The Use of Acceleration to Code for Animal Behaviours; A Case Study in Free-Ranging Eurasian Beavers Castor fiber

Classification of sex-dependent specific behaviours by tri-axial acceleration in the tegu lizard Salvator merianae

Validated patterns of behaviour detected by tri-axial acceleration in the laboratory can be used for remote measurements of free-living animals. The tegu lizard naturally occupies diverse biomes in South America and presents ecological threats in regions where it was artificially introduced. We aimed to validate the use of tri-axial acceleration to distinguish among behaviours of male and female tegus in captivity by comparing observed behaviours to recorded acceleration data. Adult animals were externally fitted with an accelerometer fixed between their scapulae to quantify anteroposterior, lateral, and dorsoventral acceleration. Video recordings of cameras positioned on the walls of the animal-holding arena documented behaviours. Behaviour patterns, such as resting, walking, and eating, were identified for both sexes, and nest building in females and courtship and copulation in males. Random Forest algorithm was used to validate the behaviour patterns from accelerometry data based on two models, random split (70 % training-30 % validation; RS) and leave-one-out (divided by individual; LOO). Although LOO showed lower accuracies than RS for all the acceleration data, nest building in females and copulation in males had high accuracies in both models. In contrast, the lowest accuracies for walking and eating indicates they may involve more inconsistent movement patterns. Comparing the results from RS and LOO, female behaviours may be more identifiable in the field using triaxial accelerometry than males. The identification of behaviours by accelerometry, especially related to reproduction, without the necessity for direct observation of the tegus would be helpful for conservation purposes, for both natural and invasive populations.

Investigation through Animal-Computer Interaction: A Proof-of-Concept Study for the Behavioural Experimentation of Colour Vision in Zoo-Housed Primates

Zoos are an important repository of animals, which have a wide range of visual systems, providing excellent opportunities to investigate many comparative questions in sensory ecology. However, behavioural testing must be carried out in an animal welfare-friendly manner, which is practical for zoo staff. Here, we present a proof-of-concept study to facilitate behavioural research on the sensory ecology of captive primates. A system consisting of a tablet computer and an automated feeder connected wirelessly was developed and presented to captive primate species to evaluate interactions with and without previous training. A colour stimulus, analogous to the Ishihara test, was used to check the level of interaction with the device, supporting future studies on sensory ecology with zoo animals. Animals were able to use the system successfully and displayed signs of learning to discriminate between the visual stimuli presented. We identified no risk for small primates in their interactions with the experimental setup without the presence of keepers. The use of electronic devices should be approached with caution to prevent accidents, as a standard practice for environmental enrichment for larger animals (e.g., spider monkeys). In the long term, the system developed here will allow us to address complex comparative questions about the functions of different visual systems in captive animals (i.e., dichromatic, trichromatic, etc.).

From land to ocean: One month for southern elephant seal pups to acquire aquatic skills prior to their first departure to sea

Weaned southern elephant seals (SES) quickly transition from terrestrial to aquatic life after a 5- to 6-week post-weaning period. At sea, juveniles and adult elephant seals present extreme, continuous diving behaviour. Previous studies have highlighted the importance of the post-weaning period for weanlings to prepare for the physiological challenges of their future sea life. However, very little is known about how their body condition during this period may influence the development of their behaviour and brain activities. To characterise changes in the behavioural and brain activity of weanlings prior to ocean departure, we implemented a multi-logger approach combining measurements of movements (related to behaviour), pressure (related to diving), and brain electrical activity. As pups age, the amount of time allocated to resting decreases in favour of physical activity. Most resting (9.6 ± 1.2 h/day) takes place during daytime, with periods of slow-wave sleep representing 4.9 ± 0.9 h/day during the first 2 weeks. Furthermore, an increasing proportion of physical activity transitions from land to shore. Additionally, pups in poorer condition (lean group) are more active earlier than those in better condition (corpulent group). Finally, at weaning, clear circadian activity with two peaks at dawn and dusk is observed, and this pattern remains unchanged during the 4 weeks on land. This circadian pattern matches the one observed in adults at sea, with more prey catches at dawn and dusk, raising the question of whether it is endogenous or triggered by the mother during lactation.

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.

Accelerometer-based predictions of behaviour elucidate factors affecting the daily activity patterns of spotted hyenas

Animal activity patterns are highly variable and influenced by internal and external factors, including social processes. Quantifying activity patterns in natural settings can be challenging, as it is difficult to monitor animals over long time periods. Here, we developed and validated a machine-learning-based classifier to identify behavioural states from accelerometer data of wild spotted hyenas (Crocuta crocuta), social carnivores that live in large fission-fusion societies. By combining this classifier with continuous collar-based accelerometer data from five hyenas, we generated a complete record of activity patterns over more than one month. We used these continuous behavioural sequences to investigate how past activity, individual idiosyncrasies, and social synchronization influence hyena activity patterns. We found that hyenas exhibit characteristic crepuscular-nocturnal daily activity patterns. Time spent active was independent of activity level on previous days, suggesting that hyenas do not show activity compensation. We also found limited evidence for an effect of individual identity on activity, and showed that pairs of hyenas who synchronized their activity patterns must have spent more time together. This study sheds light on the patterns and drivers of activity in spotted hyena societies, and also provides a useful tool for quantifying behavioural sequences from accelerometer data.

Validation of inertial measurement units to detect and predict horse behaviour while stabled

BACKGROUND

Musculoskeletal injuries are observed in Thoroughbred racehorses and may become catastrophic. Currently, there are limited methods for early detection of such injuries. Most injuries develop gradually due to accumulated damage, providing the opportunity for early detection. Horses experiencing pain or lameness may exhibit changes in behaviour so the development of an objective, real-time system monitoring horse behaviour may enable detection of bone injuries before catastrophic failure.

OBJECTIVES

To determine whether intensive observational methods of assessing horse behaviour can be replaced by use of inertial measurement units (IMUs).

STUDY DESIGN

Validation study assessing IMU use against video observation.

METHODS

Six hospitalised Thoroughbreds (algorithm training data) and 19 Thoroughbred racehorses in-training (algorithm testing data) were equipped with an IMU placed on the lateral side of both forelimbs (left fore, LF; right fore, RF) and monitored in a stable for 4 h. An algorithm was developed to classify behaviour and then validated against video recordings.

RESULTS

Standing was the most prevalent behaviour (LF 88.8%, 95% confidence interval [CI] 88.7-89.0; RF 88.5%, 95% CI 88.4-88.7). IMU classification of recumbent and standing activities showed excellent agreement (sensitivity) with video observation (>98%). This was followed by stepping (LF 89.4%, RF 85.5%) then weight-shifting (LF 54.3%, RF 61.5%). Predictions from the algorithm showed misclassification of 2.5% (LF 5500/225 352, RF 5218/210 170). Excluding standing, misclassification was 6.8% (1705/25 158) and 7.5% (1812/24 077) for the left and right forelimbs, respectively, with pawing and weight-shifting most frequently misclassified.

MAIN LIMITATIONS

Increasing the number of horses and types of behaviours observed may improve predictions.

CONCLUSIONS

IMUs displayed a high sensitivity to movement on a small number of horses, and with further validation they have the potential to effectively monitor behaviour of racehorses in training. However, more sensitive methods may be needed to validate the classification of weight-shifting behaviour. Future studies should evaluate the association between each behaviour and musculoskeletal injury.

Structural and Functional Analysis of a Prokaryotic (6-4) Photolyase from the Aquatic Pathogen Vibrio Cholerae†

Photolyases are flavoproteins, which are able to repair UV-induced DNA lesions in a light-dependent manner. According to their substrate, they can be distinguished as CPD- and (6-4) photolyases. While CPD-photolyases repair the predominantly occurring cyclobutane pyrimidine dimer lesion, (6-4) photolyases catalyze the repair of the less prominent (6-4) photoproduct. The subgroup of prokaryotic (6-4) photolyases/FeS-BCP is one of the most ancient types of flavoproteins in the ubiquitously occurring photolyase & cryptochrome superfamily (PCSf). In contrast to canonical photolyases, prokaryotic (6-4) photolyases possess a few particular characteristics, including a lumazine derivative as antenna chromophore besides the catalytically essential flavin adenine dinucleotide as well as an elongated linker region between the N-terminal α/β-domain and the C-terminal all-α-helical domain. Furthermore, they can harbor an additional short subdomain, located at the C-terminus, with a binding site for a [4Fe-4S] cluster. So far, two crystal structures of prokaryotic (6-4) photolyases have been reported. Within this study, we present the high-resolution structure of the prokaryotic (6-4) photolyase from Vibrio cholerae and its spectroscopic characterization in terms of in vitro photoreduction and DNA-repair activity.

Evaluating Behavior Recognition Pipeline of Laying Hens Using Wearable Inertial Sensors

Recently, animal welfare has gained worldwide attention. The concept of animal welfare encompasses the physical and mental well-being of animals. Rearing layers in battery cages (conventional cages) may violate their instinctive behaviors and health, resulting in increased animal welfare concerns. Therefore, welfare-oriented rearing systems have been explored to improve their welfare while maintaining productivity. In this study, we explore a behavior recognition system using a wearable inertial sensor to improve the rearing system based on continuous monitoring and quantifying behaviors. Supervised machine learning recognizes a variety of 12 hen behaviors where various parameters in the processing pipeline are considered, including the classifier, sampling frequency, window length, data imbalance handling, and sensor modality. A reference configuration utilizes a multi-layer perceptron as a classifier; feature vectors are calculated from the accelerometer and angular velocity sensor in a 1.28 s window sampled at 100 Hz; the training data are unbalanced. In addition, the accompanying results would allow for a more intensive design of similar systems, estimation of the impact of specific constraints on parameters, and recognition of specific behaviors.

Testing the Accuracy of Wearable Technology to Assess Sleep Behaviour in Domestic Dogs: A Prospective Tool for Animal Welfare Assessment in Kennels

Sleep is a physiological process that has been shown to impact both physical and psychological heath of individuals when compromised; hence, it has the potential to be used as an indicator of animal welfare. Nonetheless, evaluating sleep in non-human species normally involves manipulation of the subjects (i.e., placement of electrodes on the cranium), and most studies are conducted in a laboratory setting, which limits the generalisability of information obtained, and the species investigated. In this study, we evaluated an alternative method of assessing sleep behaviour in domestic dogs, using a wearable sensor, and compared the measurements obtained to behavioural observations to evaluate accuracy. Differences between methods ranged from 0.13% to 59.3% for diurnal observations and 0.1% to 95.9% for nocturnal observations for point-by-point observations. Comparisons between methods showed significant differences in certain behaviours, such as inactivity and activity for diurnal recordings. However, total activity and total sleep recorded did not differ statistically between methods. Overall, the wearable technology tested was found to be a useful, and a less-time consuming, tool in comparison to direct behavioural observations for the evaluation of behaviours and their indication of wellbeing in dogs. The agreement between the wearable technology and directly observed data ranged from 75% to 99% for recorded behaviours, and these results are similar to previous findings in the literature.

Quantifying allo-grooming in wild chacma baboons (Papio ursinus) using tri-axial acceleration data and machine learning

Quantification of activity budgets is pivotal for understanding how animals respond to changes in their environment. Social grooming is a key activity that underpins various social processes with consequences for health and fitness. Traditional methods use direct (focal) observations to calculate grooming rates, providing systematic but sparse data. Accelerometers, in contrast, can quantify activity budgets continuously but have not been used to quantify social grooming. We test whether grooming can be accurately identified using machine learning (random forest model) trained on labelled acceleration data from wild chacma baboons (Papio ursinus). We successfully identified giving and receiving grooming with high precision (81% and 91%) and recall (87% and 79%). Giving grooming was associated with a distinct rhythmical signal along the surge axis. Receiving grooming had similar acceleration signals to resting, and thus was more difficult to assign. We applied our machine learning model to n = 680 collar data days from n = 12 baboons and found that grooming rates obtained from accelerometers were significantly and positively correlated with direct observation rates for giving but not receiving grooming. The ability to collect continuous grooming data in wild populations will allow researchers to re-examine and expand upon long-standing questions regarding the formation and function of grooming bonds.

Developing a classification system to assign activity states to two species of freshwater turtles

Research in ecology often requires robust assessment of animal behaviour, but classifying behavioural patterns in free-ranging animals and in natural environments can be especially challenging. New miniaturised bio-logging devices such as accelerometers are increasingly available to record animal behaviour remotely, and thereby address the gap in knowledge related to behaviour of free-ranging animals. However, validation of these data is rarely conducted and classification model transferability across closely-related species is often not tested. Here, we validated accelerometer and water sensor data to classify activity states in two free-ranging freshwater turtle species (Blanding's turtle, Emydoidea blandingii, and Painted turtle, Chrysemys picta). First, using only accelerometer data, we developed a decision tree to separate motion from motionless states, and second, we included water sensor data to classify the animal as being motionless or in-motion on land or in water. We found that accelerometers separated in-motion from motionless behaviour with > 83% accuracy, whereas models also including water sensor data predicted states in terrestrial and aquatic locations with > 77% accuracy. Despite differences in values separating activity states between the two species, we found high model transferability allowing cross-species application of classification models. Note that reducing sampling frequency did not affect predictive accuracy of our models up to a sampling frequency of 0.0625 Hz. We conclude that the use of accelerometers in animal research is promising, but requires prior data validation and development of robust classification models, and whenever possible cross-species assessment should be conducted to establish model generalisability.

Temporal Activity Patterns of the Eurasian Beaver and Coexisting Species in a Mediterranean Ecosystem

Analyses of temporal partitioning and overlaps in activity rhythms are pivotal to shed light on interspecific coexistence between similar species or prey and predators. In this work, we assessed the overlap of activity rhythms between the Eurasian beaver Castor fiber and its potential competitors and predators through camera trapping in an area in Central Italy. Interspecific overlaps of temporal activity patterns were estimated for the beavers, potential predators (the red fox Vulpes vulpes and the grey wolf Canis lupus), and a potential competitor, the coypu Myocastor coypus. The beavers showed a mostly crepuscular behaviour. Although high temporal overlap was observed between the Eurasian beavers and the red foxes and grey wolves, the activity of the beavers did not overlap with that of the predators. Accordingly, the beavers were more active on the darkest nights, i.e., avoiding bright moonlight.

Categorising cheetah behaviour using tri-axial accelerometer data loggers: a comparison of model resolution and data logger performance

BACKGROUND

Extinction is one of the greatest threats to the living world, endangering organisms globally, advancing conservation to the forefront of species research. To maximise the efficacy of conservation efforts, understanding the ecological, physiological, and behavioural requirements of vulnerable species is vital. Technological advances, particularly in remote sensing, enable researchers to continuously monitor movement and behaviours of multiple individuals simultaneously with minimal human intervention. Cheetahs, Acinonyx jubatus, constitute a "vulnerable" species for which only coarse behaviours have been elucidated. The aims of this study were to use animal-attached accelerometers to (1) determine fine-scale behaviours in cheetahs, (2) compare the performances of different devices in behaviour categorisation, and (3) provide a behavioural categorisation framework.

METHODS

Two different accelerometer devices (CEFAS, frequency: 30 Hz, maximum capacity: ~ 2 g; GCDC, frequency: 50 Hz, maximum capacity: ~ 8 g) were mounted onto collars, fitted to five individual captive cheetahs. The cheetahs chased a lure around a track, during which time their behaviours were videoed. Accelerometer data were temporally aligned with corresponding video footage and labelled with one of 17 behaviours. Six separate random forest models were run (three per device type) to determine the categorisation accuracy for behaviours at a fine, medium, and coarse resolution.

RESULTS

Fine- and medium-scale models had an overall categorisation accuracy of 83-86% and 84-88% respectively. Non-locomotory behaviours were best categorised on both loggers with GCDC outperforming CEFAS devices overall. On a coarse scale, both devices performed well when categorising activity (86.9% (CEFAS) vs. 89.3% (GCDC) accuracy) and inactivity (95.5% (CEFAS) vs. 95.0% (GCDC) accuracy). This study defined cheetah behaviour beyond three categories and accurately determined stalking behaviours by remote sensing. We also show that device specification and configuration may affect categorisation accuracy, so we recommend deploying several different loggers simultaneously on the same individual.

CONCLUSION

The results of this study will be useful in determining wild cheetah behaviour. The methods used here allowed broad-scale (active/inactive) as well as fine-scale (e.g. stalking) behaviours to be categorised remotely. These findings and methodological approaches will be useful in monitoring the behaviour of wild cheetahs and other species of conservation interest.

Decision rules for determining terrestrial movement and the consequences for filtering high-resolution global positioning system tracks: a case study using the African lion (Panthera leo)

The combined use of global positioning system (GPS) technology and motion sensors within the discipline of movement ecology has increased over recent years. This is particularly the case for instrumented wildlife, with many studies now opting to record parameters at high (infra-second) sampling frequencies. However, the detail with which GPS loggers can elucidate fine-scale movement depends on the precision and accuracy of fixes, with accuracy being affected by signal reception. We hypothesized that animal behaviour was the main factor affecting fix inaccuracy, with inherent GPS positional noise (jitter) being most apparent during GPS fixes for non-moving locations, thereby producing disproportionate error during rest periods. A movement-verified filtering (MVF) protocol was constructed to compare GPS-derived speed data with dynamic body acceleration, to provide a computationally quick method for identifying genuine travelling movement. This method was tested on 11 free-ranging lions (Panthera leo) fitted with collar-mounted GPS units and tri-axial motion sensors recording at 1 and 40 Hz, respectively. The findings support the hypothesis and show that distance moved estimates were, on average, overestimated by greater than 80% prior to GPS screening. We present the conceptual and mathematical protocols for screening fix inaccuracy within high-resolution GPS datasets and demonstrate the importance that MVF has for avoiding inaccurate and biased estimates of movement.

How often should dead-reckoned animal movement paths be corrected for drift?

BACKGROUND

Understanding what animals do in time and space is important for a range of ecological questions, however accurate estimates of how animals use space is challenging. Within the use of animal-attached tags, radio telemetry (including the Global Positioning System, 'GPS') is typically used to verify an animal's location periodically. Straight lines are typically drawn between these 'Verified Positions' ('VPs') so the interpolation of space-use is limited by the temporal and spatial resolution of the system's measurement. As such, parameters such as route-taken and distance travelled can be poorly represented when using VP systems alone. Dead-reckoning has been suggested as a technique to improve the accuracy and resolution of reconstructed movement paths, whilst maximising battery life of VP systems. This typically involves deriving travel vectors from motion sensor systems and periodically correcting path dimensions for drift with simultaneously deployed VP systems. How often paths should be corrected for drift, however, has remained unclear.

METHODS AND RESULTS

Here, we review the utility of dead-reckoning across four contrasting model species using different forms of locomotion (the African lion Panthera leo, the red-tailed tropicbird Phaethon rubricauda, the Magellanic penguin Spheniscus magellanicus, and the imperial cormorant Leucocarbo atriceps). Simulations were performed to examine the extent of dead-reckoning error, relative to VPs, as a function of Verified Position correction (VP correction) rate and the effect of this on estimates of distance moved. Dead-reckoning error was greatest for animals travelling within air and water. We demonstrate how sources of measurement error can arise within VP-corrected dead-reckoned tracks and propose advancements to this procedure to maximise dead-reckoning accuracy.

CONCLUSIONS

We review the utility of VP-corrected dead-reckoning according to movement type and consider a range of ecological questions that would benefit from dead-reckoning, primarily concerning animal-barrier interactions and foraging strategies.

Using Machine Learning for Remote Behaviour Classification-Verifying Acceleration Data to Infer Feeding Events in Free-Ranging Cheetahs

Behavioural studies of elusive wildlife species are challenging but important when they are threatened and involved in human-wildlife conflicts. Accelerometers (ACCs) and supervised machine learning algorithms (MLAs) are valuable tools to remotely determine behaviours. Here we used five captive cheetahs in Namibia to test the applicability of ACC data in identifying six behaviours by using six MLAs on data we ground-truthed by direct observations. We included two ensemble learning approaches and a probability threshold to improve prediction accuracy. We used the model to then identify the behaviours in four free-ranging cheetah males. Feeding behaviours identified by the model and matched with corresponding GPS clusters were verified with previously identified kill sites in the field. The MLAs and the two ensemble learning approaches in the captive cheetahs achieved precision (recall) ranging from 80.1% to 100.0% (87.3% to 99.2%) for resting, walking and trotting/running behaviour, from 74.4% to 81.6% (54.8% and 82.4%) for feeding behaviour and from 0.0% to 97.1% (0.0% and 56.2%) for drinking and grooming behaviour. The model application to the ACC data of the free-ranging cheetahs successfully identified all nine kill sites and 17 of the 18 feeding events of the two brother groups. We demonstrated that our behavioural model reliably detects feeding events of free-ranging cheetahs. This has useful applications for the determination of cheetah kill sites and helping to mitigate human-cheetah conflicts.

Food availability and long-term predation risk interactively affect antipredator response

Food availability and temporal variation in predation risk are both important determinants of the magnitude of antipredator responses, but their effects have rarely been examined simultaneously, particularly in wild prey. Here, we determine how food availability and long-term predation risk affect antipredator responses to acute predation risk by monitoring the foraging response of free-ranging snowshoe hares (Lepus americanus) to an encounter with a Canada lynx (Lynx canadensis) in Yukon, Canada, over four winters (2015-2016 to 2018-2019). We examined how this response was influenced by natural variation in long-term predation risk (2-month mortality rate of hares) while providing some individuals with supplemental food. On average, snowshoe hares reduced foraging time up to 10 h after coming into close proximity (≤75 m) with lynx, and reduced foraging time an average of 15.28 ± 7.08 min per lynx encounter. Hares tended to respond more strongly when the distance to lynx was shorter. More importantly, the magnitude of hares' antipredator response to a lynx encounter was affected by the interaction between food-supplementation and long-term predation risk. Food-supplemented hares reduced foraging time more than control hares after a lynx encounter under low long-term risk, but decreased the magnitude of the response as long-term risk increased. In contrast, control hares increased the magnitude of their response as long-term risk increased. Our findings show that food availability and long-term predation risk interactively drive the magnitude of reactive antipredator response to acute predation risk. Determining the factors driving the magnitude of antipredator responses would contribute to a better understanding of the indirect effects of predators on prey populations.

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.

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.

Novel use of an activity monitor to model jumping behaviors in cats

OBJECTIVE

To develop methods to identify and characterize activity monitor (AM) data signatures for jumps performed by cats.

ANIMALS

13 healthy, client-owned cats without evidence of osteoarthritis or degenerative joint disease.

PROCEDURES

Each cat was fitted with the same AM, individually placed in an observation room, then simultaneously recorded by 3 video cameras during the observation period (5 to 8 hours). Each cat was encouraged to jump up (JU), jump down (JD), and jump across (JA) during the observation period. Output from the AM was manually annotated for jumping events, each of which was characterized by functional data analysis yielding relevant coefficients. The coefficients were then used in linear discriminant analysis to differentiate recorded jumps as JUs, JDs, or JAs. To assess the model's ability to distinguish among the 3 jump types, a leave-one-out cross-validation method was used, and the misclassification error rate of the overall categorization of the model was calculated.

RESULTS

Of 731 jumping events, 29 were misclassified. Overall, the mean misclassification error rate per cat was 5.4% (range, 0% to 12.5%), conversely indicating a correct classification rate per cat of 94.6%.

CONCLUSIONS AND CLINICAL RELEVANCE

Results indicated that the model was successful in correctly identifying JUs, JDs, and JAs in healthy cats. With advancements in AM technology and data processing, there is potential for the model to be applied in clinical settings as a means to obtain objective outcome measures.

Behavioural inference from signal processing using animal-borne multi-sensor loggers: a novel solution to extend the knowledge of sea turtle ecology

The identification of sea turtle behaviours is a prerequisite to predicting the activities and time-budget of these animals in their natural habitat over the long term. However, this is hampered by a lack of reliable methods that enable the detection and monitoring of certain key behaviours such as feeding. This study proposes a combined approach that automatically identifies the different behaviours of free-ranging sea turtles through the use of animal-borne multi-sensor recorders (accelerometer, gyroscope and time-depth recorder), validated by animal-borne video-recorder data. We show here that the combination of supervised learning algorithms and multi-signal analysis tools can provide accurate inferences of the behaviours expressed, including feeding and scratching behaviours that are of crucial ecological interest for sea turtles. Our procedure uses multi-sensor miniaturized loggers that can be deployed on free-ranging animals with minimal disturbance. It provides an easily adaptable and replicable approach for the long-term automatic identification of the different activities and determination of time-budgets in sea turtles. This approach should also be applicable to a broad range of other species and could significantly contribute to the conservation of endangered species by providing detailed knowledge of key animal activities such as feeding, travelling and resting.

Estimates for energy expenditure in free-living animals using acceleration proxies: A reappraisal

It is fundamentally important for many animal ecologists to quantify the costs of animal activities, although it is not straightforward to do so. The recording of triaxial acceleration by animal-attached devices has been proposed as a way forward for this, with the specific suggestion that dynamic body acceleration (DBA) be used as a proxy for movement-based power. Dynamic body acceleration has now been validated frequently, both in the laboratory and in the field, although the literature still shows that some aspects of DBA theory and practice are misunderstood. Here, we examine the theory behind DBA and employ modelling approaches to assess factors that affect the link between DBA and energy expenditure, from the deployment of the tag, through to the calibration of DBA with energy use in laboratory and field settings. Using data from a range of species and movement modes, we illustrate that vectorial and additive DBA metrics are proportional to each other. Either can be used as a proxy for energy and summed to estimate total energy expended over a given period, or divided by time to give a proxy for movement-related metabolic power. Nonetheless, we highlight how the ability of DBA to predict metabolic rate declines as the contribution of non-movement-related factors, such as heat production, increases. Overall, DBA seems to be a substantive proxy for movement-based power but consideration of other movement-related metrics, such as the static body acceleration and the rate of change of body pitch and roll, may enable researchers to refine movement-based metabolic costs, particularly in animals where movement is not characterized by marked changes in body acceleration.

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.

Performance of GPS units for deployment on semiaquatic animals

Global Positioning System (GPS) technology is widely used in wildlife research to study animal movement and habitat use. In order to evaluate the quality and reliability of GPS data, the factors influencing the performance of these devices must be known, especially for semiaquatic species, because terrestrial and aquatic habitat might affect GPS performance differently. We evaluated the location error and fix success rate of three GPS receiver models in stationary tests and on a semi-aquatic mammal, the Eurasian beaver (Castor fiber). The location error during stationary tests was on average 15.7 m, and increased with increasing canopy closure, slope, and horizontal dilution of precision, potentially leading to the erroneous classification of GPS positions when studying habitat use in animals. In addition, the position of the GPS antenna (flat versus 90° tilted) affected the location error, suggesting that animal behavior affects GPS performance. The fix success rate was significantly higher during stationary tests compared to when GPS units were deployed on beavers (94% versus 86%). Further, GPS receivers did not obtain any positions underwater and underground, the latter potentially allowing the estimation of activity periods in animals that use lodges or burrows as shelter. We discuss the possibilities for data screening, the use of buffer zones along the shoreline, and combination with other data loggers to avoid the erroneous classification of GPS positions when studying habitat use.

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.

Combined use of two supervised learning algorithms to model sea turtle behaviours from tri-axial acceleration data

Accelerometers are becoming ever more important sensors in animal-attached technology, providing data that allow determination of body posture and movement and thereby helping to elucidate behaviour in animals that are difficult to observe. We sought to validate the identification of sea turtle behaviours from accelerometer signals by deploying tags on the carapace of a juvenile loggerhead (Caretta caretta), an adult hawksbill (Eretmochelys imbricata) and an adult green turtle (Chelonia mydas) at Aquarium La Rochelle, France. We recorded tri-axial acceleration at 50 Hz for each species for a full day while two fixed cameras recorded their behaviours. We identified behaviours from the acceleration data using two different supervised learning algorithms, Random Forest and Classification And Regression Tree (CART), treating the data from the adult animals as separate from the juvenile data. We achieved a global accuracy of 81.30% for the adult hawksbill and green turtle CART model and 71.63% for the juvenile loggerhead, identifying 10 and 12 different behaviours, respectively. Equivalent figures were 86.96% for the adult hawksbill and green turtle Random Forest model and 79.49% for the juvenile loggerhead, for the same behaviours. The use of Random Forest combined with CART algorithms allowed us to understand the decision rules implicated in behaviour discrimination, and thus remove or group together some 'confused' or under--represented behaviours in order to get the most accurate models. This study is the first to validate accelerometer data to identify turtle behaviours and the approach can now be tested on other captive sea turtle species.

Social information in equine movement gestalts

One model of signal evolution is based on the notion that behaviours become increasingly detached from their original biological functions to obtain a communicative value. Selection may not always favour the evolution of such transitions, for instance, if signalling is costly due to predators usurping signal production. Here, we collected inertial movement sensing data recorded from multiple locations in free-ranging horses (Equus caballus), which we subjected to a machine learning algorithm to extract kinematic gestalt profiles. This yielded surprisingly rich and multi-layered sets of information. In particular, we were able to discriminate identity, breed, sex and some personality traits from the overall movement patterns of freely moving subjects. Our study suggests that, by attending to movement gestalts, domestic horses, and probably many other group-living animals, have access to rich social information passively but reliably made available by conspecifics, a finding that we discuss in relation with current signal evolution theories.

Diving behavior in a free-living, semi-aquatic herbivore, the Eurasian beaver Castor fiber

Semi-aquatic mammals have secondarily returned to the aquatic environment, although they spend a major part of their life operating in air. Moving both on land, as well as in, and under water is challenging because such species are considered to be imperfectly adapted to both environments. We deployed accelerometers combined with a depth sensor to study the diving behavior of 12 free-living Eurasian beavers Castor fiber in southeast Norway between 2009 and 2011 to examine the extent to which beavers conformed with mass-dependent dive capacities, expecting them to be poorer than wholly aquatic species. Dives were generally shallow (<1 m) and of short duration (<30 s), suggesting that the majority of dives were aerobic. Dive parameters such as maximum diving depth, dive duration, and bottom phase duration were related to the effort during different dive phases and the maximum depth reached. During the descent, mean vectorial dynamic body acceleration (VeDBA-a proxy for movement power) was highest near the surface, probably due to increased upthrust linked to fur- and lung-associated air. Inconsistently though, mean VeDBA underwater was highest during the ascent when this air would be expected to help drive the animals back to the surface. Higher movement costs during ascents may arise from transporting materials up, the air bubbling out of the fur, and/or the animals' exhaling during the bottom phase of the dive. In a manner similar to other homeotherms, beavers extended both dive and bottom phase durations with diving depth. Deeper dives tended to have a longer bottom phase, although its duration was shortened with increased VeDBA during the bottom phase. Water temperature did not affect diving behavior. Overall, the beavers' dive profile (depth, duration) was similar to other semi-aquatic freshwater divers. However, beavers dived for only 2.8% of their active time, presumably because they do not rely on diving for food acquisition.

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.

Estimating activity of Japanese macaques (Macaca fuscata) using accelerometers

Accelerometers have been used to study both terrestrial and aquatic wildlife, mainly for mammal and bird species. In terrestrial mammals, there is a bias toward ungulates and carnivores, with fewer studies on nonhuman primates. In this study, we tested the use of accelerometers for studying the activity of Japanese macaques (Macaca fuscata). We modeled the activity of a male and a female subject by matching continuous focal observations from video recordings to sensor parameters derived from collar-mounted accelerometers. Models achieved classification performance (AUC) of greater than 90% for both subjects, with similar results when subjects were cross-validated. Accelerometer-based estimates of activity had comparable accuracies to estimates from instantaneous sampling at 1 min and 5 min intervals. We further demonstrated the use of model estimates for analyzing circadian rhythm and night time activity of M. fuscata. Our results add support to the feasibility of using accelerometers for studying activity of nonhuman primates. We discussed the limitations, benefits and potential applications of remote-sensing technology like accelerometers for advancing primalotogical studies.

Combined use of tri-axial accelerometers and GPS reveals the flexible foraging strategy of a bird in relation to weather conditions

Tri-axial accelerometry has proved to be a useful technique to study animal behavior with little direct observation, and also an effective way to measure energy expenditure, allowing a refreshing revisit to optimal foraging theory. This theory predicts that individuals should gain the most energy for the lowest cost in terms of time and energy when foraging, in order to maximize their fitness. However, during a foraging trip, central-place foragers could face different trade-offs during the commuting and searching parts of the trip, influencing behavioral decisions. Using the lesser kestrel (Falco naumanni) as an example we study the time and energy costs of different behaviors during the commuting and searching parts of a foraging trip. Lesser kestrels are small insectivorous falcons that behave as central-place foragers during the breeding season. They can commute by adopting either time-saving flapping flights or energy-saving soaring-gliding flights, and capture prey by using either time-saving active hovering flights or energy-saving perch-hunting. We tracked 6 lesser kestrels using GPS and tri-axial accelerometers during the breeding season. Our results indicate that males devoted more time and energy to flight behaviors than females in agreement with being the sex responsible for food provisioning to the nest. During the commuting flights, kestrels replaced flapping with soaring-gliding flights as solar radiation increased and thermal updrafts got stronger. In the searching part, they replaced perch-hunting with hovering as wind speed increased and they experienced a stronger lift. But also, they increased the use of hovering as air temperature increased, which has a positive influence on the activity level of the preferred prey (large grasshoppers). Kestrels maintained a constant energy expenditure per foraging trip, although flight and hunting strategies changed dramatically with weather conditions, suggesting a fixed energy budget per trip to which they adjusted their commuting and searching strategies in response to weather conditions.

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.

Wireless inertial measurement of head kinematics in freely-moving rats

While miniature inertial sensors offer a promising means for precisely detecting, quantifying and classifying animal behaviors, versatile inertial sensing devices adapted for small, freely-moving laboratory animals are still lacking. We developed a standalone and cost-effective platform for performing high-rate wireless inertial measurements of head movements in rats. Our system is designed to enable real-time bidirectional communication between the headborne inertial sensing device and third party systems, which can be used for precise data timestamping and low-latency motion-triggered applications. We illustrate the usefulness of our system in diverse experimental situations. We show that our system can be used for precisely quantifying motor responses evoked by external stimuli, for characterizing head kinematics during normal behavior and for monitoring head posture under normal and pathological conditions obtained using unilateral vestibular lesions. We also introduce and validate a novel method for automatically quantifying behavioral freezing during Pavlovian fear conditioning experiments, which offers superior performance in terms of precision, temporal resolution and efficiency. Thus, this system precisely acquires movement information in freely-moving animals, and can enable objective and quantitative behavioral scoring methods in a wide variety of experimental situations.