to Greenland; humpback whales were also present in the Caribbean. All species occurred throughout all regions in the winter, suggesting that baleen whales are widely distributed during these months. Each of the species showed significant changes in acoustic occurrence after 2010. Similar to NARWs, sei whales had higher acoustic occurrence in mid-Atlantic regions after 2010. Fin, blue, and sei whales were more frequently detected in the northern latitudes of the study area after 2010. Despite this general northward shift, all four species were detected less on the Scotian Shelf area after 2010, matching documented shifts in prey availability in this region. A decade of acoustic observations have shown important distributional changes over the range of baleen whales, mirroring known climatic shifts and identifying new habitats that will require further protection from anthropogenic threats like fixed fishing gear, shipping, and noise pollution.

Deep neural networks for automated detection of marine mammal species.

Deep neural networks have advanced the field of detection and classification and allowed for effective identification of signals in challenging data sets. Numerous time-critical conservation needs may benefit from these methods. We developed and empirically studied a variety of deep neural networks to detect the vocalizations of endangered North Atlantic right whales (Eubalaena glacialis). We compared the performance of these deep architectures to that of traditional detection algorithms for the primary vocalization produced by this species, the upcall. We show that deep-learning architectures are capable of producing false-positive rates that are orders of magnitude lower than alternative algorithms while substantially increasing the ability to detect calls. We demonstrate that a deep neural network trained with recordings from a single geographic region recorded over a span of days is capable of generalizing well to data from multiple years and across the species' range, and that the low false positives make the output of the algorithm amenable to quality control for verification. The deep neural networks we developed are relatively easy to implement with existing software, and may provide new insights applicable to the conservation of endangered species.

50 to 30-Hz triplet and singlet down sweep vocalizations produced by sei whales (Balaenoptera borealis) in the western North Atlantic Ocean.

The life history, distribution, and acoustic ecology of the sei whale (Balaenoptera borealis) in the western North Atlantic Ocean remains poorly understood. In this study an array of bottom-mounted recorders captured previously undocumented low frequency 50 to 30-Hz triplet and singlet down sweep vocalizations in close association with signature 82 to 34-Hz sei whale down sweep vocalizations. Spatiotemporal correlations of acoustically tracked sei whales confirm the original vocalizations are produced by sei whales. The 50 to 34-Hz down sweep call types were characterized with a suite of five spectral and temporal measurements. The pattern and repetition of the full acoustic suite is suggestive of song structure and warrants further investigation. The discovery of vocalizations attributed specifically to sei whales enables historic acoustic records to be re-evaluated for the presence of this species throughout its range.

A description of echolocation clicks recorded in the presence of True's beaked whale (Mesoplodon mirus).

True's beaked whales (Mesoplodon mirus) were encountered on two separate shipboard surveys on 24 July 2016 and 16 September 2017 in the western North Atlantic Ocean. Recordings were made using a hydrophone array towed 300 m behind the ship. In 2016, three different groups were sighted within 1500 m of the ship; clicks were recorded for 26 min. In 2017, a single group of five whales was tracked over the course of five hours in which the ship maintained a distance <4000 m from the group. A total of 2938 frequency-modulated (FM) clicks and 7 buzzes were recorded from both encounters. Plausible inter-click-intervals (ICIs) were calculated from 2763 clicks, and frequency and duration measurements were calculated from 2150 good quality FM clicks. The median peak frequencies were 43.1 kHz (2016, n = 718) and 43.5 kHz (2017, n = 1432). Median ICIs were 0.17 s (2016) and 0.19 s (2017). The spectra and measurements of the recorded clicks closely resemble Gervais's beaked whale clicks (Mesoplodon europaeus) and distinguishing between the two species in acoustic data sets proves difficult. The acoustic behavior of True's beaked whales was previously unknown; this study provides a description of echolocation clicks produced by this species.

Songbird dynamics under the sea: acoustic interactions between humpback whales suggest song mediates male interactions.

The function of song has been well studied in numerous taxa and plays a role in mediating both intersexual and intrasexual interactions. Humpback whales are among few mammals who sing, but the role of sexual selection on song in this species is poorly understood. While one predominant hypothesis is that song mediates male-male interactions, the mechanism by which this may occur has never been explored. We applied metrics typically used to assess songbird interactions to examine song sequences and movement patterns of humpback whale singers. We found that males altered their song presentation in the presence of other singers; focal males increased the rate at which they switched between phrase types (p = 0.005), and tended to increase the overall evenness of their song presentation (p = 0.06) after a second male began singing. Two-singer dyads overlapped their song sequences significantly more than expected by chance. Spatial analyses revealed that change in distance between singers was related to whether both males kept singing (p = 0.012), with close approaches leading to song cessation. Overall, acoustic interactions resemble known mechanisms of mediating intrasexual interactions in songbirds. Future work should focus on more precisely resolving how changes in song presentation may be used in competition between singing males.

Long-term passive acoustic recordings track the changing distribution of North Atlantic right whales (Eubalaena glacialis) from 2004 to 2014.

Given new distribution patterns of the endangered North Atlantic right whale (NARW; Eubalaena glacialis) population in recent years, an improved understanding of spatio-temporal movements are imperative for the conservation of this species. While so far visual data have provided most information on NARW movements, passive acoustic monitoring (PAM) was used in this study in order to better capture year-round NARW presence. This project used PAM data from 2004 to 2014 collected by 19 organizations throughout the western North Atlantic Ocean. Overall, data from 324 recorders (35,600 days) were processed and analyzed using a classification and detection system. Results highlight almost year-round habitat use of the western North Atlantic Ocean, with a decrease in detections in waters off Cape Hatteras, North Carolina in summer and fall. Data collected post 2010 showed an increased NARW presence in the mid-Atlantic region and a simultaneous decrease in the northern Gulf of Maine. In addition, NARWs were widely distributed across most regions throughout winter months. This study demonstrates that a large-scale analysis of PAM data provides significant value to understanding and tracking shifts in large whale movements over long time scales.

Geographic variation in Risso's dolphin echolocation click spectra.

Discrimination of bioacoustic signals to the species or population level is critical for using passive acoustic monitoring to study cetacean ecology. Risso's dolphins off southern California have distinctive peaks and notches in their echolocation clicks, but it was unknown whether Risso's dolphins from other geographic areas have similarly distinctive click spectra and whether populations are acoustically distinct. This study investigates using clicks for species and population identification by characterizing the spectral structure of Risso's dolphin echolocation clicks recorded over wide-ranging geographic regions including the U.S. waters of the North Atlantic Ocean, Gulf of Mexico, and North Pacific Ocean; and international waters of the Eastern Tropical Pacific. All recordings with Risso's dolphin clicks exhibited the spectral peak and notch pattern described off southern California, indicating the presence of peak banding patterns is useful for species discrimination. Geographic regions were a significant explanatory factor for variability in the frequencies of click spectral peaks, with relatively higher frequency peaks and notches found off Hawaii compared to California waters and off the southeast U.S. compared to the Gulf of Mexico. In the North Atlantic Ocean, a latitudinal cline in frequencies was evident. Potential causes of acoustic variation within and among acoustic encounters are evaluated.

Using multipath reflections to obtain dive depths of beaked whales from a towed hydrophone array.

Beaked whales are deep divers, emitting echolocation clicks while at depth. Little is known about the dive behavior of most species; however, passive acoustic data collected with towed hydrophone arrays can provide depth information using multipath reflections of clicks coupled with a two-dimensional localization of the individual. Data were collected during a shipboard survey in the western North Atlantic Ocean using a towed linear hydrophone array. Beaked whale tracks were classified as either Cuvier's (Ziphius cavirostris) or Gervais'/True's (Mesoplodon europaeus/Mesoplodon mirus). Weighted species average depths and weighted species standard deviations were 1158 m ± 287 m for Cuvier's (n = 24), and 870 m ± 151 m for Gervais'/True's (n = 15). Depth uncertainties ranged from 3% to 142% of the average depth. Slant ranges were corrected for depth to provide average horizontal perpendicular distance estimates. The average horizontal perpendicular distance distribution exhibited fewer detections in the first bin than the second. This is the first report of dive depths for Gervais'/True's beaked whales and use of this method to obtain depths for beaked whales using a towed linear array.

Beaked whales demonstrate a marked acoustic response to the use of shipboard echosounders.

The use of commercial echosounders for scientific and industrial purposes is steadily increasing. In addition to traditional navigational and fisheries uses, commercial sonars are used extensively for oceanographic research, benthic habitat mapping, geophysical exploration, and ecosystem studies. Little is known about the effects of these acoustic sources on marine animals, though several studies have already demonstrated behavioural responses of cetaceans to shipboard echosounders. Some species of cetaceans are known to be particularly sensitive to acoustic disturbance, including beaked whales. In 2011 and 2013, we conducted cetacean assessment surveys in the western North Atlantic in which a suite of Simrad EK60 echosounders was used to characterize the distribution of prey along survey tracklines. Echosounders were alternated daily between active and passive mode, to determine whether their use affected visual and acoustic detection rates of beaked whales. A total of 256 groups of beaked whales were sighted, and 118 definitive acoustic detections were recorded. Regression analyses using generalized linear models (GLM) found that sea state and region were primary factors in determining visual sighting rates, while echosounder state was the primary driver for acoustic detections, with significantly fewer detections (only 3%) occurring when echosounders were active. These results indicate that beaked whales both detect and change their behaviour in response to commercial echosounders. The mechanism of this response is unknown, but could indicate interruption of foraging activity or vessel avoidance, with potential implications for management and mitigation of anthropogenic impacts.

Description of sounds associated with Sowerby's beaked whales (Mesoplodon bidens) in the western North Atlantic Ocean.

Several groups of Sowerby's beaked whales (Mesoplodon bidens) were encountered on July 4, 2011, during a shipboard cetacean survey conducted off the eastern seaboard of the United States. Acoustic recordings were collected using a three-element towed hydrophone array. Many echolocation clicks were recorded during the encounter, but no tonal sounds were detected. A total of 2969 echolocation clicks were included in analyses of frequency and temporal characteristics. A Gaussian mixture model with four mixtures was fitted to the histogram of peak frequencies; four subsets of clicks were designated. The majority of clicks (n = 2048) contained a median peak frequency of 33 kHz, while the others contained a median peak frequency of 25 kHz (n = 324), 51 kHz (n = 304), or 67 kHz (n = 293). Most clicks did not contain a clear frequency-modulated upsweep, though some clicks exhibited a slight sweep from 30-36 kHz. Seven burst pulses were detected in the encounter, two of which were of high enough quality for detailed analysis. The acoustic characteristics of Sowerby's beaked whales have not previously been described; the current study will facilitate incorporation of these data into passive acoustic monitoring programs in the North Atlantic Ocean.

Calling depths of baleen whales from single sensor data: development of an autocorrelation method using multipath localization.

Multipath localization techniques have not previously been applied to baleen whale vocalizations due to difficulties in application to tonal vocalizations. Here it is shown that an autocorrelation method coupled with the direct reflected time difference of arrival localization technique can successfully resolve location information. A derivation was made to model the autocorrelation of a direct signal and its overlapping reflections to illustrate that an autocorrelation may be used to extract reflection information from longer duration signals containing a frequency sweep, such as some calls produced by baleen whales. An analysis was performed to characterize the difference in behavior of the autocorrelation when applied to call types with varying parameters (sweep rate, call duration). The method's feasibility was tested using data from playback transmissions to localize an acoustic transducer at a known depth and location. The method was then used to estimate the depth and range of a single North Atlantic right whale (Eubalaena glacialis) and humpback whale (Megaptera novaeangliae) from two separate experiments.