Tracking Of Whales Research Articles

Localization and tracking of beluga whales in aerial video using deep learning

Localization and tracking of beluga whales in aerial video using deep learning

Reinforcement learning-based framework for whale rendezvous via autonomous sensing robots.

Rendezvous with sperm whales for biological observations is made challenging by their prolonged dive patterns. Here, we propose an algorithmic framework that codevelops multiagent reinforcement learning-based routing (autonomy module) and synthetic aperture radar-based very high frequency (VHF) signal-based bearing estimation (sensing module) for maximizing rendezvous opportunities of autonomous robots with sperm whales. The sensing module is compatible with low-energy VHF tags commonly used for tracking wildlife. The autonomy module leverages in situ noisy bearing measurements of whale vocalizations, VHF tags, and whale dive behaviors to enable time-critical rendezvous of a robot team with multiple whales in simulation. We conducted experiments at sea in the native habitat of sperm whales using an "engineered whale"-a speedboat equipped with a VHF-emitting tag, emulating five distinct whale tracks, with different whale motions. The sensing module shows a median bearing error of 10.55° to the tag. Using bearing measurements to the engineered whale from an acoustic sensor and our sensing module, our autonomy module gives an aggregate rendezvous success rate of 81.31% for a 500-meter rendezvous distance using three robots in postprocessing. A second class of fielded experiments that used acoustic-only bearing measurements to three untagged sperm whales showed an aggregate rendezvous success rate of 68.68% for a 1000-meter rendezvous distance using two robots in postprocessing. We further validated these algorithms with several ablation studies using a sperm whale visual encounter dataset collected by marine biologists.

Characterizing humpback whale behavior along the North-Norwegian coast

BackgroundStudying movement patterns of individual animals over time can give insight into how they interact with the environment and optimize foraging strategies. Humpback whales (Megaptera novaeangliae) undertake long seasonal migrations between feeding areas in polar regions and breeding grounds in tropical areas. During the last decade, several individuals have had up to a 3-month stop-over period around specific fjord-areas in Northern Norway to feed on Norwegian spring-spawning (NSS-) herring (Culpea harengus L.). Their behavioral patterns during this period are not well understood, including why some whales seemingly leave the fjords and then later return within the same season.MethodsTo investigate whale behavior during this seasonal stopover, we classified humpback whale tracks into five distinct movement modes; ranging, encamped, nomadic, roundtrip and semi-roundtrip. A behavioral change point analysis (BCPA) was used to select homogeneous segments based on persistence velocity. Then, net squared displacement (NSD) over time was modeled to differentiate movement modes. This study also manually identified longer roundtrips away from the fjords that lasted several days and examined movement modes within these.ResultsInside the fjord systems, encamped mode was most prevalent in December–January, suggesting the whales were mainly foraging on overwintering NSS-herring in this area. During the same winter seasons, half of the whales left the fjords and then returned. We hypothesize that these trips serve as ‘searching trips’ during which the whales seek better feeding opportunities outside the fjords. If better foraging conditions are not found, they return to the fjords to continue their feeding. The overall most common mode was ranging (54%), particularly seen during the start of their southwards migration and in areas outside the fjord systems, indicating that the whales mainly moved over larger distances in the offshore habitat.ConclusionsThis study serves as a baseline for future studies investigating both the searching trip theory and humpback whale behavior in general, and confirms that this method could be useful to analyze local scale movement patterns of satellite tagged whales.

Annual phenology and migration routes to breeding grounds in western-central North Pacific sei whales

The sei whale (Balaenoptera borealis) is an important species among baleen whales in the North Pacific and plays a significant role in the ecosystem. Despite the importance of this species, information regarding its migration patterns and breeding locations remains limited. To enhance the understanding of the phenology of North Pacific sei whales, we deployed satellite-monitored tags on these whales in the western and central North Pacific from 2017 to 2023. We fitted 55 sei whale tracks to a state-space model to describe the whales’ seasonal movements at feeding grounds and their migratory behavior. The whales typically leave their feeding grounds between November and December, with migration pathways extending from off Japan to the west of the Hawaiian Islands. These southward transits converge in the waters of the Marshall Islands and north of Micronesia between 20° N and 7° N, which appear to be breeding grounds. After a brief stay at these breeding grounds, the whales migrate northward from January to February, reaching their feeding grounds around 30°N by March. To the best of our knowledge, this is the first study to present the phenology of feeding and breeding seasons and the migration pattern of North Pacific sei whales.

A decade of change and stability for fin whale song in the North Atlantic

Fin whale song is a ubiquitous low-frequency pulsing that has been recorded in every ocean basin. In recent years, the complexity of fin whale song has been realized. In the North Atlantic, this song is made up of two low-frequency note types (A and B notes) and one higher-frequency note type (upsweeps). The A and B notes are produced with both singlet and doublet inter-note intervals. These song patterns are interwoven throughout an individual’s song and have been observed to shift over time, both gradually and suddenly. Like many baleen whales, the population size of fin whales is poorly defined. Passive acoustic monitoring has been suggested as a way to estimate abundance if a cue rate or calling rate can be determined. In this study, we used recordings from 119 fin whale tracks from 2013–2023 generated from passive acoustic recorders in the North Atlantic to define fin whale song patterns in this ocean basin, identify changes in these song patterns over time, and calculate cue rates. The fin whales recorded sang in 4 distinct patterns. Inter-note intervals were 25 s for A-A singlets, 13 s for B-B singlets, 12/20 s for B-B doublets, and 10/11 s for A-B doublets. B-B doublet inter-note intervals significantly increased from May 2019–April 2023 and A-B inter-note intervals showed an intra-annual increasing trend that reset each summer. Fin whale song in the North Atlantic also often includes higher frequency upsweep notes immediately prior to B notes. While A and B note peak frequency was steady over these years, upsweep peak frequency significantly decreased, continuing a similar trend that has been observed over 30 years. Despite these changes, the dominant song pattern remained consistent and cue rate for actively singing whales was relatively stable, which is in contrast to patterns observed in the North Pacific. Many of these nuances in fin whale singing behavior are newly described herein, and these findings are vital for monitoring baseline behavior, behavioral trends over time, and ultimately estimating abundance of a poorly understood species.

Automated detection of fin whale calls recorded with distributed acoustic sensing

As the global biodiversity crisis intensifies, wildlife monitoring has become a necessity. From an oceanographic perspective, the tracking of baleen whales allows studies of their habitat use and can support call density estimates, which are indicative of ocean health. Distributed Acoustic Sensing (DAS) is a promising observational technique that measures strain rates along a spare optical fiber with a spatial resolution of tens of meters to distances of ∼100 km. Over 4-days in November 2021, a public domain DAS dataset was collected on the submarine cables of the Ocean Observatories Initiative Regional Cabled Array off the coast of central Oregon. The experiment recorded the acoustic signals from tens of thousands of fin whale calls as well as blue whale calls, ship signals and T-phases. We will describe preliminary efforts to develop automated methods to detect and localize fin whale calls using small representative examples of the DAS data. Our longer-term goal is to extend this approach to the full data set to understand the sensitivity of DAS to fin whale calls as function of cable geometry and seafloor characteristics and to study the distribution of calling fin whales in coastal waters off central Oregon. [Work supported by ONR.]

An acoustic tracking model based on deep learning using two hydrophones and its reverberation transfer hypothesis, applied to whale tracking

Acoustic tracking of whales’ underwater cruises is essential for protecting marine ecosystems. For cetacean conservationists, fewer hydrophones will provide more convenience in capturing high-mobility whale positions. Currently, it has been possible to use two hydrophones individually to accomplish direction finding or ranging. However, traditional methods only aim at estimating one of the spatial parameters and are susceptible to the detrimental effects of reverberation superimposition. To achieve complete whale tracking under reverberant interference, in this study, an intelligent acoustic tracking model (CIAT) is proposed, which allows both horizontal direction discrimination and distance/depth perception by mining unpredictable features of position information directly from the received signals of two hydrophones. Specifically, the horizontal direction is discriminated by an enhanced cross-spectral analysis to make full use of the exact frequency of received signals and eliminate the interference of non-source signals, and the distance/depth direction combines convolutional neural network (CNN) with transfer learning to address the adverse effects caused by unavoidable acoustic reflections and reverberation superposition. Experiments with real recordings show that 0.13 km/MAE is achieved within 8 km. Our work not only provides satisfactory prediction performance, but also effectively avoids the reverberation effect of long-distance signal propagation, opening up a new avenue for underwater target tracking.

Satellite tracking of minke whales (Balaenoptera acutorostrata) off the coast of northern Norway

Two minke whales were tagged with satellite-linked radio transmitters off the coast of northern Norway in order to obtain data on daily locations, movements and swimming speed. One whale was tagged in September 1994, south of Lofoten at the entrance to the Vestfjorden, and one whale was tagged in August 1999 just north of Vesterålen. The whale tagged in 1994 was successfully tracked for 31 days (located 1.5 times/day on average). The whale tagged in 1999 was successfully tracked for 19 days (located 3.0 times/day on average), although the first locations were not obtained until 18 days after the instrumentation. The whale tagged in 1994 travelled between two apparent feeding areas on the west coast of northern Norway: one in the mouth of Vestfjorden and the other along the continental slope north of Vesterålen. The whale tagged in 1999 moved to an area inside Vestfjorden and remained there until early September, after which it began a southward movement out of Vestfjorden offshore to the edge of the continental shelf. Both whales were presumably feeding on herring (Clupea harengus), which is particularly abundant in these waters at this time of year. The two minke whales travelled 78 and 79 km/day when distances between all positions were used, and 66 and 53 km/day when the daily average positions (all qualities) were used. Both calculations illustrate that minke whales can move considerable distances on a daily basis.

Minke whales change their swimming behavior with respect to their calling behavior, nearby conspecifics, and the environment in the central North Pacific

Behavioral responses to sonar have been observed in a number of baleen whales, including minke whales (Balaenoptera acutorostrata). Previous studies used acoustic minke whale boing detections to localize and track individual whales on the U.S. Pacific Missile Range Facility (PMRF) in Kaua ‘i, Hawai‘i before, during, and after Navy training activities. These analyses showed significant changes in central North Pacific minke whale distribution and swimming behavior during Navy sonar events. For the purposes of contextualizing changes in animal movement relative to Navy sonar, we expanded on this research to examine the natural variation in minke whale movement when Navy sonar was not present. This study included 2,245 acoustically derived minke whale tracks spanning the years 2012–2017 over all months that minke whales were detected (October–May). Minke whale movement was examined relative to calling season, day of the year, hour of day, wind speed, calling state (nominal or rapid), and distance to the nearest calling conspecific. Hidden Markov models were used to identify two kinematic states (slower, less directional movement and faster, more directional movement). The findings indicate that minke whales were more likely to travel in a faster and more directional state when they were calling rapidly, when other vocalizing minke whales were nearby, during certain times of the day and calling seasons, and in windier conditions, but these changes in movement were less intense than the changes observed during exposure to Navy sonar, when swim speeds were the fastest. These results start to put behavioral responses to Navy sonar into an environmental context to understand the severity of responses relative to natural changes in behavior.

First evidence for fin whale migration into the Pacific from Antarctic feeding grounds at Elephant Island.

This study presents the first long-distance tracks of fin whales (Balaenoptera physalus) equipped with satellite transmitters off the Antarctic Peninsula. Southern Hemisphere fin whales were severely depleted by twentieth century industrial whaling, yet recently, they have returned to historical feeding grounds off the northern Antarctic Peninsula, forming large aggregations in austral summers. To date, our knowledge only extended to summer behaviour, while information regarding migration routes and the location of breeding and wintering grounds are lacking. During the austral autumn of 2021, we deployed nsatellite transmitters on four fin whales at Elephant Island. Two transmitters stopped working while the animals were still at the feeding grounds, while two continued to transmit during the transition from feeding activity to migration. Both migrating animals left the feeding ground on 15 April 2021, travelling northward into the Pacific and up along the Chilean coast. The most northerly position received before all tags stopped transmitting on 1 May 2021 was at 48°S. These tracks provide initial evidence of seasonal migratory routes and a first indication toward possible locations of winter destinations. This information, even if preliminary, is critical for investigations of population connectivity, population structure and the identification of breeding grounds of Southern Hemisphere fin whales.

Acoustic monitoring and analyses of air gun, pile driving, vessel, and ambient sounds during the 2015 seismic surveys on the Sakhalin shelf

During the summer of 2015, four 4D seismic surveys were conducted on the northeastern Sakhalin shelf near the feeding grounds of the Korean-Okhotsk (western) gray whale (Eschrichtius robustus) population. In addition to the seismic surveys, onshore pile driving activities and vessel operations occurred. Forty autonomous underwater acoustic recorders provided data in the 2 Hz to15 kHz frequency band. Recordings were analyzed to evaluate the characteristics of impulses propagating from the seismic sources. Acoustic metrics analyzed comprised peak sound pressure level (PK), mean square sound pressure level (SPL), sound exposure level (SEL), T100%, T90% (the time intervals that contain the full and 90% of the energy of the impulse), and kurtosis. The impulses analyzed differed significantly due to the variability and complexity of propagation in the shallow water of the northeast Sakhalin shelf. At larger ranges, a seismic precursor propagated in the seabed ahead of the acoustic impulse, and the impulses often interfered with each other, complicating analyses. Additional processing of recordings allowed evaluation and documentation of relevant metrics for pile driving, vessel sounds, and ambient background levels. The computed metrics were used to calibrate acoustic models, generating time resolved estimates of the acoustic levels from seismic surveys, pile driving, and vessel operations on a gray whale distribution grid and along observed gray whale tracks. This paper describes the development of the metrics and the calibrated acoustic models, both of which will be used in work quantifying gray whale behavioral and distribution responses to underwater sounds and to determine whether these observed responses have the potential to impact important parameters at the population level (e.g., reproductive success).

Spatiotemporal Overlap of Baleen Whales and Krill Fisheries in the Western Antarctic Peninsula Region

In Antarctica, abundant consumers rely on Antarctic krill for food, but krill are also the subject of a commercial fishery. The fishery overlaps in time and space with the foraging areas of these consumers, thus potential competition between krill fisheries and krill consumers is a major management concern. The fishery is managed by the Commission for the Conservation of Antarctic Marine Living Resources with an ecosystem approach, according to which fishing should not interfere with either the population growth of krill, or krill-dependent consumers. Krill catches have become increasingly spatially concentrated in a small number of hotspots, raising concerns about how local depletion of krill impacts consumers. Such concentrated fishing demonstrates that there is a mismatch between the spatial and temporal scale at which krill fisheries are currently managed, and that at which fisheries operate and consumers forage. Information on the seasonal dynamics of predator abundance and their foraging behaviour is fundamental to future precautionary management of the krill fishery. We analysed the spatiotemporal distribution of two major krill consumers – humpback and minke whales – and that of krill fishing, off the Western Antarctic Peninsula. We used whale tracking data (58 humpback whale tracks and 19 minke whale tracks) to develop spatial random forest models predicting the monthly distribution of whale foraging areas from January-July. Using these predictions, we calculated spatiotemporally-explicit geographic overlap between whales and fisheries, the latter represented by krill fishing effort and catch data. Over the krill fishing season, fishing effort and catch hotspots shifted to the southwest, into the Bransfield Strait where effort and catch was highest. Predicted humpback whale foraging areas increased in the Bransfield Strait over the same period, while predicted minke whale foraging areas showed an opposite trend. For both we predicted a whale-fishing interaction hotspot in the Bransfield Strait, strongest in April and May. Our results illustrate the fine spatial scale of likely interactions between baleen whales and the krill fishery, and their concentration over the season, underlining the need for fishery management more closely aligned to the spatiotemporal scale of likely predator-fishery interactions.

Monitoring fin and blue whales in the lower St. Lawrence Seaway with onshore seismometers

AbstractThe Lower St. Lawrence Seaway (LSLS), in eastern Canada, is an important habitat for several species of endangered baleen whale. As we seek to reduce the hazards that these endangered species face from human activity, there is increasing demand for detailed knowledge of their habitat use. Only a sparse network of hydrophones exists in the LSLS to remotely observe whales. However, there is also a network of onshore seismometers, designed to monitor earthquakes, that have sufficiently high sample rates to record fin and blue whale calls. We present a simple method for detecting band‐limited, regularly repeating calls, such as the 20 Hz calls of fin and blue whales, and apply the method to build a catalog of fin and blue whale detections at 14 onshore seismometers across the LSLS, over approximately a 4‐year period. The resulting catalog contains over 600 000 fin whale calls and almost 60 000 blue whale calls. Individual calls are rarely detected at more than one seismometer. Fin whale calls recorded onshore often consist of multiple seismic phases arriving as a ∼2 sequence. Onshore seismometers provide a valuable, previously unused source of data for monitoring baleen whales. However, in the LSLS, the current seismometer network cannot provide high‐precision whale tracking alone, so a denser deployment of onshore and/or offshore seismometers is required.

Singing Fin Whale Swimming Behavior in the Central North Pacific

Male fin whales sing using 20 Hz pulses produced in regular patterns of inter-note intervals, but little is known about fin whale swimming behavior while they are singing. Even less is known about fin whales in Hawaiian waters because they have rarely been sighted during surveys and passive acoustic monitoring has been limited to sparse hydrophone systems that do not have localization capabilities. We hypothesized that fin whale kinematics may be related to their singing behavior, or external variables such as time and sea state. To investigate this hypothesis, we analyzed 115 tracks containing 50,034 unique notes generated from passive acoustic recordings on an array of 14 hydrophones from 2011 to 2017 at the U.S. Navy Pacific Missile Range Facility off Kauai, Hawaii. Fin whales swam at an average speed of 1.1 m/s over relatively direct paths. We incorporated the whales' speed and turning angle into hidden Markov models to identify different behavioral states based on the whales' movements. We found that fin whale kinematic behavioral state was related to the vocalization rate (also known as cue rate) and time of day. When cue rate was higher, fin whales were more likely to swim slower and turn more than when cue rate was lower. During the night, fin whales were also more likely to swim slower and turn more than during the day. In addition, we examined whether the presence of singing fin whales was related to time and sea state using generalized additive models. Fin whale track presence was affected by day of the year and song season, and possibly also wind speed and wave height. Although the track kinematics from the fin whale tracks presented here are limited to a subset of whales that are acoustically active, they provide some of the only detailed movements of fin whales in the region and can be compared against fin whale swim speeds in other regions. Understanding how fin whale swimming behavior varies based on their vocalization patterns, time, and environmental factors will help us to contextualize potential changes in whale behavior during Navy training and testing on the range.

Acoustic tracking of fin whales: Habitat use and movement patterns within a Canadian Pacific fjord system.

Fin whale 20 Hz calls were detected, localized, and tracked using a 10 km aperture network of three acoustic receivers deployed for 11 months in a Pacific Canadian fjord system. The area has been historically important for fin whales and is located along a route that tankers will begin using in 2024. A total of 6712 calls were localized, and trajectories were fitted for 55 acoustic tracks. Fin whale tracks occurred throughout the monitoring site. Call activity peaked in September and was low during winter months. Swimming characteristics varied significantly between day- and nighttime: at night, whales swam faster (7.1 vs 4.0 km/h median, +75.2%), which resulted in longer (+34.7%), less predictable (-70.6%) tracks as compared to daylight hours. Call frequencies varied between 16 and 32 Hz. Beside stereotypical song frequencies, fin whales also used irregular frequency components, which contributed the majority of calls in the summer but did not occur in the winter. The results suggest that the area is primarily used as a summer feeding ground, where fin whales follow a diel behavioral cycle. The observed activity patterns will aid in the assessment of strike risk and harassment mitigation and provide a baseline to document behavioral change.

Asset Tracking Whales—First Deployment of a Custom-Made GPS/GSM Suction Cup Tag on Migrating Humpback Whales

The study of marine mammals is greatly enhanced through fine scale data on habitat use. Here we used a commonly available asset tracker Global Positioning System/Global Systems for Mobile Communication (GPS/GSM) integrated into a CATS suction cup tag to test its feasibility in providing real time location position on migrating humpback whales in coastal waters of eastern Australia. During two deployments—one on a suspected male and another on a female humpback whale—the tags provided location points with relatively high accuracy for both individuals albeit different swim behavior and surface intervals. In combination with an integrated archival data logger, the tag also provided detailed information on fine scale habitat use such as dive profiles. However, surface intervals were too short to allow for an upload of location data during deployment. Further improvements of the tag design will allow remote access to location data after deployment. Preliminary results suggested location acquisition was better when the tag was positioned well above the midline of the whale body. The technology promises less expensive, more reliable and more accurate short-term tracking of humpback whales compared to satellite relay tags, and it has the potential to be deployed on other marine mammals in coastal waters.

Predicting seasonal movements and distribution of the sperm whale using machine learning algorithms.

Implementation of effective conservation planning relies on a robust understanding of the spatiotemporal distribution of the target species. In the marine realm, this is even more challenging for species rarely seen at the sea surface due to their extreme diving behavior like the sperm whales. Our study aims at (a) investigating the seasonal movements, (b) predicting the potential distribution, and (c) assessing the diel vertical behavior of this species in the Mascarene Archipelago in the south‐west Indian Ocean. Using 21 satellite tracks of sperm whales and eight environmental predictors, 14 supervised machine learning algorithms were tested and compared to predict the whales' potential distribution during the wet and dry season, separately. Fourteen of the whales remained in close proximity to Mauritius, while a migratory pattern was evidenced with a synchronized departure for eight females that headed towards Rodrigues Island. The best performing algorithm was the random forest, showing a strong affinity of the whales for sea surface height during the wet season and for bottom temperature during the dry season. A more dispersed distribution was predicted during the wet season, whereas a more restricted distribution to Mauritius and Reunion waters was found during the dry season, probably related to the breeding period. A diel pattern was observed in the diving behavior, likely following the vertical migration of squids. The results of our study fill a knowledge gap regarding seasonal movements and habitat affinities of this vulnerable species, for which a regional IUCN assessment is still missing in the Indian Ocean. Our findings also confirm the great potential of machine learning algorithms in conservation planning and provide highly reproductible tools to support dynamic ocean management.

Study on an Automatic Sperm Whale Tracking System by a Drone

Study on an Automatic Sperm Whale Tracking System by a Drone

Detection probability and density estimation of fin whales by a seaglider

The effectiveness of gliders is evaluated for passive acoustic density estimation of fin whales. An estimate of the probability of detection as a function of range (or, equivalently, effective survey area) is required to estimate density from acoustic data collected by a single-hydrophone glider platform. A cabled hydrophone array was used to estimate fin whale localizations and tracks concurrently with a glider survey. Fin whale tracks were used as detection trials and a detection function for snapshots containing fin whale 20 Hz pulses recorded by the glider was modeled using a generalized additive model. Detection probability was strongly dependent on 40 Hz noise levels recorded on the glider. At the median noise level of 97 dB re 1μPa2/Hz, detection probability was near one at zero horizontal distance, and maximum detection range was near 40 km. The estimated effective survey area at this noise level was 870 km2. Using estimates of vocal rates and group size from tagged and tracked fin whales, respectively, the density was estimated as 2.4 fin whales per 1000 km2 (coefficient of variation 0.55). The framework presented here could be applied to other baleen whale species to advance the use of gliders for density estimation of cetaceans.

An automated passive acoustic monitoring system for real time sperm whale (Physeter macrocephalus) threat prevention in the Mediterranean Sea

The sperm whale (Physeter macrocephalus) is one of the largest and most widely distributed marine mammals. The Mediterranean population, classified as Endangered in the IUCN Red List, is affected by a number of anthropogenic threats, including noise increase. The individuals of this species tend to spend long periods “rafting” and socializing at the surface inbetween deep dives and this make them highly vulnerable to ship strikes. In the Pelagos Sanctuary, 8% of photo-identified sperm whales has evidence of wounds from propellers. In the framework of the Life+ Nature WHALESAFE, an interference avoidance system capable of detecting and tracking sperm whales in the range of about 7 km was developed in order to identify the threats and potentially prevent collisions and other risks; this tool issues warning messages in real time to ships. In this work, the technical details of the WHALESAFE detector and of the two reconstruction algorithms developed for the sperm whale tracking will be described. The analysis of a cetacean signal has been used to test the performances of WHALESAFE, revealing that the average error on the position is 50 m at a distance of 1 km. The accuracy in the reconstruction of the cetacean position fits the requirements for an efficient animal tracking and species preservation.