Using accelerometers to remotely and automatically characterize behavior in small animals

Identifying animal behaviours from accelerometers: Improving predictive accuracy of machine learning by refining the variables selected, data frequency, and sample duration

Observing animals in the wild often poses extreme challenges, but animal-borne accelerometers are increasingly revealing unobservable behaviours. Automated machine learning streamlines behaviour identification from the substantial datasets generated during multi-animal, long-term studies; however, the accuracy of such models depends on the qualities of the training data. We examined how data processing influenced the predictive accuracy of random forest (RF) models, leveraging the easily observed domestic cat (Felis catus) as a model organism for terrestrial mammalian behaviours. Nine indoor domestic cats were equipped with collar-mounted tri-axial accelerometers, and behaviours were recorded alongside video footage. From this calibrated data, eight datasets were derived with (i) additional descriptive variables, (ii) altered frequencies of acceleration data (40 Hz vs. a mean over 1 s) and (iii) standardised durations of different behaviours. These training datasets were used to generate RF models that were validated against calibrated cat behaviours before identifying the behaviours of five free-ranging tag-equipped cats. These predictions were compared to those identified manually to validate the accuracy of the RF models for free-ranging animal behaviours. RF models accurately predicted the behaviours of indoor domestic cats (F-measure up to 0.96) with discernible improvements observed with post-data-collection processing. Additional variables, standardised durations of behaviours and higher recording frequencies improved model accuracy. However, prediction accuracy varied with different behaviours, where high-frequency models excelled in identifying fast-paced behaviours (e.g. locomotion), whereas lower-frequency models (1 Hz) more accurately identified slower, aperiodic behaviours such as grooming and feeding, particularly when examining free-ranging cat behaviours. While RF modelling offered a robust means of behaviour identification from accelerometer data, field validations were important to validate model accuracy for free-ranging individuals. Future studies may benefit from employing similar data processing methods that enhance RF behaviour identification accuracy, with extensive advantages for investigations into ecology, welfare and management of wild animals.

Resting disparity in quoll semelparity: examining the sex-linked behaviours of wild roaming northern quolls (Dasyurus hallucatus) during breeding season

Semelparity is a breeding strategy whereby an individual invests large amounts of resources into a single breeding season, leading to the death of the individual. Male northern quolls (Dasyurus hallucatus) are the largest known mammal to experience a post-breeding die-off; however, the cause of their death is unknown, dissimilar from causes in other semelparous dasyurids. To identify potential differences between male northern quolls that breed once, and females that can breed for up to four seasons, the behaviours, activity budgets, speeds and distances travelled were examined. Northern quolls were captured on Groote Eylandt off the coast of the Northern Territory, Australia, and were fitted with accelerometers. A machine learning algorithm (Self-organizing Map) was trained on more than 76 h of recorded footage of quoll behaviours and used to predict behaviours in 42 days of data from wild roaming quolls (7M : 6F). Male northern quolls were more active (male 1.27 g, s.d. = 0.41; female 1.18 g, s.d. = 0.36), spent more time walking (13.09% male: 8.93% female) and engaged in less lying/resting behaviour than female northern quolls (7.67% male: 23.65% female). Reduced resting behaviour among males could explain the post-breeding death as the deterioration in appearance reflects that reported for sleep-deprived rodents.

An implantable neurophysiology platform: Broadening research capabilities in free-living and non-traditional animals

Animal-borne sensors that can record and transmit data (“biologgers”) are becoming smaller and more capable at a rapid pace. Biologgers have provided enormous insight into the covert lives of many free-ranging animals by characterizing behavioral motifs, estimating energy expenditure, and tracking movement over vast distances, thereby serving both scientific and conservational endpoints. However, given that biologgers are usually attached externally, access to the brain and neurophysiological data has been largely unexplored outside of the laboratory, limiting our understanding of how the brain adapts to, interacts with, or addresses challenges of the natural world. For example, there are only a handful of studies in free-living animals examining the role of sleep, resulting in a wake-centric view of behavior despite the fact that sleep often encompasses a large portion of an animal’s day and plays a vital role in maintaining homeostasis. The growing need to understand sleep from a mechanistic viewpoint and probe its function led us to design an implantable neurophysiology platform that can record brain activity and inertial data, while utilizing a wireless link to enable a suite of forward-looking capabilities. Here, we describe our design approach and demonstrate our device’s capability in a standard laboratory rat as well as a captive fox squirrel. We also discuss the methodological and ethical implications of deploying this new class of device “into the wild” to fill outstanding knowledge gaps.

Video Validation of Tri-Axial Accelerometer for Monitoring Zoo-Housed Tamandua tetradactyla Activity Patterns in Response to Changes in Husbandry Conditions

Accelerometers are a technology that is increasingly used in the evaluation of animal behaviour. A tri-axial accelerometer attached to a vest was used on Tamandua tetradactyla individuals (n = 10) at Biodiversity Park. First, the influence of using a vest on the animals’ behaviour was evaluated (ABA-type: A1 and A2, without a vest; B, with a vest; each stage lasted 24 h), and no changes were detected. Second, their behaviour was monitored using videos and the accelerometer simultaneously (experimental room, 20 min). The observed behaviours were correlated with the accelerometer data, and summary measures (X, Y and Z axes) were obtained. Additionally, the overall dynamic body acceleration was calculated, determining a threshold to discriminate activity/inactivity events (variance = 0.0055). Then, based on a 24 h complementary test (video sampling every 5 min), the sensitivity (85.91%) and precision (100%) of the accelerometer were assessed. Animals were exposed to an ABA-type experimental design: A1 and A2: complex enclosure; B: decreased complexity (each stage lasted 24 h). An increase in total activity (%) was revealed using the accelerometer (26.15 ± 1.50, 29.29 ± 2.25, and 35.36 ± 3.15, respectively). Similar activity levels were detected using video analysis. The results demonstrate that the use of the accelerometer is reliable to determine the activity. Considering that the zoo-housed lesser anteaters exhibit a cathemeral activity pattern, this study contributes to easily monitoring their activities and responses to different management procedures supporting welfare programs, as well as ex situ conservation.

A methodology for preprocessing structured big data in the behavioral sciences

The characteristics of big data, including high volume, increased variety, and velocity, pose special challenges for data analysis. As these characteristics generally preclude manual data inspection and processing, researchers must often use computational methodologies to deal with this type of data; techniques that may be unfamiliar to nonspecialists, including behavioral scientists. However, previous data analytics methodologies within the field of computer science, developed to handle the generic tasks of data collection, preprocessing, and analysis, can be appropriated for use in other disciplines. These methodologies involve a sequential pipeline of quality checks to prepare data sets for analysis and application. Building upon these methodologies, this paper describes the Big Data Quality & Statistical Assurance (BDQSA) model, applicable for researchers in the behavioral sciences. It involves a series of data preprocessing tasks, to achieve data understanding, as well as data screening, cleaning, and transformation. These are followed by a statistical quality phase, which includes extraction of the relevant data subset, type conversions, ensuring sample representativeness when appropriate, and assessing statistical assumptions. The resulting model thereby provides methodological guidance for the preprocessing of behavioral science big data, aimed at ensuring acceptable data quality before analysis is undertaken. Sample R code snippets demonstrating the application of this model are provided throughout the paper.

A terrain treadmill to study animal locomotion through large obstacles

A challenge to understanding locomotion in complex 3-D terrain with large obstacles is to create tools for controlled, systematic experiments. Recent terrain arenas allow observations at small spatiotemporal scales (∼10 body length or cycles). Here, we create a terrain treadmill to enable high-resolution observation of animal locomotion through large obstacles over large spatiotemporal scales. An animal moves through modular obstacles on an inner sphere, while a rigidly-attached, concentric, transparent outer sphere rotates with the opposite velocity via closed-loop feedback to keep the animal atop. During sustained locomotion, a discoid cockroach moved through pillar obstacles for up to 25 minutes (2500 cycles) over 67 m (1500 body lengths). Over 12 trials totaling∼1 hour, the animal was maintained within a radius of 1 body length (4.5 cm) on top of the sphere 90% of the time. The high-resolution observation enables study of diverse locomotor behaviors and quantification of animal-obstacle interaction.

Wildlife Monitoring on the Edge: A Performance Evaluation of Embedded Neural Networks on Microcontrollers for Animal Behavior Classification

Monitoring animals’ behavior living in wild or semi-wild environments is a very interesting subject for biologists who work with them. The difficulty and cost of implanting electronic devices in this kind of animals suggest that these devices must be robust and have low power consumption to increase their battery life as much as possible. Designing a custom smart device that can detect multiple animal behaviors and that meets the mentioned restrictions presents a major challenge that is addressed in this work. We propose an edge-computing solution, which embeds an ANN in a microcontroller that collects data from an IMU sensor to detect three different horse gaits. All the computation is performed in the microcontroller to reduce the amount of data transmitted via wireless radio, since sending information is one of the most power-consuming tasks in this type of devices. Multiples ANNs were implemented and deployed in different microcontroller architectures in order to find the best balance between energy consumption and computing performance. The results show that the embedded networks obtain up to 97.96% ± 1.42% accuracy, achieving an energy efficiency of 450 Mops/s/watt.

Validating accelerometry-derived proxies of energy expenditure using the doubly labelled water method in the smallest penguin species

Understanding energy use is central to understanding an animal's physiological and behavioural ecology. However, directly measuring energy expenditure in free-ranging animals is inherently difficult. The doubly labelled water (DLW) method is widely used to investigate energy expenditure in a range of taxa. Although reliable, DLW data collection and analysis is both financially costly and time consuming. Dynamic body acceleration (e.g. VeDBA) calculated from animal-borne accelerometers has been used to determine behavioural patterns, and is increasingly being used as a proxy for energy expenditure. Still its performance as a proxy for energy expenditure in free-ranging animals is not well established and requires validation against established methods. In the present study, the relationship between VeDBA and the at-sea metabolic rate calculated from DLW was investigated in little penguins (Eudyptula minor) using three approaches. Both in a simple correlation and activity-specific approaches were shown to be good predictors of at-sea metabolic rate. The third approach using activity-specific energy expenditure values obtained from literature did not accurately calculate the energy expended by individuals. However, all three approaches were significantly strengthened by the addition of mean horizontal travel speed. These results provide validation for the use of accelerometry as a proxy for energy expenditure and show how energy expenditure may be influenced by both individual behaviour and environmental conditions.

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.

Remote Recognition of Moving Behaviors for Captive Harbor Seals Using a Smart-Patch System via Bluetooth Communication

Animal telemetry has been recognized as a core platform for exploring animal species due to future opportunities in terms of its contribution toward marine fisheries and living resources. Herein, biologging systems with pressure sensors are successfully implemented via open-source hardware platforms, followed by immediate application to captive harbor seals (HS). Remotely captured output voltage signals in real-time mode via Bluetooth communication were reproducibly and reliably recorded on the basis of hours using a smartphone built with data capturing software with graphic user interface (GUI). Output voltages, corresponding to typical behaviors on the captive HS, such as stopping (A), rolling (B), flapping (C), and sliding (D), are clearly obtained, and their analytical interpretation on captured electrical signals are fully validated via a comparison study with consecutively captured images for each motion of the HS. Thus, the biologging system with low cost and light weight, which is fully compatible with a conventional smartphone, is expected to potentially contribute toward future anthology of seal animals.

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.

Observing the unwatchable: Integrating automated sensing, naturalistic observations and animal social network analysis in the age of big data

In the 4.5 decades since Altmann (1974) published her seminal paper on the methods for the observational study of behaviour, automated detection and analysis of social interaction networks have fundamentally transformed the ways that ecologists study social behaviour. Methodological developments for collecting data remotely on social behaviour involve indirect inference of associations, direct recordings of interactions and machine vision. These recent technological advances are improving the scale and resolution with which we can dissect interactions among animals. They are also revealing new intricacies of animal social interactions at spatial and temporal resolutions as well as in ecological contexts that have been hidden from humans, making the unwatchable seeable. We first outline how these technological applications are permitting researchers to collect exquisitely detailed information with little observer bias. We further recognize new emerging challenges from these new reality-mining approaches. While technological advances in automating data collection and its analysis are moving at an unprecedented rate, we urge ecologists to thoughtfully combine these new tools with classic behavioural and ecological monitoring methods to place our understanding of animal social networks within fundamental biological contexts.

Breeders are less active foragers than non-breeders in wild Damaraland mole-rats

Eusocial societies are characterized by a clear division of labour between non-breeding workers and breeding queens, and queens often do not contribute to foraging, defence and other maintenance tasks. It has been suggested that the structure and organization of social mole-rat groups resembles that of eusocial insect societies. However, the division of labour has rarely been investigated in wild mole-rats, and it is unknown whether breeders show decreased foraging activity compared with non-breeding helpers in natural groups. Here, we show that, in wild Damaraland mole-rats (Fukomys damarensis), breeders show lower activity in foraging areas than non-breeding group members. Both breeders and non-breeders displayed variation in activity across the different seasons. Our results suggest that group living allows social mole-rat breeders to reduce their investment in energetically costly behaviour, or alternatively, that the high cost of reproduction in this species forces a behavioural trade-off against foraging investment.

Methods for Monitoring Large Terrestrial Animals in the Wild

Reliable information about wildlife is absolutely important for making informed management decisions. The issues with the effectiveness of the control and monitoring of both large and small wild animals are relevant to assess and protect the world’s biodiversity. Monitoring becomes part of the methods in wildlife ecology for observation, assessment, and forecasting of the human environment. World practice reveals the potential of the joint application of both proven traditional and modern technologies using specialized equipment to organize environmental control and management processes. Monitoring large terrestrial animals require an individual approach due to their low density and larger habitat. Elk/moose are such animals. This work aims to evaluate the methods for monitoring large wild animals, suitable for controlling the number of elk/moose in the framework of nature conservation activities. Using different models allows determining the population size without affecting the animals and without significant financial costs. Although, the accuracy of each model is determined by its postulates implementation and initial conditions that need statistical data. Depending on the geographical, climatic, and economic conditions in each territory, it is possible to use different tools and equipment (e.g., cameras, GPS sensors, and unmanned aerial vehicles), a flexible variation of which will allow reaching the golden mean between the desires and capabilities of researchers.

Challenges of machine learning model validation using correlated behaviour data: Evaluation of cross-validation strategies and accuracy measures

Automated monitoring of the movements and behaviour of animals is a valuable research tool. Recently, machine learning tools were applied to many species to classify units of behaviour. For the monitoring of wild species, collecting enough data for training models might be problematic, thus we examine how machine learning models trained on one species can be applied to another closely related species with similar behavioural conformation. We contrast two ways to calculate accuracies, termed here as overall and threshold accuracy, because the field has yet to define solid standards for reporting and measuring classification performances. We measure 21 dogs and 7 wolves, and find that overall accuracies are between 51 and 60% for classifying 8 behaviours (lay, sit, stand, walk, trot, run, eat, drink) when training and testing data are from the same species and between 41 and 51% when training and testing is cross-species. We show that using data from dogs to predict the behaviour of wolves is feasible. We also show that optimising the model for overall accuracy leads to similar overall and threshold accuracies, while optimizing for threshold accuracy leads to threshold accuracies well above 80%, but yielding very low overall accuracies, often below the chance level. Moreover, we show that the most common method for dividing the data between training and testing data (random selection of test data) overestimates the accuracy of models when applied to data of new specimens. Consequently, we argue that for the most common goals of animal behaviour recognition overall accuracy should be the preferred metric. Considering, that often the goal is to collect movement data without other methods of observation, we argue that training data and testing data should be divided by individual and not randomly.

Use of a Tri-Axial Accelerometer Can Reliably Detect Play Behaviour in Newborn Calves

Traditionally, the welfare assessment of farm animals has focused on health and production outcomes. Positive welfare is, however, not merely the absence of negative welfare and is an important part of a life worth living. Play behaviour is widely considered to be an indicator of positive emotions because it is a "luxury" behaviour. Direct visual observation is considered the most accurate method of behavioural analysis, but it is time consuming and laborious. There is increasing interest in the use of remote monitoring technology to quantify behaviour. We compared the data output ("motion index" (MI)) from a commercially available tri-axial accelerometer fitted to newborn dairy calves to video footage of the same calves, with a focus on play behaviour. The motion index values over 48 h were positively correlated with both the duration of play behaviour and the number of play bouts. The motion index threshold in each sample interval with the optimal sensitivity and specificity for the identification of play behaviour was MI ≥ 2.5 at a 1 min resolution (sensitivity (Se) = 98.0%; specificity (Sp) = 92.9%) and MI ≥ 24.5 at a 15 min resolution (Se = 98.0%; Sp = 89.9%), but these values consistently overestimated the overall proportion of sample intervals in which play was observed. The MI that best reflected the results obtained from visual one-zero sampling was MI ≥ 23 for 1 min intervals and MI ≥ 62 for 15 min intervals-this may therefore be the basis of a more conservative approach to the identification of play behaviour from accelerometer-generated data. Our results indicate that accelerometer-generated data can usefully indicate the amount of play behaviour shown by newborn calves for up to 48 h, providing an efficient method for identifying this important parameter in future work.

Exploration profiles drive activity patterns and temporal niche specialization in a wild rodent

Individual niche specialization can have important consequences for competition, fitness, and, ultimately, population dynamics and ecological speciation. The temporal window and the level of daily activity are niche components that may vary with sex, breeding season, food supply, population density, and predator’s circadian rhythm. More recently, ecologists emphasized that traits such as dispersal and space use could depend on personality differences. Boldness and exploration have been shown to correlate with variation in foraging patterns, habitat use, and home range. Here, we assessed the link between exploration, measured from repeated novel environment tests, activity patterns, and temporal niche specialization in wild eastern chipmunks (Tamias striatus). Intrinsic differences in exploration should drive daily activity patterns through differences in energy requirements, space use, or the speed to access resources. We used collar-mounted accelerometers to assess whether individual exploration profiles predicted: 1) daily overall dynamic body acceleration, reflecting overall activity levels; 2) mean activity duration and the rate of activity sequences, reflecting the structure of daily activity; and 3) patterns of dawn and dusk activity, reflecting temporal niche differentiation. Exploration and overall activity levels were weakly related. However, both dawn activity and rate of activity sequences increased with the speed of exploration. Overall, activity patterns varied according to temporal variability in food conditions. This study emphasizes the role of intrinsic behavioral differences in activity patterns in a wild animal population. Future studies will help us understand how yearly seasonality in reproduction, food abundance, and population density modulate personality-dependent foraging patterns and temporal niche specialization.

The Design of an Automated System for the Analysis of the Activity and Emotional Patterns of Dogs with Wearable Sensors Using Machine Learning

The safety and welfare of companion animals such as dogs has become a large challenge in the last few years. To assess the well-being of a dog, it is very important for human beings to understand the activity pattern of the dog, and its emotional behavior. A wearable, sensor-based system is suitable for such ends, as it will be able to monitor the dogs in real-time. However, the question remains unanswered as to what kind of data should be used to detect the activity patterns and emotional patterns, as does another: what should be the location of the sensors for the collection of data and how should we automate the system? Yet these questions remain unanswered, because to date, there is no such system that can address the above-mentioned concerns. The main purpose of this study was (1) to develop a system that can detect the activities and emotions based on the accelerometer and gyroscope signals and (2) to automate the system with robust machine learning techniques for implementing it for real-time situations. Therefore, we propose a system which is based on the data collected from 10 dogs, including nine breeds of various sizes and ages, and both genders. We used machine learning classification techniques for automating the detection and evaluation process. The ground truth fetched for the evaluation process was carried out by taking video recording data in frame per second and the wearable sensors data were collected in parallel with the video recordings. Evaluation of the system was performed using an ANN (artificial neural network), random forest, SVM (support vector machine), KNN (k nearest neighbors), and a naïve Bayes classifier. The robustness of our system was evaluated by taking independent training and validation sets. We achieved an accuracy of 96.58% while detecting the activity and 92.87% while detecting emotional behavior, respectively. This system will help the owners of dogs to track their behavior and emotions in real-life situations for various breeds in different scenarios.

Ecological specialization, variability in activity patterns and response to environmental change

Differences in temporal patterns of activity can modulate the ambient conditions to which organisms are exposed, providing an important mechanism for responding to environmental change. Such differences may be particularly relevant to ecological generalists, which are expected to encounter a wider range of environmental conditions. Here, we compare temporal patterns of activity for partially sympatric populations of a generalist (the lodgepole chipmunk, Tamias speciosus) and a more specialized congener (the alpine chipmunk, Tamias alpinus) that have displayed divergent responses to the past century of environmental change. Although mean activity budgets were similar between species, analyses of individual-level variation in locomotion revealed that T. alpinus exhibited a narrower range of activity patterns than Tspeciosus Further analyses revealed that T. alpinus was more active earlier in the day, when temperatures were cooler, and that activity patterns for both species changed with increased interspecific co-occurrence. These results are consistent with the greater responsiveness of T. alpinus to changes in environmental conditions. In addition to highlighting the utility of accelerometers for collecting behavioural data, our findings add to a growing body of evidence, suggesting that the greater phenotypic variability displayed by ecological generalists may be critical to in situ responses to environmental change.

Behavioural compass: animal behaviour recognition using magnetometers

BACKGROUND

Animal-borne data loggers today often house several sensors recording simultaneously at high frequency. This offers opportunities to gain fine-scale insights into behaviour from individual-sensor as well as integrated multi-sensor data. In the context of behaviour recognition, even though accelerometers have been used extensively, magnetometers have recently been shown to detect specific behaviours that accelerometers miss. The prevalent constraint of limited training data necessitates the importance of identifying behaviours with high robustness to data from new individuals, and may require fusing data from both these sensors. However, no study yet has developed an end-to-end approach to recognise common animal behaviours such as foraging, locomotion, and resting from magnetometer data in a common classification framework capable of accommodating and comparing data from both sensors.

METHODS

We address this by first leveraging magnetometers' similarity to accelerometers to develop biomechanical descriptors of movement: we use the static component given by sensor tilt with respect to Earth's local magnetic field to estimate posture, and the dynamic component given by change in sensor tilt with time to characterise movement intensity and periodicity. We use these descriptors within an existing hybrid scheme that combines biomechanics and machine learning to recognise behaviour. We showcase the utility of our method on triaxial magnetometer data collected on ten wild Kalahari meerkats (Suricata suricatta), with annotated video recordings of each individual serving as groundtruth. Finally, we compare our results with accelerometer-based behaviour recognition.

RESULTS

The overall recognition accuracy of > 94% obtained with magnetometer data was found to be comparable to that achieved using accelerometer data. Interestingly, higher robustness to inter-individual variability in dynamic behaviour was achieved with the magnetometer, while the accelerometer was better at estimating posture.

CONCLUSIONS

Magnetometers were found to accurately identify common behaviours, and were particularly robust to dynamic behaviour recognition. The use of biomechanical considerations to summarise magnetometer data makes the hybrid scheme capable of accommodating data from either or both sensors within the same framework according to each sensor's strengths. This provides future studies with a method to assess the added benefit of using magnetometers for behaviour recognition.

Multiscale modeling of mechanotransduction in the utricle

We review recent progress in using numerical models to relate utricular hair bundle and otoconial membrane (OM) structure to the functional requirements imposed by natural behavior in turtles. The head movements section reviews the evolution of experimental attempts to understand vestibular system function with emphasis on turtles, including data showing that accelerations occurring during natural head movements achieve higher magnitudes and frequencies than previously assumed. The structure section reviews quantitative anatomical data documenting topographical variation in the structures underlying macromechanical and micromechanical responses of the turtle utricle to head movement: hair bundles, OM, and bundle-OM coupling. The macromechanics section reviews macromechanical models that incorporate realistic anatomical and mechanical parameters and reveal that the system is significantly underdamped, contrary to previous assumptions. The micromechanics: hair bundle motion and met currents section reviews work based on micromechanical models, which demonstrates that topographical variation in the structure of hair bundles and OM, and their mode of coupling, result in regional specializations for signaling of low frequency (or static) head position and high frequency head accelerations. We conclude that computational models based on empirical data are especially promising for investigating mechanotransduction in this challenging sensorimotor system.

A comparison of techniques for classifying behavior from accelerometers for two species of seabird

The behavior of many wild animals remains a mystery, as it is difficult to quantify behavior of species that cannot be easily followed throughout their daily or seasonal movements. Accelerometers can solve some of these mysteries, as they collect activity data at a high temporal resolution (<1 s), can be relatively small (<1 g) so they minimally disrupt behavior, and are increasingly capable of recording data for long periods. Nonetheless, there is a need for increased validation of methods to classify animal behavior from accelerometers to promote widespread adoption of this technology in ecology. We assessed the accuracy of six different behavioral assignment methods for two species of seabird, thick-billed murres (Uria lomvia) and black-legged kittiwakes (Rissa tridactyla). We identified three behaviors using tri-axial accelerometers: standing, swimming, and flying, after classifying diving using a pressure sensor for murres. We evaluated six classification methods relative to independent classifications from concurrent GPS tracking data. We used four variables for classification: depth, wing beat frequency, pitch, and dynamic acceleration. Average accuracy for all methods was >98% for murres, and 89% and 93% for kittiwakes during incubation and chick rearing, respectively. Variable selection showed that classification accuracy did not improve with more than two (kittiwakes) or three (murres) variables. We conclude that simple methods of behavioral classification can be as accurate for classifying basic behaviors as more complex approaches, and that identifying suitable accelerometer metrics is more important than using a particular classification method when the objective is to develop a daily activity or energy budget. Highly accurate daily activity budgets can be generated from accelerometer data using multiple methods and a small number of accelerometer metrics; therefore, identifying a suitable behavioral classification method should not be a barrier to using accelerometers in studies of seabird behavior and ecology.

Behavioral classification of low-frequency acceleration and temperature data from a free-ranging small mammal

The miniaturization and affordability of new technology is driving a biologging revolution in wildlife ecology with use of animal-borne data logging devices. Among many new biologging technologies, accelerometers are emerging as key tools for continuously recording animal behavior. Yet a critical, but under-acknowledged consideration in biologging is the trade-off between sampling rate and sampling duration, created by battery- (or memory-) related sampling constraints. This is especially acute among small animals, causing most researchers to sample at high rates for very limited durations. Here, we show that high accuracy in behavioral classification is achievable when pairing low-frequency acceleration recordings with temperature. We conducted 84 hr of direct behavioral observations on 67 free-ranging red squirrels (200-300 g) that were fitted with accelerometers (2 g) recording tri-axial acceleration and temperature at 1 Hz. We then used a random forest algorithm and a manually created decision tree, with variable sampling window lengths, to associate observed behavior with logger recorded acceleration and temperature. Finally, we assessed the accuracy of these different classifications using an additional 60 hr of behavioral observations, not used in the initial classification. The accuracy of the manually created decision tree classification using observational data varied from 70.6% to 91.6% depending on the complexity of the tree, with increasing accuracy as complexity decreased. Short duration behavior like running had lower accuracy than long-duration behavior like feeding. The random forest algorithm offered similarly high overall accuracy, but the manual decision tree afforded the flexibility to create a hierarchical tree, and to adjust sampling window length for behavioral states with varying durations. Low frequency biologging of acceleration and temperature allows accurate behavioral classification of small animals over multi-month sampling durations. Nevertheless, low sampling rates impose several important limitations, especially related to assessing the classification accuracy of short duration behavior.

High accuracy at low frequency: detailed behavioural classification from accelerometer data

Accelerometers are a valuable tool for studying animal behaviour and physiology where direct observation is unfeasible. However, giving biological meaning to multivariate acceleration data is challenging. Here, we describe a method that reliably classifies a large number of behaviours using tri-axial accelerometer data collected at the low sampling frequency of 1 Hz, using the dingo (Canis dingo) as an example. We used out-of-sample validation to compare the predictive performance of four commonly used classification models (random forest, k-nearest neighbour, support vector machine, and naïve Bayes). We tested the importance of predictor variable selection and moving window size for the classification of each behaviour and overall model performance. Random forests produced the highest out-of-sample classification accuracy, with our best-performing model predicting 14 behaviours with a mean accuracy of 87%. We also investigated the relationship between overall dynamic body acceleration (ODBA) and the activity level of each behaviour, given the increasing use of ODBA in ecophysiology as a proxy for energy expenditure. ODBA values for our four 'high activity' behaviours were significantly greater than all other behaviours, with an overall positive trend between ODBA and intensity of movement. We show that a random forest model of relatively low complexity can mitigate some major challenges associated with establishing meaningful ecological conclusions from acceleration data. Our approach has broad applicability to free-ranging terrestrial quadrupeds of comparable size. Our use of a low sampling frequency shows potential for deploying accelerometers over extended time periods, enabling the capture of invaluable behavioural and physiological data across different ontogenies.

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.