Supervised accelerometry analysis can identify prey capture by penguins at sea

Characterizing the suckling behavior by video and 3D-accelerometry in humpback whale calves on a breeding ground

Getting maternal milk through nursing is vital for all newborn mammals. Despite its importance, nursing has been poorly documented in humpback whales (Megaptera novaeangliae). Nursing is difficult to observe underwater without disturbing the whales and is usually impossible to observe from a ship. We attempted to observe nursing from the calf's perspective by placing CATS cam tags on three humpback whale calves in the Sainte Marie channel, Madagascar, Indian Ocean, during the breeding seasons. CATS cam tags are animal-borne multi-sensor tags equipped with a video camera, a hydrophone, and several auxiliary sensors (including a 3-axis accelerometer, a 3-axis magnetometer, and a depth sensor). The use of multi-sensor tags minimized potential disturbance from human presence. A total of 10.52 h of video recordings were collected with the corresponding auxiliary data. Video recordings were manually analyzed and correlated with the auxiliary data, allowing us to extract different kinematic features including the depth rate, speed, Fluke Stroke Rate (FSR), Overall Body Dynamic Acceleration (ODBA), pitch, roll, and roll rate. We found that suckling events lasted 18.8 ± 8.8 s on average (N = 34) and were performed mostly during dives. Suckling events represented 1.7% of the total observation time. During suckling, the calves were visually estimated to be at a 30-45° pitch angle relative to the midline of their mother's body and were always observed rolling either to the right or to the left. In our auxiliary dataset, we confirmed that suckling behavior was primarily characterized by a high average absolute roll and additionally we also found that it was likely characterized by a high average FSR and a low average speed. Kinematic features were used for supervised machine learning in order to subsequently detect suckling behavior automatically. Our study is a proof of method on which future investigations can build upon. It opens new opportunities for further investigation of suckling behavior in humpback whales and the baleen whale species.

Flexible use of a dynamic energy landscape buffers a marine predator against extreme climate variability

Animal migrations track predictable seasonal patterns of resource availability and suitable thermal habitat. As climate change alters this ‘energy landscape’, some migratory species may struggle to adapt. We examined how climate variability influences movements, thermal habitat selection and energy intake by juvenile Pacific bluefin tuna (Thunnus orientalis) during seasonal foraging migrations in the California Current. We tracked 242 tuna across 15 years (2002–2016) with high-resolution archival tags, estimating their daily energy intake via abdominal warming associated with digestion (the ‘heat increment of feeding’). The poleward extent of foraging migrations was flexible in response to climate variability, allowing tuna to track poleward displacements of thermal habitat where their standard metabolic rates were minimized. During a marine heatwave that saw temperature anomalies of up to +2.5°C in the California Current, spatially explicit energy intake by tuna was approximately 15% lower than average. However, by shifting their mean seasonal migration approximately 900 km poleward, tuna remained in waters within their optimal temperature range and increased their energy intake. Our findings illustrate how tradeoffs between physiology and prey availability structure migration in a highly mobile vertebrate, and suggest that flexible migration strategies can buffer animals against energetic costs associated with climate variability and change.

Basking shark sub-surface behaviour revealed by animal-towed cameras

While biologging tags have answered a wealth of ecological questions, the drivers and consequences of movement and activity often remain difficult to ascertain, particularly marine vertebrates which are difficult to observe directly. Basking sharks, the second largest shark species in the world, aggregate in the summer in key foraging sites but despite advances in biologging technologies, little is known about their breeding ecology and sub-surface behaviour. Advances in camera technologies holds potential for filling in these knowledge gaps by providing environmental context and validating behaviours recorded with conventional telemetry. Six basking sharks were tagged at their feeding site in the Sea of Hebrides, Scotland, with towed cameras combined with time-depth recorders and satellite telemetry. Cameras recorded a cumulative 123 hours of video data over an average 64-hour deployment and confirmed the position of the sharks within the water column. Feeding events only occurred within a metre depth and made up ¾ of the time spent swimming near the surface. Sharks maintained similar tail beat frequencies regardless of whether feeding, swimming near the surface or the seabed, where they spent surprisingly up to 88% of daylight hours. This study reported the first complete breaching event and the first sub-surface putative courtship display, with nose-to-tail chasing, parallel swimming as well as the first observation of grouping behaviour near the seabed. Social groups of sharks are thought to be very short term and sporadic, and may play a role in finding breeding partners, particularly in solitary sharks which may use aggregations as an opportunity to breed. In situ observation of basking sharks at their seasonal aggregation site through animal borne cameras revealed unprecedented insight into the social and environmental context of basking shark behaviour which were previously limited to surface observations.

Quantifying finer-scale behaviours using self-organising maps (SOMs) to link accelerometery signatures with behavioural patterns in free-roaming terrestrial animals

Collecting quantitative information on animal behaviours is difficult, especially from cryptic species or species that alter natural behaviours under observation. Using harness-mounted tri-axial accelerometers free-roaming domestic cats (Felis Catus) we developed a methodology that can precisely classify finer-scale behaviours. We further tested the effect of a prey-protector device designed to reduce prey capture. We aligned accelerometer traces collected at 50 Hz with video files (60 fps) and labelled 12 individual behaviours, then trained a supervised machine-learning algorithm using Kohonen super self-organising maps (SOM). The SOM was able to predict individual behaviours with a ~ 99.6% overall accuracy, which was slightly better than for random forest estimates using the same dataset (98.9%). There was a significant effect of sample size, with precision and sensitivity decreasing rapidly below 2000 1-s observations. We were also able to detect a behaviour specific reduction in the predictability when cats were fitted with the prey-protector device indicating it altered biomechanical gait. Our results can be applied in movement ecology, zoology and conservation, where habitat specific movement performance between predators or prey may be critical to managing species of conservation significance, or in veterinary and agricultural fields, where early detection of movement pathologies can improve animal welfare.

The early life of king penguins: ontogeny of dive capacity and foraging behaviour in an expert diver

The period of emancipation in seabirds, when juveniles change from a terrestrial existence to a life at sea, is associated with many challenges. Apart from finding favourable foraging sites, they have to develop effective prey search patterns and physiological capacities that enable them to capture sufficient prey to meet their energetic needs. Animals that dive to forage, such as king penguins (Aptenodytes patagonicus), need to acquire an adequate breath-hold capacity, allowing them to locate and capture prey at depth. To investigate the ontogeny of their dive capacity and foraging performance, we implanted juvenile king penguins before their first departure to sea and also adult breeders with a data-logger recording pressure and temperature. We found that juvenile king penguins possess a remarkable dive capacity when leaving their natal colony, enabling them to conduct dives in excess of 100 m within their first week at sea. Despite this, juvenile dive/foraging performance, investigated in relation to dive depth, remained below the adult level throughout their first year at sea, probably reflecting physiological limitations as a result of incomplete maturation. A significantly shallower foraging depth of juveniles, particularly during their first 5 months at sea, could also indicate differences in foraging strategy and targeted prey. The initially greater wiggle rate suggests that juveniles fed opportunistically and also targeted different prey from adults and/or that many of the wiggles of juveniles reflect unsuccessful prey-capture attempts, indicating a lower foraging proficiency. After 5 months, this difference disappeared, suggesting sufficient physical maturation and improvement of juvenile foraging skills.

Caution is warranted when using animal space-use and movement to infer behavioral states

BACKGROUND

Identifying the behavioral state for wild animals that can't be directly observed is of growing interest to the ecological community. Advances in telemetry technology and statistical methodologies allow researchers to use space-use and movement metrics to infer the underlying, latent, behavioral state of an animal without direct observations. For example, researchers studying ungulate ecology have started using these methods to quantify behaviors related to mating strategies. However, little work has been done to determine if assumed behaviors inferred from movement and space-use patterns correspond to actual behaviors of individuals.

METHODS

Using a dataset with male and female white-tailed deer location data, we evaluated the ability of these two methods to correctly identify male-female interaction events (MFIEs). We identified MFIEs using the proximity of their locations in space as indicators of when mating could have occurred. We then tested the ability of utilization distributions (UDs) and hidden Markov models (HMMs) rendered with single sex location data to identify these events.

RESULTS

For white-tailed deer, male and female space-use and movement behavior did not vary consistently when with a potential mate. There was no evidence that a probability contour threshold based on UD volume applied to an individual's UD could be used to identify MFIEs. Additionally, HMMs were unable to identify MFIEs, as single MFIEs were often split across multiple states and the primary state of each MFIE was not consistent across events.

CONCLUSIONS

Caution is warranted when interpreting behavioral insights rendered from statistical models applied to location data, particularly when there is no form of validation data. For these models to detect latent behaviors, the individual needs to exhibit a consistently different type of space-use and movement when engaged in the behavior. Unvalidated assumptions about that relationship may lead to incorrect inference about mating strategies or other behaviors.

How animals distribute themselves in space: energy landscapes of Antarctic avian predators

BACKGROUND

Energy landscapes provide an approach to the mechanistic basis of spatial ecology and decision-making in animals. This is based on the quantification of the variation in the energy costs of movements through a given environment, as well as how these costs vary in time and for different animal populations. Organisms as diverse as fish, mammals, and birds will move in areas of the energy landscape that result in minimised costs and maximised energy gain. Recently, energy landscapes have been used to link energy gain and variable energy costs of foraging to breeding success, revealing their potential use for understanding demographic changes.

METHODS

Using GPS-temperature-depth and tri-axial accelerometer loggers, stable isotope and molecular analyses of the diet, and leucocyte counts, we studied the response of gentoo (Pygoscelis papua) and chinstrap (Pygoscelis antarcticus) penguins to different energy landscapes and resources. We compared species and gentoo penguin populations with contrasting population trends.

RESULTS

Between populations, gentoo penguins from Livingston Island (Antarctica), a site with positive population trends, foraged in energy landscape sectors that implied lower foraging costs per energy gained compared with those around New Island (Falkland/Malvinas Islands; sub-Antarctic), a breeding site with fluctuating energy costs of foraging, breeding success and populations. Between species, chinstrap penguins foraged in sectors of the energy landscape with lower foraging costs per bottom time, a proxy for energy gain. They also showed lower physiological stress, as revealed by leucocyte counts, and higher breeding success than gentoo penguins. In terms of diet, we found a flexible foraging ecology in gentoo penguins but a narrow foraging niche for chinstraps.

CONCLUSIONS

The lower foraging costs incurred by the gentoo penguins from Livingston, may favour a higher breeding success that would explain the species' positive population trend in the Antarctic Peninsula. The lower foraging costs in chinstrap penguins may also explain their higher breeding success, compared to gentoos from Antarctica but not their negative population trend. Altogether, our results suggest a link between energy landscapes and breeding success mediated by the physiological condition.

Behaviour Classification on Giraffes (Giraffa camelopardalis) Using Machine Learning Algorithms on Triaxial Acceleration Data of Two Commonly Used GPS Devices and Its Possible Application for Their Management and Conservation

Averting today's loss of biodiversity and ecosystem services can be achieved through conservation efforts, especially of keystone species. Giraffes (Giraffa camelopardalis) play an important role in sustaining Africa's ecosystems, but are 'vulnerable' according to the IUCN Red List since 2016. Monitoring an animal's behavior in the wild helps to develop and assess their conservation management. One mechanism for remote tracking of wildlife behavior is to attach accelerometers to animals to record their body movement. We tested two different commercially available high-resolution accelerometers, e-obs and Africa Wildlife Tracking (AWT), attached to the top of the heads of three captive giraffes and analyzed the accuracy of automatic behavior classifications, focused on the Random Forests algorithm. For both accelerometers, behaviors of lower variety in head and neck movements could be better predicted (i.e., feeding above eye level, mean prediction accuracy e-obs/AWT: 97.6%/99.7%; drinking: 96.7%/97.0%) than those with a higher variety of body postures (such as standing: 90.7-91.0%/75.2-76.7%; rumination: 89.6-91.6%/53.5-86.5%). Nonetheless both devices come with limitations and especially the AWT needs technological adaptations before applying it on animals in the wild. Nevertheless, looking at the prediction results, both are promising accelerometers for behavioral classification of giraffes. Therefore, these devices when applied to free-ranging animals, in combination with GPS tracking, can contribute greatly to the conservation of giraffes.

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.

Hidden Markov models identify major movement modes in accelerometer and magnetometer data from four albatross species

BACKGROUND

Inertial measurement units (IMUs) with high-resolution sensors such as accelerometers are now used extensively to study fine-scale behavior in a wide range of marine and terrestrial animals. Robust and practical methods are required for the computationally-demanding analysis of the resulting large datasets, particularly for automating classification routines that construct behavioral time series and time-activity budgets. Magnetometers are used increasingly to study behavior, but it is not clear how these sensors contribute to the accuracy of behavioral classification methods. Development of effective  classification methodology is key to understanding energetic and life-history implications of foraging and other behaviors.

METHODS

We deployed accelerometers and magnetometers on four species of free-ranging albatrosses and evaluated the ability of unsupervised hidden Markov models (HMMs) to identify three major modalities in their behavior: 'flapping flight', 'soaring flight', and 'on-water'. The relative contribution of each sensor to classification accuracy was measured by comparing HMM-inferred states with expert classifications identified from stereotypic patterns observed in sensor data.

RESULTS

HMMs provided a flexible and easily interpretable means of classifying behavior from sensor data. Model accuracy was high overall (92%), but varied across behavioral states (87.6, 93.1 and 91.7% for 'flapping flight', 'soaring flight' and 'on-water', respectively). Models built on accelerometer data alone were as accurate as those that also included magnetometer data; however, the latter were useful for investigating slow and periodic behaviors such as dynamic soaring at a fine scale.

CONCLUSIONS

The use of IMUs in behavioral studies produces large data sets, necessitating the development of computationally-efficient methods to automate behavioral classification in order to synthesize and interpret underlying patterns. HMMs provide an accessible and robust framework for analyzing complex IMU datasets and comparing behavioral variation among taxa across habitats, time and space.

Assessment of Machine Learning Models to Identify Port Jackson Shark Behaviours Using Tri-Axial Accelerometers

Movement ecology has traditionally focused on the movements of animals over large time scales, but, with advancements in sensor technology, the focus can become increasingly fine scale. Accelerometers are commonly applied to quantify animal behaviours and can elucidate fine-scale (<2 s) behaviours. Machine learning methods are commonly applied to animal accelerometry data; however, they require the trial of multiple methods to find an ideal solution. We used tri-axial accelerometers (10 Hz) to quantify four behaviours in Port Jackson sharks (Heterodontus portusjacksoni): two fine-scale behaviours (<2 s)-(1) vertical swimming and (2) chewing as proxy for foraging, and two broad-scale behaviours (>2 s-mins)-(3) resting and (4) swimming. We used validated data to calculate 66 summary statistics from tri-axial accelerometry and assessed the most important features that allowed for differentiation between the behaviours. One and two second epoch testing sets were created consisting of 10 and 20 samples from each behaviour event, respectively. We developed eight machine learning models to assess their overall accuracy and behaviour-specific accuracy (one classification tree, five ensemble learners and two neural networks). The support vector machine model classified the four behaviours better when using the longer 2 s time epoch (F-measure 89%; macro-averaged F-measure: 90%). Here, we show that this support vector machine (SVM) model can reliably classify both fine- and broad-scale behaviours in Port Jackson sharks.

Machine learning for modeling animal movement

Animal movement drives important ecological processes such as migration and the spread of infectious disease. Current approaches to modeling animal tracking data focus on parametric models used to understand environmental effects on movement behavior and to fill in missing tracking data. Machine Learning and Deep learning algorithms are powerful and flexible predictive modeling tools but have rarely been applied to animal movement data. In this study we present a general framework for predicting animal movement that is a combination of two steps: first predicting movement behavioral states and second predicting the animal's velocity. We specify this framework at the individual level as well as for collective movement. We use Random Forests, Neural and Recurrent Neural Networks to compare performance predicting one step ahead as well as long range simulations. We compare results against a custom constructed Stochastic Differential Equation (SDE) model. We apply this approach to high resolution ant movement data. We found that the individual level Machine Learning and Deep Learning methods outperformed the SDE model for one step ahead prediction. The SDE model did comparatively better at simulating long range movement behaviour. Of the Machine Learning and Deep Learning models the Long Short Term Memory (LSTM) individual level model did best at long range simulations. We also applied the Random Forest and LSTM individual level models to model gull migratory movement to demonstrate the generalizability of this framework. Machine Learning and deep learning models are easier to specify compared to traditional parametric movement models which can have restrictive assumptions. However, machine learning and deep learning models are less interpretable than parametric movement models. The type of model used should be determined by the goal of the study, if the goal is prediction, our study provides evidence that machine learning and deep learning models could be useful tools.

Multi-predator assemblages, dive type, bathymetry and sex influence foraging success and efficiency in African penguins

Marine predators adapt their hunting techniques to locate and capture prey in response to their surrounding environment. However, little is known about how certain strategies influence foraging success and efficiency. Due to the miniaturisation of animal tracking technologies, a single individual can be equipped with multiple data loggers to obtain multi-scale tracking information. With the addition of animal-borne video data loggers, it is possible to provide context-specific information for movement data obtained over the video recording periods. Through a combination of video data loggers, accelerometers, GPS and depth recorders, this study investigated the influence of habitat, sex and the presence of other predators on the foraging success and efficiency of the endangered African penguin, Spheniscus demersus, from two colonies in Algoa Bay, South Africa. Due to limitations in the battery life of video data loggers, a machine learning model was developed to detect prey captures across full foraging trips. The model was validated using prey capture signals detected in concurrently recording accelerometers and animal-borne cameras and was then applied to detect prey captures throughout the full foraging trip of each individual. Using GPS and bathymetry information to inform the position of dives, individuals were observed to perform both pelagic and benthic diving behaviour. Females were generally more successful on pelagic dives than males, suggesting a trade-off between manoeuvrability and physiological diving capacity. By contrast, males were more successful in benthic dives, at least for Bird Island (BI) birds, possibly due to their larger size compared to females, allowing them to exploit habitat deeper and for longer durations. Both males at BI and both sexes at St Croix (SC) exhibited similar benthic success rates. This may be due to the comparatively shallower seafloor around SC, which could increase the likelihood of females capturing prey on benthic dives. Observation of camera data indicated individuals regularly foraged with a range of other predators including penguins and other seabirds, predatory fish (sharks and tuna) and whales. The presence of other seabirds increased individual foraging success, while predatory fish reduced it, indicating competitive exclusion by larger heterospecifics. This study highlights novel benthic foraging strategies in African penguins and suggests that individuals could buffer the effects of changes to prey availability in response to climate change. Furthermore, although group foraging was prevalent in the present study, its influence on foraging success depends largely on the type of heterospecifics present.

Optimizing the use of biologgers for movement ecology research

The paradigm-changing opportunities of biologging sensors for ecological research, especially movement ecology, are vast, but the crucial questions of how best to match the most appropriate sensors and sensor combinations to specific biological questions and how to analyse complex biologging data, are mostly ignored. Here, we fill this gap by reviewing how to optimize the use of biologging techniques to answer questions in movement ecology and synthesize this into an Integrated Biologging Framework (IBF). We highlight that multisensor approaches are a new frontier in biologging, while identifying current limitations and avenues for future development in sensor technology. We focus on the importance of efficient data exploration, and more advanced multidimensional visualization methods, combined with appropriate archiving and sharing approaches, to tackle the big data issues presented by biologging. We also discuss the challenges and opportunities in matching the peculiarities of specific sensor data to the statistical models used, highlighting at the same time the large advances which will be required in the latter to properly analyse biologging data. Taking advantage of the biologging revolution will require a large improvement in the theoretical and mathematical foundations of movement ecology, to include the rich set of high-frequency multivariate data, which greatly expand the fundamentally limited and coarse data that could be collected using location-only technology such as GPS. Equally important will be the establishment of multidisciplinary collaborations to catalyse the opportunities offered by current and future biologging technology. If this is achieved, clear potential exists for developing a vastly improved mechanistic understanding of animal movements and their roles in ecological processes and for building realistic predictive models.

Augmenting biologging with supervised machine learning to study in situ behavior of the medusa Chrysaora fuscescens

Zooplankton play critical roles in marine ecosystems, yet their fine-scale behavior remains poorly understood because of the difficulty in studying individuals in situ. Here, we combine biologging with supervised machine learning (ML) to propose a pipeline for studying in situ behavior of larger zooplankton such as jellyfish. We deployed the ITAG, a biologging package with high-resolution motion sensors designed for soft-bodied invertebrates, on eight Chrysaora fuscescens in Monterey Bay, using the tether method for retrieval. By analyzing simultaneous video footage of the tagged jellyfish, we developed ML methods to: (1) identify periods of tag data corrupted by the tether method, which may have compromised prior research findings, and (2) classify jellyfish behaviors. Our tools yield characterizations of fine-scale jellyfish activity and orientation over long durations, and we conclude that it is essential to develop behavioral classifiers on in situ rather than laboratory data.

Summary: High-resolution motion sensors paired with supervised machine learning can be used to infer fine-scale in situ behavior of zooplankton over long durations.

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.

Recent prey capture experience and dynamic habitat quality mediate short-term foraging site fidelity in a seabird

Foraging site fidelity allows animals to increase their efficiency by returning to profitable feeding areas. However, the mechanisms underpinning why animals 'stay' or 'switch' sites have rarely been investigated. Here, we explore how habitat quality and prior prey capture experience influence short-term site fidelity by the little penguin (Eudyptula minor). Using 88 consecutive foraging trips by 20 brooding penguins, we found that site fidelity was higher after foraging trips where environmental conditions were favourable, and after trips where prey capture success was high. When penguins exhibited lower site fidelity, the number of prey captures relative to the previous trip increased, suggesting that switches in foraging location were an adaptive strategy in response to low prey capture rates. Penguins foraged closer to where other penguins foraged on the same day than they did to the location of their own previous foraging site, and caught more prey when they foraged close together. This suggests that penguins aggregated flexibly when prey was abundant and accessible. Our results illustrate how foraging predators can integrate information about prior experience with contemporary information such as social cues. This gives insight into how animals combine information adaptively to exploit changing prey distribution in a dynamic environment.

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.

Processing of acceleration and dive data on-board satellite relay tags to investigate diving and foraging behaviour in free-ranging marine predators

Biologging technologies are changing the way in which the marine environment is observed and monitored. However, because device retrieval is typically required to access the high-resolution data they collect, their use is generally restricted to those animals that predictably return to land. Data abstraction and transmission techniques aim to address this, although currently these are limited in scope and do not incorporate, for example, acceleration measurements which can quantify animal behaviours and movement patterns over fine-scales.In this study, we present a new method for the collection, abstraction and transmission of accelerometer data from free-ranging marine predators via the Argos satellite system. We test run the technique on 20 juvenile southern elephant seals Mirounga leonina from the Kerguelen Islands during their first months at sea following weaning. Using retrieved archival data from nine individuals that returned to the colony, we compare and validate abstracted transmissions against outputs from established accelerometer processing procedures.Abstracted transmissions included estimates, across five segments of a dive profile, of time spent in prey catch attempt (PrCA) behaviours, swimming effort and pitch. These were then summarised and compared to archival outputs across three dive phases: descent, bottom and ascent. Correlations between the two datasets were variable but generally good (dependent on dive phase, marginal R2 values of between .45 and .6 to >.9) and consistent between individuals. Transmitted estimates of PrCA behaviours and swimming effort were positively biased to those from archival processing.Data from this study represent some of the first remotely transmitted quantifications from accelerometers. The methods presented and analysed can be used to provide novel insight towards the behaviours and movements of free-ranging marine predators, such as juvenile southern elephant seals, from whom logger retrieval is challenging. Future applications could however benefit from some adaption, particularly to reduce positive bias in transmitted PrCA behaviours and swimming effort, for which this study provides useful insight.

Integrating multiple technologies to understand the foraging behaviour of Hawaiian monk seals

The objective of this research was to investigate and describe the foraging behaviour of monk seals in the main Hawaiian Islands. Specifically, our goal was to identify a metric to classify foraging behaviour from telemetry instruments. We deployed accelerometers, seal-mounted cameras and GPS tags on six monk seals during 2012-2014 on the islands of Molokai, Kauai and Oahu. We used pitch, calculated from the accelerometer, to identify search events and thus classify foraging dives. A search event and consequent 'foraging dive' occurred when the pitch was greater than or equal to 70° at a depth less than or equal to -3 m. By integrating data from the accelerometers with video and GPS, we were able to ground-truth this classification method and identify environmental variables associated with each foraging dive. We used Bayesian logistic regression to identify the variables that influenced search events. Dive depth, body motion (mean overall dynamic body acceleration during the dive) and proximity to the sea floor were the best predictors of search events for these seals. Search events typically occurred on long, deep dives, with more time spent at the bottom (more than 50% bottom time). We can now identify where monk seals are foraging in the main Hawaiian Islands (MHI) and what covariates influence foraging behaviour in this region. This increased understanding will inform management strategies and supplement outreach and recovery efforts.

Integrating multiple technologies to understand the foraging behaviour of Hawaiian monk seals

The objective of this research was to investigate and describe the foraging behaviour of monk seals in the main Hawaiian Islands. Specifically, our goal was to identify a metric to classify foraging behaviour from telemetry instruments. We deployed accelerometers, seal-mounted cameras and GPS tags on six monk seals during 2012-2014 on the islands of Molokai, Kauai and Oahu. We used pitch, calculated from the accelerometer, to identify search events and thus classify foraging dives. A search event and consequent 'foraging dive' occurred when the pitch was greater than or equal to 70° at a depth less than or equal to -3 m. By integrating data from the accelerometers with video and GPS, we were able to ground-truth this classification method and identify environmental variables associated with each foraging dive. We used Bayesian logistic regression to identify the variables that influenced search events. Dive depth, body motion (mean overall dynamic body acceleration during the dive) and proximity to the sea floor were the best predictors of search events for these seals. Search events typically occurred on long, deep dives, with more time spent at the bottom (more than 50% bottom time). We can now identify where monk seals are foraging in the main Hawaiian Islands (MHI) and what covariates influence foraging behaviour in this region. This increased understanding will inform management strategies and supplement outreach and recovery efforts.

Seeing It All: Evaluating Supervised Machine Learning Methods for the Classification of Diverse Otariid Behaviours

Constructing activity budgets for marine animals when they are at sea and cannot be directly observed is challenging, but recent advances in bio-logging technology offer solutions to this problem. Accelerometers can potentially identify a wide range of behaviours for animals based on unique patterns of acceleration. However, when analysing data derived from accelerometers, there are many statistical techniques available which when applied to different data sets produce different classification accuracies. We investigated a selection of supervised machine learning methods for interpreting behavioural data from captive otariids (fur seals and sea lions). We conducted controlled experiments with 12 seals, where their behaviours were filmed while they were wearing 3-axis accelerometers. From video we identified 26 behaviours that could be grouped into one of four categories (foraging, resting, travelling and grooming) representing key behaviour states for wild seals. We used data from 10 seals to train four predictive classification models: stochastic gradient boosting (GBM), random forests, support vector machine using four different kernels and a baseline model: penalised logistic regression. We then took the best parameters from each model and cross-validated the results on the two seals unseen so far. We also investigated the influence of feature statistics (describing some characteristic of the seal), testing the models both with and without these. Cross-validation accuracies were lower than training accuracy, but the SVM with a polynomial kernel was still able to classify seal behaviour with high accuracy (>70%). Adding feature statistics improved accuracies across all models tested. Most categories of behaviour -resting, grooming and feeding—were all predicted with reasonable accuracy (52–81%) by the SVM while travelling was poorly categorised (31–41%). These results show that model selection is important when classifying behaviour and that by using animal characteristics we can strengthen the overall accuracy.

Putting the behavior into animal movement modeling: Improved activity budgets from use of ancillary tag information

Animal movement research relies on biotelemetry, and telemetry-based locations are increasingly augmented with ancillary information. This presents an underutilized opportunity to enhance movement process models. Given tags designed to record specific behaviors, efforts are increasing to update movement models beyond reliance solely upon horizontal movement information to improve inference of space use and activity budgets. We present two state-space models adapted to incorporate ancillary data to inform three discrete movement states: directed, resident, and an activity state. These were developed for two case studies: (1) a "haulout" model for Weddell seals, and (2) an "activity" model for Antarctic fur seals which intersperse periods of diving activity and inactivity. The methodology is easily implementable with any ancillary data that can be expressed as a proportion (or binary) indicator. A comparison of the models augmented with ancillary information and unaugmented models confirmed that many behavioral states appeared mischaracterized in the latter. Important changes in subsequent activity budgets occurred. Haulout accounted for 0.17 of the overall Weddell seal time budget, with the estimated proportion of time spent in a resident state reduced from a posterior median of 0.69 (0.65-0.73; 95% HPDI) to 0.54 (0.50-0.58 HPDI). The drop was more dramatic in the Antarctic fur seal case, from 0.57 (0.52-0.63 HPDI) to 0.22 (0.20-0.25 HPDI), with 0.35 (0.31-0.39 HPDI) of time spent in the inactive (nondiving) state. These findings reinforce previously raised contentions about the drawbacks of behavioral states inferred solely from horizontal movements. Our findings have implications for assessing habitat requirements; estimating energetics and consumption; and management efforts such as mitigating fisheries interactions. Combining multiple sources of information within integrated frameworks should improve inference of relationships between movement decisions and fitness, the interplay between resource and habitat dependencies, and their changes at the population and landscape level.

The influence of preceding dive cycles on the foraging decisions of Antarctic fur seals

The foraging strategy of many animals is thought to be determined by their past experiences. However, few empirical studies have investigated whether this is true in diving animals. We recorded three-dimensional movements and mouth-opening events from three Antarctic fur seals during their foraging trips to examine how they adapt their behaviour based on past experience--continuing to search for prey in the same area or moving to search in a different place. Each dive cycle was divided into a transit phase and a feeding phase. The linear horizontal distance travelled after feeding phases in each dive was affected by the mouth-opening rate during the previous 244 s, which typically covered two to three dive cycles. The linear distance travelled tended to be shorter when the mouth-opening rate in the previous 244 s was higher, i.e. seals tended to stay in the same areas with high prey-encounter rates. These results indicate that Antarctic fur seals follow decision-making strategies based on the past foraging experience over time periods longer than the immediately preceding dive.

Using accelerometers to remotely and automatically characterize behavior in small animals

Activity budgets in wild animals are challenging to measure via direct observation because data collection is time consuming and observer effects are potentially confounding. Although tri-axial accelerometers are increasingly employed for this purpose, their application in small-bodied animals has been limited by weight restrictions. Additionally, accelerometers engender novel complications, as a system is needed to reliably map acceleration to behaviors. In this study, we describe newly developed, tiny acceleration-logging devices (1.5-2.5 g) and use them to characterize behavior in two chipmunk species. We collected paired accelerometer readings and behavioral observations from captive individuals. We then employed techniques from machine learning to develop an automatic system for coding accelerometer readings into behavioral categories. Finally, we deployed and recovered accelerometers from free-living, wild chipmunks. This is the first time to our knowledge that accelerometers have been used to generate behavioral data for small-bodied (<100 g), free-living mammals.

High sea surface temperatures driven by a strengthening current reduce foraging success by penguins

The world’s oceans are undergoing rapid, regionally specific warming. Strengthening western boundary currents play a role in this phenomenon, with sea surface temperatures (SST) in their paths rising faster than the global average. To understand how dynamic oceanography influences food availability in these ocean warming “hotspots”, we use a novel prey capture signature derived from accelerometry to understand how the warm East Australian Current shapes foraging success by a meso-predator, the little penguin. This seabird feeds on low trophic level species that are sensitive to environmental change. We found that in 2012, prey capture success by penguins was high when SST was low relative to the long-term mean. In 2013 prey capture success was low, coincident with an unusually strong penetration of warm water. Overall there was an optimal temperature range for prey capture around 19–21 °C, with lower success at both lower and higher temperatures, mirroring published relationships between commercial sardine catch and SST. Spatially, higher SSTs corresponded to a lower probability of penguins using an area, and lower prey capture success. These links between high SST and reduced prey capture success by penguins suggest negative implications for future resource availability in a system dominated by a strengthening western boundary current.

Dive characteristics can predict foraging success in Australian fur seals (Arctocephalus pusillus doriferus) as validated by animal-borne video

Dive characteristics and dive shape are often used to infer foraging success in pinnipeds. However, these inferences have not been directly validated in the field with video, and it remains unclear if this method can be applied to benthic foraging animals. This study assessed the ability of dive characteristics from time-depth recorders (TDR) to predict attempted prey capture events (APC) that were directly observed on animal-borne video in Australian fur seals (Arctocephalus pusillus doriferus, n=11). The most parsimonious model predicting the probability of a dive with ≥1 APC on video included only descent rate as a predictor variable. The majority (94%) of the 389 total APC were successful, and the majority of the dives (68%) contained at least one successful APC. The best model predicting these successful dives included descent rate as a predictor. Comparisons of the TDR model predictions to video yielded a maximum accuracy of 77.5% in classifying dives as either APC or non-APC or 77.1% in classifying dives as successful verses unsuccessful. Foraging intensity, measured as either total APC per dive or total successful APC per dive, was best predicted by bottom duration and ascent rate. The accuracy in predicting total APC per dive varied based on the number of APC per dive with maximum accuracy occurring at 1 APC for both total (54%) and only successful APC (52%). Results from this study linking verified foraging dives to dive characteristics potentially opens the door to decades of historical TDR datasets across several otariid species.

Measuring abnormal movements in free-swimming fish with accelerometers: implications for quantifying tag and parasite load

Animal-borne data loggers allow movement, associated behaviours and energy expenditure in fish to be quantified without direct observations. As with any tagging, tags that are attached externally may adversely affect fish behaviour, swimming efficiency and survival. We report on free-swimming wild Atlantic cod (Gadus morhua) held in a large mesocosm that exhibited distinctly aberrant rotational swimming (scouring) when externally tagged with accelerometer data loggers. To quantify the phenomenon, the cod were tagged with two sizes of loggers (18 and 6 g; <2% body mass) that measured tri-axial acceleration at 50 Hz. An automated algorithm, based on body angular rotation, was designed to extract the scouring movements from the acceleration signal (98% accuracy). The algorithm also identified the frequency pattern and associated energy expenditure of scouring in relation to tag load (% body weight). The average per cent time spent scouring (5%) was independent of tag load. The vector of the dynamic body acceleration (VeDBA), used as a proxy for energy expenditure, increased with tag load (r(2)=0.51), and suggests that fish with large tags spent more energy when scouring than fish with small tags. The information allowed us to determine potential detrimental effects of an external tag on fish behaviour and how these effects may be mitigated by tag size. The algorithm can potentially identify similar rotational movements associated with spawning, courtship, feeding and parasite-load shedding in the wild. The results infer a more careful interpretation of data derived from external tags and the careful consideration of tag type, drag, buoyancy and placement, as well as animal buoyancy and species.

Navigating uncertain waters: a critical review of inferring foraging behaviour from location and dive data in pinnipeds

In the last thirty years, the emergence and progression of biologging technology has led to great advances in marine predator ecology. Large databases of location and dive observations from biologging devices have been compiled for an increasing number of diving predator species (such as pinnipeds, sea turtles, seabirds and cetaceans), enabling complex questions about animal activity budgets and habitat use to be addressed. Central to answering these questions is our ability to correctly identify and quantify the frequency of essential behaviours, such as foraging. Despite technological advances that have increased the quality and resolution of location and dive data, accurately interpreting behaviour from such data remains a challenge, and analytical methods are only beginning to unlock the full potential of existing datasets. This review evaluates both traditional and emerging methods and presents a starting platform of options for future studies of marine predator foraging ecology, particularly from location and two-dimensional (time-depth) dive data. We outline the different devices and data types available, discuss the limitations and advantages of commonly-used analytical techniques, and highlight key areas for future research. We focus our review on pinnipeds - one of the most studied taxa of marine predators - but offer insights that will be applicable to other air-breathing marine predator tracking studies. We highlight that traditionally-used methods for inferring foraging from location and dive data, such as first-passage time and dive shape analysis, have important caveats and limitations depending on the nature of the data and the research question. We suggest that more holistic statistical techniques, such as state-space models, which can synthesise multiple track, dive and environmental metrics whilst simultaneously accounting for measurement error, offer more robust alternatives. Finally, we identify a need for more research to elucidate the role of physical oceanography, device effects, study animal selection, and developmental stages in predator behaviour and data interpretation.

Interpreting behaviors from accelerometry: a method combining simplicity and objectivity

Quantifying the behavior of motile, free-ranging animals is difficult. The accelerometry technique offers a method for recording behaviors but interpretation of the data is not straightforward. To date, analysis of such data has either involved subjective, study-specific assignments of behavior to acceleration data or the use of complex analyses based on machine learning. Here, we present a method for automatically classifying acceleration data to represent discrete, coarse-scale behaviors. The method centers on examining the shape of histograms of basic metrics readily derived from acceleration data to objectively determine threshold values by which to separate behaviors. Through application of this method to data collected on two distinct species with greatly differing behavioral repertoires, kittiwakes, and humans, the accuracy of this approach is demonstrated to be very high, comparable to that reported for other automated approaches already published. The method presented offers an alternative to existing methods as it uses biologically grounded arguments to distinguish behaviors, it is objective in determining values by which to separate these behaviors, and it is simple to implement, thus making it potentially widely applicable. The R script coding the method is provided.

Identification of Prey Captures in Australian Fur Seals (Arctocephalus pusillus doriferus) Using Head-Mounted Accelerometers: Field Validation with Animal-Borne Video Cameras

This study investigated prey captures in free-ranging adult female Australian fur seals (Arctocephalus pusillus doriferus) using head-mounted 3-axis accelerometers and animal-borne video cameras. Acceleration data was used to identify individual attempted prey captures (APC), and video data were used to independently verify APC and prey types. Results demonstrated that head-mounted accelerometers could detect individual APC but were unable to distinguish among prey types (fish, cephalopod, stingray) or between successful captures and unsuccessful capture attempts. Mean detection rate (true positive rate) on individual animals in the testing subset ranged from 67-100%, and mean detection on the testing subset averaged across 4 animals ranged from 82-97%. Mean False positive (FP) rate ranged from 15-67% individually in the testing subset, and 26-59% averaged across 4 animals. Surge and sway had significantly greater detection rates, but also conversely greater FP rates compared to heave. Video data also indicated that some head movements recorded by the accelerometers were unrelated to APC and that a peak in acceleration variance did not always equate to an individual prey item. The results of the present study indicate that head-mounted accelerometers provide a complementary tool for investigating foraging behaviour in pinnipeds, but that detection and FP correction factors need to be applied for reliable field application.