Development of an automated method of detecting stereotyped feeding events in multisensor data from tagged rorqual whales

Using quantile regression and relative entropy to assess the period of anomalous behavior of marine mammals following tagging

Tagging of animals induces a variable stress response which following release will obscure natural behavior. It is of scientific relevance to establish methods that assess recovery from such behavioral perturbation and generalize well to a broad range of animals, while maintaining model transparency. We propose two methods that allow for subdivision of animals based on covariates, and illustrate their use onN = 20 narwhals (Monodon monoceros) andN = 4 bowhead whales (Balaena mysticetus), captured and instrumented with Acousonde™ behavioral tags, but with a framework that easily generalizes to other marine animals and sampling units. The narwhals were divided into two groups based on handling time, short (t < 58 min) and long (t ≥ 58 min), to measure the effect on recovery. Proxies for energy expenditure (VeDBA) and rapid movement (jerk) were derived from accelerometer data. Diving profiles were characterized using two metrics (target depth and dive duration) derived from depth data. For accelerometer data, recovery was estimated using quantile regression (QR) on the log-transformed response, whereas depth data were addressed using relative entropy (RE) between hourly distributions of dive duration (partitioned into three target depth ranges) and the long-term average distribution. Quantile regression was used to address location-based behavior to accommodate distributional shifts anticipated in aquatic locomotion. For all narwhals, we found fast recovery in the tail of the distribution (<3 h) compared with a variable recovery at the median (∼1-10 h) and with a significant difference between groups separated by handling time. Estimates of bowhead whale recovery times showed fast median recovery (<3 h) and slow recovery at the tail (>6 h), but were affected by substantial uncertainty. For the diving profiles, as characterized by the component pair (target depth, dive duration), the recovery was slower (narwhals-long:t < 16 h; narwhals-short:t < 10 h; bowhead whales: <9 h) and with a difference between narwhals with short vs long handling times. Using simple statistical concepts, we have presented two transparent and general methods for analyzing high-resolution time series data from marine animals, addressing energy expenditure, activity, and diving behavior, and which allows for comparison between groups of animals based on well-defined covariates.

Tag-based estimates of bottlenose dolphin swimming behavior and energetics

Current estimates of marine mammal hydrodynamic forces tend to be made using camera-based kinematic data for a limited number of fluke strokes during a prescribed swimming task. In contrast, biologging tag data yield kinematic measurements from thousands of strokes, enabling new insights into swimming behavior and mechanics. However, there have been limited tag-based estimates of mechanical work and power. In this work, we investigated bottlenose dolphin (Tursiops truncatus) swimming behavior using tag-measured kinematics and a hydrodynamic model to estimate propulsive power, work and cost of transport. Movement data were collected from six animals during prescribed straight-line swimming trials to investigate swimming mechanics over a range of sustained speeds (1.9-6.1 m s-1). Propulsive power ranged from 66 W to 3.8 kW over 282 total trials. During the lap trials, the dolphins swam at depths that mitigated wave drag, reducing overall drag throughout these mid- to high-speed tasks. Data were also collected from four individuals during undirected daytime (08:30-18:00 h) swimming to examine how self-selected movement strategies are used to modulate energetic efficiency and effort. Overall, self-selected swimming speeds (individual means ranging from 1.0 to 1.96 m s-1) tended to minimize cost of transport, and were on the lower range of animal-preferred speeds reported in literature. The results indicate that these dolphins moderate propulsive effort and efficiency through a combination of speed and depth regulation. This work provides new insights into dolphin swimming behavior in both prescribed tasks and self-selected swimming, and presents a path forward for continuous estimates of mechanical work and power from wild animals.

The use of machine learning to detect foraging behaviour in whale sharks: a new tool in conservation

In this study we present the first attempt at modelling the feeding behaviour of whale sharks using a machine learning analytical method. A total of eight sharks were monitored with tri-axial accelerometers and their foraging behaviours were visually observed. Our results highlight that the random forest model is a valid and robust approach to predict the feeding behaviour of the whale shark. In conclusion this novel approach exposes the practicality of this method to serve as a conservation tool and the capability it offers in monitoring potential disturbances of the species.

Insight into the kinematics of blue whale surface foraging through drone observations and prey data

To understand how predators optimize foraging strategies, extensive knowledge of predator behavior and prey distribution is needed. Blue whales employ an energetically demanding lunge feeding method that requires the whales to selectively feed where energetic gain exceeds energetic loss, while also balancing oxygen consumption, breath holding capacity, and surface recuperation time. Hence, blue whale foraging behavior is primarily driven by krill patch density and depth, but many studies have not fully considered surface feeding as a significant foraging strategy in energetic models. We collected predator and prey data on a blue whale (Balaenoptera musculus brevicauda) foraging ground in New Zealand in February 2017 to assess the distributional and behavioral response of blue whales to the distribution and density of krill prey aggregations. Krill density across the study region was greater toward the surface (upper 20 m), and blue whales were encountered where prey was relatively shallow and more dense. This relationship was particularly evident where foraging and surface lunge feeding were observed. Furthermore, New Zealand blue whales also had relatively short dive times (2.83 ± 0.27 SE min) as compared to other blue whale populations, which became even shorter at foraging sightings and where surface lunge feeding was observed. Using an unmanned aerial system (UAS; drone) we also captured unique video of a New Zealand blue whale's surface feeding behavior on well-illuminated krill patches. Video analysis illustrates the whale's potential use of vision to target prey, make foraging decisions, and orient body mechanics relative to prey patch characteristics. Kinematic analysis of a surface lunge feeding event revealed biomechanical coordination through speed, acceleration, head inclination, roll, and distance from krill patch to maximize prey engulfment. We compared these lunge kinematics to data previously reported from tagged blue whale lunges at depth to demonstrate strong similarities, and provide rare measurements of gape size, and krill response distance and time. These findings elucidate the predator-prey relationship between blue whales and krill, and provide support for the hypothesis that surface feeding by New Zealand blue whales is an important component to their foraging ecology used to optimize their energetic efficiency. Understanding how blue whales make foraging decisions presents logistical challenges, which may cause incomplete sampling and biased ecological knowledge if portions of their foraging behavior are undocumented. We conclude that surface foraging could be an important strategy for blue whales, and integration of UAS with tag-based studies may expand our understanding of their foraging ecology by examining surface feeding events in conjunction with behaviors at depth.

Why whales are big but not bigger: Physiological drivers and ecological limits in the age of ocean giants

The largest animals are marine filter feeders, but the underlying mechanism of their large size remains unexplained. We measured feeding performance and prey quality to demonstrate how whale gigantism is driven by the interplay of prey abundance and harvesting mechanisms that increase prey capture rates and energy intake. The foraging efficiency of toothed whales that feed on single prey is constrained by the abundance of large prey, whereas filter-feeding baleen whales seasonally exploit vast swarms of small prey at high efficiencies. Given temporally and spatially aggregated prey, filter feeding provides an evolutionary pathway to extremes in body size that are not available to lineages that must feed on one prey at a time. Maximum size in filter feeders is likely constrained by prey availability across space and time.

Hidden Markov models reveal temporal patterns and sex differences in killer whale behavior

Behavioral data can be important for effective management of endangered marine predators, but can be challenging to obtain. We utilized suction cup-attached biologging tags equipped with stereo hydrophones, triaxial accelerometers, triaxial magnetometers, pressure and temperature sensors, to characterize the subsurface behavior of an endangered population of killer whales (Orcinus orca). Tags recorded depth, acoustic and movement behavior on fish-eating killer whales in the Salish Sea between 2010-2014. We tested the hypotheses that (a) distinct behavioral states can be characterized by integrating movement and acoustic variables, (b) subsurface foraging occurs in bouts, with distinct periods of searching and capture temporally separated from travel, and (c) the probabilities of transitioning between behavioral states differ by sex. Using Hidden Markov modeling of two acoustic and four movement variables, we identified five temporally distinct behavioral states. Persistence in the same state on a subsequent dive had the greatest likelihood, with the exception of deep prey pursuit, indicating that behavior was clustered in time. Additionally, females spent more time at the surface than males, and engaged in less foraging behavior. These results reveal significant complexity and sex differences in subsurface foraging behavior, and underscore the importance of incorporating behavior into the design of conservation strategies.

Kinematic signatures of prey capture from archival tags reveal sex differences in killer whale foraging activity

Studies of odontocete foraging ecology have been limited by the challenges of observing prey capture events and outcomes underwater. We sought to determine whether subsurface movement behavior recorded from archival tags could accurately identify foraging events by fish-eating killer whales. We used multisensor bio-logging tags attached by suction cups to Southern Resident killer whales (Orcinus orca) to: (1) identify a stereotyped movement signature that co-occurred with visually confirmed prey capture dives; (2) construct a prey capture dive detector and validate it against acoustically confirmed prey capture dives; and (3) demonstrate the utility of the detector by testing hypotheses about foraging ecology. Predation events were significantly predicted by peaks in the rate of change of acceleration (‘jerk peak’), roll angle and heading variance. Detection of prey capture dives by movement signatures enabled substantially more dives to be included in subsequent analyses compared with previous surface or acoustic detection methods. Males made significantly more prey capture dives than females and more dives to the depth of their preferred prey, Chinook salmon. Additionally, only half of the tag deployments on females (5 out of 10) included a prey capture dive, whereas all tag deployments on males exhibited at least one prey capture dive (12 out of 12). This dual approach of kinematic detection of prey capture coupled with hypothesis testing can be applied across odontocetes and other marine predators to investigate the impacts of social, environmental and anthropogenic factors on foraging ecology.

Development and application of a machine learning algorithm for classification of elasmobranch behaviour from accelerometry data

Discerning behaviours of free-ranging animals allows for quantification of their activity budget, providing important insight into ecology. Over recent years, accelerometers have been used to unveil the cryptic lives of animals. The increased ability of accelerometers to store large quantities of high resolution data has prompted a need for automated behavioural classification. We assessed the performance of several machine learning (ML) classifiers to discern five behaviours performed by accelerometer-equipped juvenile lemon sharks (Negaprion brevirostris) at Bimini, Bahamas (25°44'N, 79°16'W). The sharks were observed to exhibit chafing, burst swimming, headshaking, resting and swimming in a semi-captive environment and these observations were used to ground-truth data for ML training and testing. ML methods included logistic regression, an artificial neural network, two random forest models, a gradient boosting model and a voting ensemble (VE) model, which combined the predictions of all other (base) models to improve classifier performance. The macro-averaged F-measure, an indicator of classifier performance, showed that the VE model improved overall classification (F-measure 0.88) above the strongest base learner model, gradient boosting (0.86). To test whether the VE model provided biologically meaningful results when applied to accelerometer data obtained from wild sharks, we investigated headshaking behaviour, as a proxy for prey capture, in relation to the variables: time of day, tidal phase and season. All variables were significant in predicting prey capture, with predations most likely to occur during early evening and less frequently during the dry season and high tides. These findings support previous hypotheses from sporadic visual observations.