Development of a machine learning detector for North Atlantic humpback whale song.

The study of humpback whale song using passive acoustic monitoring devices requires bioacousticians to manually review hours of audio recordings to annotate the signals. To vastly reduce the time of manual annotation through automation, a machine learning model was developed. Convolutional neural networks have made major advances in the previous decade, leading to a wide range of applications, including the detection of frequency modulated vocalizations by cetaceans. A large dataset of over 60 000 audio segments of 4 s length is collected from the North Atlantic and used to fine-tune an existing model for humpback whale song detection in the North Pacific (see Allen, Harvey, Harrell, Jansen, Merkens, Wall, Cattiau, and Oleson (2021). Front. Mar. Sci. 8, 607321). Furthermore, different data augmentation techniques (time-shift, noise augmentation, and masking) are used to artificially increase the variability within the training set. Retraining and augmentation yield F-score values of 0.88 on context window basis and 0.89 on hourly basis with false positive rates of 0.05 on context window basis and 0.01 on hourly basis. If necessary, usage and retraining of the existing model is made convenient by a framework (AcoDet, acoustic detector) built during this project. Combining the tools provided by this framework could save researchers hours of manual annotation time and, thus, accelerate their research.

A brief overview of current approaches for underwater sound analysis and reporting.

Soundscapes have substantially changed since the industrial revolution and in response to biodiversity loss and climate change. Human activities such as shipping, resource exploration and offshore construction alter natural ecosystems through sound, which can impact marine species in complex ways. The study of underwater sound is multi-disciplinary, spanning the fields of acoustics, physics, animal physiology and behaviour to marine ecology and conservation. These different backgrounds have led to the use of various disparate terms, metrics, and summary statistics, which can hamper comparisons between studies. Different types of equipment, analytical pathways, and reporting can lead to different results for the same sound source, with implications for impact assessments. For meaningful comparisons and derivation of appropriate thresholds, mitigation, and management approaches, it is necessary to develop common standards. This paper presents a brief overview of acoustic metrics, analysis approaches and reporting standards used in the context of long-term monitoring of soundscapes.

Recommended metrics for quantifying underwater noise impacts on North Atlantic right whales.

Anthropogenic underwater noise has been identified as a potentially serious stressor for the critically endangered North Atlantic right whale (NARW). The Government of Canada is undertaking steps to better characterize the noise sources of most concern and their associated impacts, but there is currently an insufficient understanding of which noise sources are most impacting NARW in their Canadian habitat. This knowledge gap together with the myriad possible methods and metrics for quantifying underwater noise presents a confounding and challenging problem that risks delaying timely mitigation. This study presents the results from a 2020 workshop aimed at developing a series of metrics recommended specifically for better characterizing the types of noise deemed of greatest concern for NARW in Canadian waters. The recommendations provide a basis for more targeted research on noise impacts and set the stage for more effective management and protection of NARW, with potential conservation applications to similar species.

Robust North Atlantic right whale detection using deep learning models for denoising.

This paper proposes a robust system for detecting North Atlantic right whales by using deep learning methods to denoise noisy recordings. Passive acoustic recordings of right whale vocalisations are subject to noise contamination from many sources, such as shipping and offshore activities. When such data are applied to uncompensated classifiers, accuracy falls substantially. To build robustness into the detection process, two separate approaches that have proved successful for image denoising are considered. Specifically, a denoising convolutional neural network and a denoising autoencoder, each of which is applied to spectrogram representations of the noisy audio signal, are developed. Performance is improved further by matching the classifier training to include the vestigial signal that remains in clean estimates after the denoising process. Evaluations are performed first by adding white, tanker, trawler, and shot noises at signal-to-noise ratios from -10 to +5 dB to clean recordings to simulate noisy conditions. Experiments show that denoising gives substantial improvements to accuracy, particularly when using the vestigial-trained classifier. A final test applies the proposed methods to previously unseen noisy right whale recordings and finds that denoising is able to improve performance over the baseline clean-trained model in this new noise environment.

A review of the UK and British Channel Islands practical tidal stream energy resource.

This review provides a critical, multi-faceted assessment of the practical contribution tidal stream energy can make to the UK and British Channel Islands future energy mix. Evidence is presented that broadly supports the latest national-scale practical resource estimate, of 34 TWh/year, equivalent to 11% of the UK's current annual electricity demand. The size of the practical resource depends in part on the economic competitiveness of projects. In the UK, 124 MW of prospective tidal stream capacity is currently eligible to bid for subsidy support (MeyGen 1C, 80 MW; PTEC, 30 MW; and Morlais, 14 MW). It is estimated that the installation of this 124 MW would serve to drive down the levelized cost of energy (LCoE), through learning, from its current level of around 240 £ / MWh to below 150 £ / MWh , based on a mid-range technology learning rate of 17%. Doing so would make tidal stream cost competitive with technologies such as combined cycle gas turbines, biomass and anaerobic digestion. Installing this 124 MW by 2031 would put tidal stream on a trajectory to install the estimated 11.5 GW needed to generate 34 TWh/year by 2050. The cyclic, predictable nature of tidal stream power shows potential to provide additional, whole-system cost benefits. These include reductions in balancing expenditure that are not considered in conventional LCoE estimates. The practical resource is also dependent on environmental constraints. To date, no collisions between animals and turbines have been detected, and only small changes in habitat have been measured. The impacts of large arrays on stratification and predator-prey interaction are projected to be an order of magnitude less than those from climate change, highlighting opportunities for risk retirement. Ongoing field measurements will be important as arrays scale up, given the uncertainty in some environmental and ecological impact models. Based on the findings presented in this review, we recommend that an updated national-scale practical resource study is undertaken that implements high-fidelity, site-specific modelling, with improved model validation from the wide range of field measurements that are now available from the major sites. Quantifying the sensitivity of the practical resource to constraints will be important to establish opportunities for constraint retirement. Quantification of whole-system benefits is necessary to fully understand the value of tidal stream in the energy system.

Characterisation of underwater operational sound of a tidal stream turbine.

The underwater sound emitted during the operation of the Atlantis AR1500 turbine, a 1.5 MW three bladed horizontal axis tidal-stream turbine, was measured in the Pentland Firth, Scotland. Most sound was concentrated in the lower frequencies, ranging from 50 to 1000 Hz. Within 20 m of the turbine, third-octave band sound pressure levels were elevated by up to 40 dB relative to ambient conditions. In comparison, ambient noise at these frequencies fluctuated by about 5-10 dB between different tidal states. At the maximum recording distance of 2300 m from the turbine, median sound pressure levels when the turbine was operational were still over 5 dB higher than ambient noise levels alone. A higher frequency, tonal signal was observed at 20 000 Hz. This signal component appears at a constant level whenever the turbine is operational and did not change with turbine rotation rate. It is most likely produced by the turbine's generator. This study highlights the importance of empirical measurements of turbine underwater sound. It illustrates the utility and challenges of using drifting hydrophone systems to spatially map operational turbine signal levels with reduced flow noise artefacts when recording in high flow environments.

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.

Seasonal and diel acoustic presence of North Atlantic minke whales in the North Sea.

Despite frequent records from other parts of the North Atlantic, minke whales have never been acoustically recorded in the North Sea. This study investigated the detectability of pulse trains previously associated with this species in other regions, in acoustic data from ten sites along the east coast of Scotland. Since preliminary results confirmed pulse train presence, subsequently, an automated detector was applied to these data to record the seasonal and diel presence of minke whale pulse trains. Minke whales were detected from May to November, with most detections occurring in June, July and October. No acoustic detections were made in December, January or in the month of April, whilst no data were available for February and March. This pattern of acoustic presence supports available visual data and suggested an absence of minke whales from the study area during winter. Minke whale acoustic presence showed a statistically significant diel pattern, with a detection peak during night time. This study established the acoustic detectability of minke whales in the North Sea and highlights the potential of using passive acoustic monitoring to study the seasonal presence and spatial distribution of minke whales in the North Sea and wider Northeast Atlantic.

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.

Mysterious bio-duck sound attributed to the Antarctic minke whale (Balaenoptera bonaerensis).

For decades, the bio-duck sound has been recorded in the Southern Ocean, but the animal producing it has remained a mystery. Heard mainly during austral winter in the Southern Ocean, this ubiquitous sound has been recorded in Antarctic waters and contemporaneously off the Australian west coast. Here, we present conclusive evidence that the bio-duck sound is produced by Antarctic minke whales (Balaenoptera bonaerensis). We analysed data from multi-sensor acoustic recording tags that included intense bio-duck sounds as well as singular downsweeps that have previously been attributed to this species. This finding allows the interpretation of a wealth of long-term acoustic recordings for this previously acoustically concealed species, which will improve our understanding of the distribution, abundance and behaviour of Antarctic minke whales. This is critical information for a species that inhabits a difficult to access sea-ice environment that is changing rapidly in some regions and has been the subject of contentious lethal sampling efforts and ongoing international legal action.

Seasonal migrations of North Atlantic minke whales: novel insights from large-scale passive acoustic monitoring networks.

BACKGROUND: Little is known about migration patterns and seasonal distribution away from coastal summer feeding habitats of many pelagic baleen whales. Recently, large-scale passive acoustic monitoring networks have become available to explore migration patterns and identify critical habitats of these species. North Atlantic minke whales (Balaenoptera acutorostrata) perform seasonal migrations between high latitude summer feeding and low latitude winter breeding grounds. While the distribution and abundance of the species has been studied across their summer range, data on migration and winter habitat are virtually missing. Acoustic recordings, from 16 different sites from across the North Atlantic, were analyzed to examine the seasonal and geographic variation in minke whale pulse train occurrence, infer information about migration routes and timing, and to identify possible winter habitats. RESULTS: Acoustic detections show that minke whales leave their winter grounds south of 30° N from March through early April. On their southward migration in autumn, minke whales leave waters north of 40° N from mid-October through early November. In the western North Atlantic spring migrants appear to track the warmer waters of the Gulf Stream along the continental shelf, while whales travel farther offshore in autumn. Abundant detections were found off the southeastern US and the Caribbean during winter. Minke whale pulse trains showed evidence of geographic variation, with longer pulse trains recorded south of 40° N. Very few pulse trains were recorded during summer in any of the datasets. CONCLUSION: This study highlights the feasibility of using acoustic monitoring networks to explore migration patterns of pelagic marine mammals. Results confirm the presence of minke whales off the southeastern US and the Caribbean during winter months. The absence of pulse train detections during summer suggests either that minke whales switch their vocal behaviour at this time of year, are absent from available recording sites or that variation in signal structure influenced automated detection. Alternatively, if pulse trains are produced in a reproductive context by males, these data may indicate their absence from the selected recording sites. Evidence of geographic variation in pulse train duration suggests different behavioural functions or use of these calls at different latitudes.

Passive acoustic tracking of singing humpback whales (Megaptera novaeangliae) on a northwest Atlantic feeding ground.

Passive acoustic tracking provides an unobtrusive method of studying the movement of sound-producing animals in the marine environment where traditional tracking methods may be costly or infeasible. We used passive acoustic tracking to characterize the fine-scale movements of singing humpback whales (Megaptera novaeangliae) on a northwest Atlantic feeding ground. Male humpback whales produce complex songs, a phenomenon that is well documented in tropical regions during the winter breeding season, but also occurs at higher latitudes during other times of year. Acoustic recordings were made throughout 2009 using an array of autonomous recording units deployed in the Stellwagen Bank National Marine Sanctuary. Song was recorded during spring and fall, and individual singing whales were localized and tracked throughout the array using a correlation sum estimation method on the time-synchronized recordings. Tracks were constructed for forty-three song sessions, revealing a high level of variation in movement patterns in both the spring and fall seasons, ranging from slow meandering to faster directional movement. Tracks were 30 min to 8 h in duration, and singers traveled distances ranging from 0.9 to 20.1 km. Mean swimming speed was 2.06 km/h (SD 0.95). Patterns and rates of movement indicated that most singers were actively swimming. In one case, two singers were tracked simultaneously, revealing a potential acoustic interaction. Our results provide a first description of the movements of singers on a northwest Atlantic feeding ground, and demonstrate the utility of passive acoustic tracking for studying the fine-scale movements of cetaceans within the behavioral context of their calls. These methods have further applications for conservation and management purposes, particularly by enhancing our ability to estimate cetacean densities using passive acoustic monitoring.

Changes in humpback whale song occurrence in response to an acoustic source 200 km away.

The effect of underwater anthropogenic sound on marine mammals is of increasing concern. Here we show that humpback whale (Megaptera novaeangliae) song in the Stellwagen Bank National Marine Sanctuary (SBNMS) was reduced, concurrent with transmissions of an Ocean Acoustic Waveguide Remote Sensing (OAWRS) experiment approximately 200 km away. We detected the OAWRS experiment in SBNMS during an 11 day period in autumn 2006. We compared the occurrence of song for 11 days before, during and after the experiment with song over the same 33 calendar days in two later years. Using a quasi-Poisson generalized linear model (GLM), we demonstrate a significant difference in the number of minutes with detected song between periods and years. The lack of humpback whale song during the OAWRS experiment was the most substantial signal in the data. Our findings demonstrate the greatest published distance over which anthropogenic sound has been shown to affect vocalizing baleen whales, and the first time that active acoustic fisheries technology has been shown to have this effect. The suitability of Ocean Acoustic Waveguide Remote Sensing technology for in-situ, long term monitoring of marine ecosystems should be considered, bearing in mind its possible effects on non-target species, in particular protected species.