Sonarlogger: Enabling long-term underwater sonar observations

Coastal seas are under increasing pressure from extreme weather events and sea level rise, resulting in impacts such as changing hydrodynamic conditions, coastal erosion, and marine heat waves. To monitor changes in coastal marine habitats, such as reefs and macrophytes meadows, which add to the resilience of our coasts, consistent, medium- to long-term seafloor observations are needed. This project aims to deliver repeated, high-frequency sonar surveys on a stationary seabed mooring of a specific target area over a period of up to several months. A new stand-alone subsea system, the Sonarlogger, based on a battery pack, low-power logger and a high-resolution scanning sonar, was developed. It allows for long-term deployments with a customisable battery pack, WI-FI download and configurable sleep state. The system has been tested for over 130 days in dynamic coastal environments off the Belgian coast. Combined with auxiliary sensors, such as for measuring currents, waves and turbidity, this system enables comprehensive studies of morphologic changes and changing benthic ecosystems. Moreover, this system has the capacity to provide measurements of coastal environments during storms, where conventional systems may fall short, providing insights into event-based changes of the seafloor.

FUZ-SMO: A fuzzy slime mould optimizer for mitigating false alarm rates in the classification of underwater datasets using deep convolutional neural networks

Sonar sound datasets are of significant importance in the domains of underwater surveillance and marine research as they enable experts to discern intricate patterns within the depths of the water. Nevertheless, the task of classifying sonar sound datasets continues to pose significant challenges. In this study, we present a novel approach aimed at enhancing the precision and efficacy of sonar sound dataset classification. The integration of deep long-short-term memory (DLSTM) and convolutional neural networks (CNNs) models is employed in order to capitalize on their respective advantages while also utilizing distinctive feature engineering techniques to achieve the most favorable outcomes. Although DLSTM networks have demonstrated effectiveness in tasks involving sequence classification, attaining their optimal performance necessitates careful adjustment of hyperparameters. While traditional methods such as grid and random search are effective, they frequently encounter challenges related to computational inefficiencies. This study aims to investigate the unexplored capabilities of the fuzzy slime mould optimizer (FUZ-SMO) in the context of LSTM hyperparameter tuning, with the objective of addressing the existing research gap in this area. Drawing inspiration from the adaptive behavior exhibited by slime moulds, the FUZ-SMO proposes a novel approach to optimization. The amalgamated model, which combines CNN, LSTM, fuzzy, and SMO, exhibits a notable improvement in classification accuracy, outperforming conventional LSTM architectures by a margin of 2.142%. This model not only demonstrates accelerated convergence milestones but also displays significant resilience against overfitting tendencies.

Efforts to advance underwater noise management in Canada: Introduction to the Marine Pollution Bulletin Special Issue

This introduction to a special issue on approaches to managing underwater noise in Canada provides a brief overview of recent efforts to better understand and reduce anthropogenic underwater noise. Recent programs have aimed to increase understanding of anthropogenic noise in the habitats of highly endangered whales and have supported management actions such as vessel slow downs. Technical workshops have advanced the development of quiet ship design and associated technologies. Collaborative research examined noise levels in the St. Lawrence Estuary and the Arctic Ocean. Efforts to better manage noise have gone beyond shipping: enhanced mitigation measures have been put in place for naval exercises near habitats used by southern resident killer whales, while other work has focused on the identification of appropriate metrics for measuring noise. To coordinate and advance these and other efforts, the Government of Canada is developing a national Ocean Noise Strategy.

Heart rate and startle responses in diving, captive harbour porpoises (Phocoena phocoena) exposed to transient noise and sonar

Anthropogenic noise can alter marine mammal behaviour and physiology, but little is known about cetacean cardiovascular responses to exposures, despite evidence that acoustic stressors, such as naval sonars, may lead to decompression sickness. Here, we measured heart rate and movements of two trained harbour porpoises during controlled exposure to 6-9 kHz sonar-like sweeps and 40 kHz peak-frequency noise pulses, designed to evoke acoustic startle responses. The porpoises initially responded to the sonar sweep with intensified bradycardia despite unaltered behaviour/movement, but habituated rapidly to the stimuli. In contrast, 40 kHz noise pulses consistently evoked rapid muscle flinches (indicative of startles), but no behavioural or heart rate changes. We conclude that the autonomous startle response appears decoupled from, or overridden by, cardiac regulation in diving porpoises, whereas certain novel stimuli may motivate oxygen-conserving cardiovascular measures. Such responses to sound exposure may contribute to gas mismanagement for deeper-diving cetaceans.

BLAINDER-A Blender AI Add-On for Generation of Semantically Labeled Depth-Sensing Data

Common Machine-Learning (ML) approaches for scene classification require a large amount of training data. However, for classification of depth sensor data, in contrast to image data, relatively few databases are publicly available and manual generation of semantically labeled 3D point clouds is an even more time-consuming task. To simplify the training data generation process for a wide range of domains, we have developed the BLAINDER add-on package for the open-source 3D modeling software Blender, which enables a largely automated generation of semantically annotated point-cloud data in virtual 3D environments. In this paper, we focus on classical depth-sensing techniques Light Detection and Ranging (LiDAR) and Sound Navigation and Ranging (Sonar). Within the BLAINDER add-on, different depth sensors can be loaded from presets, customized sensors can be implemented and different environmental conditions (e.g., influence of rain, dust) can be simulated. The semantically labeled data can be exported to various 2D and 3D formats and are thus optimized for different ML applications and visualizations. In addition, semantically labeled images can be exported using the rendering functionalities of Blender.

Postnatal development of diving physiology: implications of anthropogenic disturbance for immature marine mammals

Marine mammals endure extended breath-holds while performing active behaviors, which has fascinated scientists for over a century. It is now known that these animals have large onboard oxygen stores and utilize oxygen-conserving mechanisms to prolong aerobically supported dives to great depths, while typically avoiding (or tolerating) hypoxia, hypercarbia, acidosis and decompression sickness (DCS). Over the last few decades, research has revealed that diving physiology is underdeveloped at birth. Here, I review the postnatal development of the body's oxygen stores, cardiorespiratory system and other attributes of diving physiology for pinnipeds and cetaceans to assess how physiological immaturity makes young marine mammals vulnerable to disturbance. Generally, the duration required for body oxygen stores to mature varies across species in accordance with the maternal dependency period, which can be over 2 years long in some species. However, some Arctic and deep-diving species achieve mature oxygen stores comparatively early in life (prior to weaning). Accelerated development in these species supports survival during prolonged hypoxic periods when calves accompany their mothers under sea ice and to the bathypelagic zone, respectively. Studies on oxygen utilization patterns and heart rates while diving are limited, but the data indicate that immature marine mammals have a limited capacity to regulate heart rate (and hence oxygen utilization) during breath-hold. Underdeveloped diving physiology, in combination with small body size, limits diving and swimming performance. This makes immature marine mammals particularly vulnerable to mortality during periods of food limitation, habitat alterations associated with global climate change, fishery interactions and other anthropogenic disturbances, such as exposure to sonar.

Rio Doce Acoustic Surveys of Fish Biomass and Aquatic Habitat

Following the failure of the Fundão mine tailings dam in Brazil, approximately 32 million cubic meters of Fe ore tailings were released into the downstream riverine system. The postevent monitoring surveys implemented the use of noninvasive acoustic methods to improve the understanding of the fish biomass distribution patterns and aquatic habitat condition of the impacted reaches of the Rio Gualaxo do Norte, Rio do Carmo, and Rio Doce. The primary focus of the program was to collect hydroacoustic measurements of fish biomass, which was accompanied by sonar imaging of aquatic habitats using high-resolution side scan and downward imaging sonar at each site visited. The biannual surveys began in April 2017 and were fundamentally a multiple control-impact design because before data (prior to the dam failure event) were not available. Up to 22 sites were visited for each survey, including reservoir and river sites. Results indicate similar levels of instream habitat between control and impact river and reservoir sites. Average volumetric biomass was not significantly different between impact sites and their corresponding controls in the August 2018 survey (latest to date). Preliminary temporal analysis of the biomass data set collected indicates that either stable or increasing levels of biomass are detected at tailings impacted sites within the Rio Doce. Integr Environ Assess Manag 2020;16:615-621. © 2020 Hydrobiology QLD Pty Ltd. Integrated Environmental Assessment and Management published by Wiley Periodicals LLC on behalf of Society of Environmental Toxicology & Chemistry (SETAC).

Navigation and perception of spatial layout in virtual echo-acoustic space

Successful navigation involves finding the way, planning routes, and avoiding collisions. Whilst previous research has shown that people can navigate using non-visual cues, it is not clear to what degree learned non-visual navigational abilities generalise to 'new' environments. Furthermore, the ability to successfully avoid collisions has not been investigated separately from the ability to perceive spatial layout or to orient oneself in space. Here, we address these important questions using a virtual echolocation paradigm in sighted people. Fourteen sighted blindfolded participants completed 20 virtual navigation training sessions over the course of 10 weeks. In separate sessions, before and after training, we also tested their ability to perceive the spatial layout of virtual echo-acoustic space. Furthermore, three blind echolocation experts completed the tasks without training, thus validating our virtual echo-acoustic paradigm. We found that over the course of 10 weeks sighted people became better at navigating, i.e. they reduced collisions and time needed to complete the route, and increased success rates. This also generalised to 'new' (i.e. untrained) virtual spaces. In addition, after training, their ability to judge spatial layout was better than before training. The data suggest that participants acquired a 'true' sensory driven navigational ability using echo-acoustics. In addition, we show that people not only developed navigational skills related to avoidance of collisions and finding safe passage, but also processes related to spatial perception and orienting. In sum, our results provide strong support for the idea that navigation is a skill which people can achieve via various modalities, here: echolocation.

Behavioral responses of individual blue whales (Balaenoptera musculus) to mid-frequency military sonar

This study measured the degree of behavioral responses in blue whales (Balaenoptera musculus) to controlled noise exposure off the southern California coast. High-resolution movement and passive acoustic data were obtained from non-invasive archival tags (n=42) whereas surface positions were obtained with visual focal follows. Controlled exposure experiments (CEEs) were used to obtain direct behavioral measurements before, during and after simulated and operational military mid-frequency active sonar (MFAS), pseudorandom noise (PRN) and controls (no noise exposure). For a subset of deep-feeding animals (n=21), active acoustic measurements of prey were obtained and used as contextual covariates in response analyses. To investigate potential behavioral changes within individuals as a function of controlled noise exposure conditions, two parallel analyses of time-series data for selected behavioral parameters (e.g. diving, horizontal movement and feeding) were conducted. This included expert scoring of responses according to a specified behavioral severity rating paradigm and quantitative change-point analyses using Mahalanobis distance statistics. Both methods identified clear changes in some conditions. More than 50% of blue whales in deep-feeding states responded during CEEs, whereas no changes in behavior were identified in shallow-feeding blue whales. Overall, responses were generally brief, of low to moderate severity, and highly dependent on exposure context such as behavioral state, source-to-whale horizontal range and prey availability. Response probability did not follow a simple exposure-response model based on received exposure level. These results, in combination with additional analytical methods to investigate different aspects of potential responses within and among individuals, provide a comprehensive evaluation of how free-ranging blue whales responded to mid-frequency military sonar.

Modeling bat prey capture in echolocating bats: The feasibility of reactive pursuit

Echolocating bats are the only mammals engaging in airborne pursuit. In this paper, we implement a reactive model of sonar based prey pursuit in bats. Our simulations include a realistic prey localization mechanism as well as a model of the bat's motor behavior. In contrast to previous work, our model incorporates bats' ability to execute rapid saccadic scanning motions keeping the prey within its field of view. Decoupling the flight direction from the gaze direction allows our model to capture erratically moving prey using reactive control. We conclude that the rapid shifts in gaze direction allow bats to deal with the narrow field of view provided by their sonar system.

Echo-acoustic flow affects flight in bats

Flying animals need to react fast to rapid changes in their environment. Visually guided animals use optic flow, generated by their movement through structured environments. Nocturnal bats cannot make use of optic flow, but rely mostly on echolocation. Here, we show that bats exploit echo-acoustic flow to negotiate flight through narrow passages. Specifically, bats' flight between lateral structures is significantly affected by the echo-acoustic salience of those structures, independent of their physical distance. This is true even though echolocation, unlike vision, provides explicit distance cues. Moreover, the bats reduced the echolocation sound levels in stronger flow, probably to compensate for the increased summary target strength of the lateral reflectors. However, bats did not reduce flight velocity under stronger echo-acoustic flow. Our results demonstrate that sensory flow is a ubiquitous principle for flight guidance, independent of the fundamentally different peripheral representation of flow across the senses of vision and echolocation.

Controlled Sonar Exposure Experiments on Cetaceans in Norwegian Waters: Overview of the 3S-Project

In mitigating the risk of sonar operations, the behavioral response of cetaceans is one of the major knowledge gaps that needs to be addressed. The 3S-Project has conducted a number of controlled exposure experiments with a realistic sonar source in Norwegian waters from 2006 to 2013. In total, the following six target species have been studied: killer, long-finned pilot, sperm, humpback, minke, and northern bottlenose whales. A total of 38 controlled sonar exposures have been conducted on these species. Responses from controlled and repeated exposure runs have been recorded using acoustic and visual observations as well as with electronic tags on the target animal. So far, the first dose-response curves as well as an overview of the scored severity of responses have been revealed. In this paper, an overview is presented of the approach for the study, including the results so far as well as the current status of the ongoing analysis.

Identifying Variations in Baseline Behavior of Killer Whales (Orcinus orca) to Contextualize Their Responses to Anthropogenic Noise

Determining the baseline behavior of a whale requires understanding natural variations occurring due to environmental context, such as changes in prey behavior. Killer whales feeding on herring consistently encircle herring schools; however, depth of feeding differs from near the surface in winter to deeper than 10 m in spring and summer. These variations in feeding depth are probably due to the depth of the prey and the balance between the costs and benefits of bringing schools of herring to the surface. Such variation in baseline behavior may incur different energetic costs and consequently change the motivation of whales to avoid a feeding area. Here, we discuss these variations in feeding behavior in the context of exposure to noise and interpret observed responses to simulated navy sonar signals.

Effects of Sound on the Behavior of Wild, Unrestrained Fish Schools

To assess and manage the impact of man-made sounds on fish, we need information on how behavior is affected. Here, wild unrestrained pelagic fish schools were observed under quiet conditions using sonar. Fish were exposed to synthetic piling sounds at different levels using custom-built sound projectors, and behavioral changes were examined. In some cases, the depth of schools changed after noise playback; full dispersal of schools was also evident. The methods we developed for examining the behavior of unrestrained fish to sound exposure have proved successful and may allow further testing of the relationship between responsiveness and sound level.

Assessing the Effectiveness of Ramp-Up During Sonar Operations Using Exposure Models

Ramp-up procedures are used to mitigate the impact of sound on marine mammals. Sound exposure models combined with observations of marine mammals responding to sound can be used to assess the effectiveness of ramp-up procedures. We found that ramp-up procedures before full-level sonar operations can reduce the risk of hearing threshold shifts with marine mammals, but their effectiveness depends strongly on the responsiveness of the animals. In this paper, we investigated the effect of sonar parameters (source level, pulse-repetition time, ship speed) on sound exposure by using a simple analytical model and highlight the mechanisms that limit the effectiveness of ramp-up procedures.

The Challenges of Analyzing Behavioral Response Study Data: An Overview of the MOCHA (Multi-study OCean Acoustics Human Effects Analysis) Project

This paper describes the MOCHA project which aims to develop novel approaches for the analysis of data collected during Behavioral Response Studies (BRSs). BRSs are experiments aimed at directly quantifying the effects of controlled dosages of natural or anthropogenic stimuli (typically sound) on marine mammal behavior. These experiments typically result in low sample size, relative to variability, and so we are looking at a number of studies in combination to maximize the gain from each one. We describe a suite of analytical tools applied to BRS data on beaked whales, including a simulation study aimed at informing future experimental design.

Behavioral Response of Reef Fish and Green Sea Turtles to Midfrequency Sonar

There is growing concern over the potential effects of high-intensity sonar on wild fish populations and commercial fisheries. Acoustic telemetry was employed to measure the movements of free-ranging reef fish and sea turtles in Port Canaveral, FL, in response to routine submarine sonar testing. Twenty-five sheepshead (Archosargus probatocephalus), 28 gray snapper (Lutjanus griseus), and 29 green sea turtles (Chelonia mydas) were tagged, with movements monitored for a period of up to 4 months using an array of passive acoustic receivers. Baseline residency was examined for fish and sea turtles before, during, and after the test event. No mortality of tagged fish or sea turtles was evident from the sonar test event. There was a significant increase in the daily residency index for both sheepshead and gray snapper at the testing wharf subsequent to the event. No broad-scale movement from the study site was observed during or immediately after the test.

Active Acoustic Monitoring of Aquatic Life

Active acoustic monitoring (AAM) can be used to study the behavioral response of marine life and to mitigate harm during high-danger anthropogenic activities. This has been done in fish studies for many decades, and there are now case studies in which AAM has been used for marine mammal monitoring as well. This includes monitoring where the ranges, AAM frequency of operation, and species are such that the AAM operation is completely outside the hearing range of the animals. However, it also includes AAM operations within the hearing range of marine life, although this does not necessarily that imply AAM is not a suitable tool. It is just not always possible to have a sufficient detection and tracking range and operate at a frequency outside the marine life hearing range. Likely, the best and most important application of AAM is when the anthropogenic activity to be conducted is temporary and presents a clear danger to aquatic life.

How effectively do horizontal and vertical response strategies of long-finned pilot whales reduce sound exposure from naval sonar?

The behaviour of a marine mammal near a noise source can modulate the sound exposure it receives. We demonstrate that two long-finned pilot whales both surfaced in synchrony with consecutive arrivals of multiple sonar pulses. We then assess the effect of surfacing and other behavioural response strategies on the received cumulative sound exposure levels and maximum sound pressure levels (SPLs) by modelling realistic spatiotemporal interactions of a pilot whale with an approaching source. Under the propagation conditions of our model, some response strategies observed in the wild were effective in reducing received levels (e.g. movement perpendicular to the source's line of approach), but others were not (e.g. switching from deep to shallow diving; synchronous surfacing after maximum SPLs). Our study exemplifies how simulations of source-whale interactions guided by detailed observational data can improve our understanding about motivations behind behaviour responses observed in the wild (e.g., reducing sound exposure, prey movement).

High thresholds for avoidance of sonar by free-ranging long-finned pilot whales (Globicephala melas)

The potential effects of exposing marine mammals to military sonar is a current concern. Dose-response relationships are useful for predicting potential environmental impacts of specific operations. To reveal behavioral response thresholds of exposure to sonar, we conducted 18 exposure/control approaches to 6 long-finned pilot whales. Source level and proximity of sonar transmitting one of two frequency bands (1-2 kHz and 6-7 kHz) were increased during exposure sessions. The 2-dimensional movement tracks were analyzed using a changepoint method to identify the avoidance response thresholds which were used to estimate dose-response relationships. No support for an effect of sonar frequency or previous exposures on the probability of response was found. Estimated response thresholds at which 50% of population show avoidance (SPLmax=170 dB re 1 μPa, SELcum=173 dB re 1 μPa(2) s) were higher than previously found for other cetaceans. The US Navy currently uses a generic dose-response relationship to predict the responses of cetaceans to naval active sonar, which has been found to underestimate behavioural impacts on killer whales and beaked whales. The navy curve appears to match more closely our results with long-finned pilot whales, though it might underestimate the probability of avoidance for pilot-whales at long distances from sonar sources.