Optimizing acceleration-based ethograms: the use of variable-time versus fixed-time segmentation

How to treat mixed behavior segments in supervised machine learning of behavioural modes from inertial measurement data

The application of supervised machine learning methods to identify behavioural modes from inertial measurements of bio-loggers has become a standard tool in behavioural ecology. Several design choices can affect the accuracy of identifying the behavioural modes. One such choice is the inclusion or exclusion of segments consisting of more than a single behaviour (mixed segments) in the machine learning model training data. Currently, the common practice is to ignore such segments during model training. In this paper we tested the hypothesis that including mixed segments in model training will improve accuracy, as the model would perform better in identifying them in the test data. We test this hypothesis using a series of data simulations on four datasets of accelerometer data coupled with behaviour observations, obtained from four study species (Damaraland mole-rats, meerkats, olive baboons, polar bears). Results show that when a substantial proportion of the test data are mixed behaviour segments (above ~ 10%), including mixed segments in machine learning model training improves the accuracy of classification. These results were consistent across the four study species, and robust to changes in segment length, sample size, and degree of mixture within the mixed segments. However, we also find that in some cases (particularly in baboons) models trained with mixed segments show reduced accuracy in classifying test data containing only single behaviour (pure) segments, compared to models trained without mixed segments. Based on these results, we recommend that when the classification model is expected to deal with a substantial proportion of mixed behaviour segments (> 10%), it is beneficial to include them in model training, otherwise, it is unnecessary but also not harmful. The exception is when there is a basis to assume that the training data contains a higher rate of mixed segments than the actual (unobserved) data to be classified-such a situation may occur particularly when training data are collected in captivity and used to classify data from the wild. In this case, excess inclusion of mixed segments in training data should probably be avoided.

Supervised versus unsupervised approaches to classification of accelerometry data

Sophisticated animal-borne sensor systems are increasingly providing novel insight into how animals behave and move. Despite their widespread use in ecology, the diversity and expanding quality and quantity of data they produce have created a need for robust analytical methods for biological interpretation. Machine learning tools are often used to meet this need. However, their relative effectiveness is not well known and, in the case of unsupervised tools, given that they do not use validation data, their accuracy can be difficult to assess. We evaluated the effectiveness of supervised (n = 6), semi-supervised (n = 1), and unsupervised (n = 2) approaches to analyzing accelerometry data collected from critically endangered California condors (Gymnogyps californianus). Unsupervised K-means and EM (expectation-maximization) clustering approaches performed poorly, with adequate classification accuracies of <0.8 but very low values for kappa statistics (range: -0.02 to 0.06). The semi-supervised nearest mean classifier was moderately effective at classification, with an overall classification accuracy of 0.61 but effective classification only of two of the four behavioral classes. Supervised random forest (RF) and k-nearest neighbor (kNN) machine learning models were most effective at classification across all behavior types, with overall accuracies >0.81. Kappa statistics were also highest for RF and kNN, in most cases substantially greater than for other modeling approaches. Unsupervised modeling, which is commonly used for the classification of a priori-defined behaviors in telemetry data, can provide useful information but likely is instead better suited to post hoc definition of generalized behavioral states. This work also shows the potential for substantial variation in classification accuracy among different machine learning approaches and among different metrics of accuracy. As such, when analyzing biotelemetry data, best practices appear to call for the evaluation of several machine learning techniques and several measures of accuracy for each dataset under consideration.

Year-round behavioural time budgets of common woodpigeons inferred from acceleration data using machine learning

Abstract

Accelerometers capture rapid changes in animal motion, and the analysis of large quantities of such data using machine learning algorithms enables the inference of broad animal behaviour categories such as foraging, flying, and resting over long periods of time. We deployed GPS-GSM/GPRS trackers with tri-axial acceleration sensors on common woodpigeons (Columba palumbus) from Hesse, Germany (forest and urban birds) and from Lisbon, Portugal (urban park). We used three machine learning algorithms, Random Forest, Support Vector Machine, and Extreme Gradient Boosting, to classify the main behaviours of the birds, namely foraging, flying, and resting and calculated time budgets over the breeding and winter season. Woodpigeon time budgets varied between seasons, with more foraging time during the breeding season than in winter. Also, woodpigeons from different sites showed differences in the time invested in foraging. The proportion of time woodpigeons spent foraging was lowest in the forest habitat from Hesse, higher in the urban habitat of Hesse, and highest in the urban park in Lisbon. The time budgets we recorded contrast to previous findings in woodpigeons and reaffirm the importance of considering different populations to fully understand the behaviour and adaptation of a particular species to a particular environment. Furthermore, the differences in the time budgets of Woodpigeons from this study and previous ones might be related to environmental change and merit further attention and the future investigation of energy budgets.

Significance statement

In this study we took advantage of accelerometer technology and machine learning methods to investigate year-round behavioural time budgets of wild common woodpigeons (Columba palumbus). Our analysis focuses on identifying coarse-scale behaviours (foraging, flying, resting) using various machine learning algorithms. Woodpigeon time budgets varied between seasons and among sites. Particularly interesting is the result showing that urban woodpigeons spend more time foraging than forest conspecifics. Our study opens an opportunity to further investigate and understand how a successful bird species such as the woodpigeon copes with increasing environmental change and urbanisation. The increase in the proportion of time devoted to foraging might be one of the behavioural mechanisms involved but opens questions about the costs associated to such increase in terms of other important behaviours.

Estimation of the Maternal Investment of Sea Turtles by Automatic Identification of Nesting Behavior and Number of Eggs Laid from a Tri-Axial Accelerometer

Simple Summary

During the reproduction period, female sea turtles come several times onto the beaches to lay their eggs. Monitoring of the nesting populations is therefore important to estimate the state of a population and its future. However, measuring the clutch size and frequency of sea turtles is tedious work that requires rigorous monitoring of the nesting site throughout the breeding season. In order to support the fieldwork, we propose an automatic method to remotely record the behavior on land of the sea turtles from animal-attached sensors; an accelerometer. The proposed method estimates, with an accuracy of 95%, the behaviors on land of sea turtles and the number of eggs laid. This automatic method should therefore help researchers monitor nesting sea turtle populations and contribute to improving global knowledge on the demographic status of these threatened species.

Abstract

Monitoring reproductive outputs of sea turtles is difficult, as it requires a large number of observers patrolling extended beaches every night throughout the breeding season with the risk of missing nesting individuals. We introduce the first automatic method to remotely record the reproductive outputs of green turtles (Chelonia mydas) using accelerometers. First, we trained a fully convolutional neural network, the V-net, to automatically identify the six behaviors shown during nesting. With an accuracy of 0.95, the V-net succeeded in detecting the Egg laying process with a precision of 0.97. Then, we estimated the number of laid eggs from the predicted Egg laying sequence and obtained the outputs with a mean relative error of 7% compared to the observed numbers in the field. Based on deployment of non-invasive and miniature loggers, the proposed method should help researchers monitor nesting sea turtle populations. Furthermore, its use can be coupled with the deployment of accelerometers at sea during the intra-nesting period, from which behaviors can also be estimated. The knowledge of the behavior of sea turtle on land and at sea during the entire reproduction period is essential to improve our knowledge of this threatened species.

R package for animal behavior classification from accelerometer data-rabc

Increasingly, animal behavior studies are enhanced through the use of accelerometry. To allow translation of raw accelerometer data to animal behaviors requires the development of classifiers. Here, we present the "rabc" (r for animal behavior classification) package to assist researchers with the interactive development of such animal behavior classifiers in a supervised classification approach. The package uses datasets consisting of accelerometer data with their corresponding animal behaviors (e.g., for triaxial accelerometer data along the x, y and z axes arranged as "x, y, z, x, y, z,…, behavior"). Using an example dataset collected on white stork (Ciconia ciconia), we illustrate the workflow of this package, including accelerometer data visualization, feature calculation, feature selection, feature visualization, extreme gradient boost model training, validation, and, finally, a demonstration of the behavior classification results.

Video and acceleration records of streaked shearwaters allows detection of two foraging behaviours associated with large marine predators

The study of seabird behaviour has largely relied on animal-borne tags to gather information, requiring interpretation to estimate at-sea behaviours. Details of shallow-diving birds’ foraging are less known than deep-diving species due to difficulty in identifying shallow dives from biologging devices. Development of smaller video loggers allow a direct view of these birds’ behaviours, at the cost of short battery capacity. However, recordings from video loggers combined with relatively low power usage accelerometers give a means to develop a reliable foraging detection method. Combined video and acceleration loggers were attached to streaked shearwaters in Funakoshi-Ohshima Island (39°24’N,141°59’E) during the breeding season in 2018. Video recordings were classified into behavioural categories (rest, transit, and foraging) and a detection method was generated from the acceleration signals. Two foraging behaviours, surface seizing and foraging dives, are reported with video recordings. Surface seizing was comprised of successive take-offs and landings (mean duration 0.6 and 1.5s, respectively), while foraging dives were shallow subsurface dives (3.2s mean duration) from the air and water surface. Birds were observed foraging close to marine predators, including dolphins and large fish. Results of the behaviour detection method were validated against video recordings, with mean true and false positive rates of 90% and 0%, 79% and 5%, and 66% and <1%, for flight, surface seizing, and foraging dives, respectively. The detection method was applied to longer duration acceleration and GPS datasets collected during the 2018 and 2019 breeding seasons. Foraging trips lasted between 1 − 8 days, with birds performing on average 16 surface seizing events and 43 foraging dives per day, comprising <1% of daily activity, while transit and rest took up 55 and 40%, respectively. This foraging detection method can address the difficulties of recording shallow-diving foraging behaviour and provides a means to measure activity budgets across shallow diving seabird species.

An evaluation of machine learning classifiers for next-generation, continuous-ethogram smart trackers

BACKGROUND

Our understanding of movement patterns and behaviours of wildlife has advanced greatly through the use of improved tracking technologies, including application of accelerometry (ACC) across a wide range of taxa. However, most ACC studies either use intermittent sampling that hinders continuity or continuous data logging relying on tracker retrieval for data downloading which is not applicable for long term study. To allow long-term, fine-scale behavioural research, we evaluated a range of machine learning methods for their suitability for continuous on-board classification of ACC data into behaviour categories prior to data transmission.

METHODS

We tested six supervised machine learning methods, including linear discriminant analysis (LDA), decision tree (DT), support vector machine (SVM), artificial neural network (ANN), random forest (RF) and extreme gradient boosting (XGBoost) to classify behaviour using ACC data from three bird species (white stork Ciconia ciconia, griffon vulture Gyps fulvus and common crane Grus grus) and two mammals (dairy cow Bos taurus and roe deer Capreolus capreolus).

RESULTS

Using a range of quality criteria, SVM, ANN, RF and XGBoost performed well in determining behaviour from ACC data and their good performance appeared little affected when greatly reducing the number of input features for model training. On-board runtime and storage-requirement tests showed that notably ANN, RF and XGBoost would make suitable on-board classifiers.

CONCLUSIONS

Our identification of using feature reduction in combination with ANN, RF and XGBoost as suitable methods for on-board behavioural classification of continuous ACC data has considerable potential to benefit movement ecology and behavioural research, wildlife conservation and livestock husbandry.

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.

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.

Less is more: On-board lossy compression of accelerometer data increases biologging capacity

GPS‐tracking devices have been used in combination with a wide range of additional sensors to study animal behaviour, physiology and interaction with their environment. Tri‐axial accelerometers allow researchers to remotely infer the behaviour of individuals, at all places and times. Collection of accelerometer data is relatively cheap in terms of energy usage, but the amount of raw data collected generally requires much storage space and is particularly demanding in terms of energy needed for data transmission.

Here, we propose compressing the raw accelerometer (ACC) data into summary statistics within the tracking device (before transmission) to reduce data size, as a means to overcome limitations in storage and energy capacity.

We explored this type of lossy data compression in the accelerometer data of tagged Bewick's swans Cygnus columbianus bewickii collected in spring 2017. Using software settings in which bouts of 2 s of both raw ACC data and summary statistics were collected in parallel but with different bout intervals to keep total data size comparable, we created the opportunity for a direct comparison of time budgets derived by the two data collection methods.

We found that the data compression in our case yielded a six times reduction in data size per bout, and concurrent, similar decreases in storage and energy use of the device. We show that with the same accuracy of the behavioural classification, the freed memory and energy of the device can be used to increase the monitoring effort, resulting in a more detailed representation of the individuals’ time budget. Rare and/or short behaviours, such as daily roost flights, were picked up significantly more when collecting summary statistics instead of raw ACC data (but note differences in sampling rate). Such level of detail can be of essential importance, for instance to make a reliable estimate of the energy budgets of individuals.

In conclusion, we argue that this type of lossy data compression can be a well‐considered choice in study situations where limitations in energy and storage space of the device pose a problem. Ultimately, these developments can allow for long‐term and nearly continuous remote monitoring of the behaviour of free‐ranging animals.

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.

Short-term effects of GPS collars on the activity, behavior, and adrenal response of scimitar-horned oryx (Oryx dammah)

GPS collars have revolutionized the field of animal ecology, providing detailed information on animal movement and the habitats necessary for species survival. GPS collars also have the potential to cause adverse effects ranging from mild irritation to severe tissue damage, reduced fitness, and death. The impact of GPS collars on the behavior, stress, or activity, however, have rarely been tested on study species prior to release. The objective of our study was to provide a comprehensive assessment of the short-term effects of GPS collars fitted on scimitar-horned oryx (Oryx dammah), an extinct-in-the-wild antelope once widely distributed across Sahelian grasslands in North Africa. We conducted behavioral observations, assessed fecal glucocorticoid metabolites (FGM), and evaluated high-resolution data from tri-axial accelerometers. Using a series of datasets and methodologies, we illustrate clear but short-term effects to animals fitted with GPS collars from two separate manufacturers (Advanced Telemetry Systems—G2110E; Vectronic Aerospace—Vertex Plus). Behavioral observations highlighted a significant increase in the amount of headshaking from pre-treatment levels, returning below baseline levels during the post-treatment period (>3 days post-collaring). Similarly, FGM concentrations increased after GPS collars were fitted on animals but returned to pre-collaring levels within 5 days of collaring. Lastly, tri-axial accelerometers, collecting data at eight positions per second, indicated a > 480 percent increase in the amount of hourly headshaking immediately after collaring. This post-collaring increase in headshaking was estimated to decline in magnitude within 4 hours after GPS collar fitting. These effects constitute a handling and/or habituation response (model dependent), with animals showing short-term responses in activity, behavior, and stress that dissipated within several hours to several days of being fitted with GPS collars. Importantly, none of our analyses indicated any long-term effects that would have more pressing animal welfare concerns.

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.

Using accelerometers to develop time-energy budgets of wild fur seals from captive surrogates

Background

Accurate time-energy budgets summarise an animal's energy expenditure in a given environment, and are potentially a sensitive indicator of how an animal responds to changing resources. Deriving accurate time-energy budgets requires an estimate of time spent in different activities and of the energetic cost of that activity. Bio-loggers (e.g., accelerometers) may provide a solution for monitoring animals such as fur seals that make long-duration foraging trips. Using low resolution to record behaviour may aid in the transmission of data, negating the need to recover the device.

Methods

This study used controlled captive experiments and previous energetic research to derive time-energy budgets of juvenile Australian fur seals (Arctocephalus pusillus) equipped with tri-axial accelerometers. First, captive fur seals and sea lions were equipped with accelerometers recording at high (20 Hz) and low (1 Hz) resolutions, and their behaviour recorded. Using this data, machine learning models were trained to recognise four states-foraging, grooming, travelling and resting. Next, the energetic cost of each behaviour, as a function of location (land or water), season and digestive state (pre- or post-prandial) was estimated. Then, diving and movement data were collected from nine wild juvenile fur seals wearing accelerometers recording at high- and low- resolutions. Models developed from captive seals were applied to accelerometry data from wild juvenile Australian fur seals and, finally, their time-energy budgets were reconstructed.

Results

Behaviour classification models built with low resolution (1 Hz) data correctly classified captive seal behaviours with very high accuracy (up to 90%) and recorded without interruption. Therefore, time-energy budgets of wild fur seals were constructed with these data. The reconstructed time-energy budgets revealed that juvenile fur seals expended the same amount of energy as adults of similar species. No significant differences in daily energy expenditure (DEE) were found across sex or season (winter or summer), but fur seals rested more when their energy expenditure was expected to be higher. Juvenile fur seals used behavioural compensatory techniques to conserve energy during activities that were expected to have high energetic outputs (such as diving).

Discussion

As low resolution accelerometry (1 Hz) was able to classify behaviour with very high accuracy, future studies may be able to transmit more data at a lower rate, reducing the need for tag recovery. Reconstructed time-energy budgets demonstrated that juvenile fur seals appear to expend the same amount of energy as their adult counterparts. Through pairing estimates of energy expenditure with behaviour this study demonstrates the potential to understand how fur seals expend energy, and where and how behavioural compensations are made to retain constant energy expenditure over a short (dive) and long (season) period.

Food quality and quantity are more important in explaining foraging of an intermediate-sized mammalian herbivore than predation risk or competition

During times of high activity by predators and competitors, herbivores may be forced to forage in patches of low-quality food. However, the relative importance in determining where and what herbivores forage still remains unclear, especially for small- and intermediate-sized herbivores. Our objective was to test the relative importance of predator and competitor activity, and forage quality and quantity on the proportion of time spent in a vegetation type and the proportion of time spent foraging by the intermediate-sized herbivore European hare (Lepus europaeus). We studied red fox (Vulpes vulpes) as a predator species and European rabbit (Oryctolagus cuniculus) as a competitor. We investigated the time spent at a location and foraging time of hare using GPS with accelerometers. Forage quality and quantity were analyzed based on hand-plucked samples of a selection of the locally most important plant species in the diet of hare. Predator activity and competitor activity were investigated using a network of camera traps. Hares spent a higher proportion of time in vegetation types that contained a higher percentage of fibers (i.e., NDF). Besides, hares spent a higher proportion of time in vegetation types that contained relatively low food quantity and quality of forage (i.e., high percentage of fibers) during days that foxes (Vulpes vulpes) were more active. Also during days that rabbits (Oryctolagus cuniculus) were more active, hares spent a higher proportion of time foraging in vegetation types that contained a relatively low quality of forage. Although predation risk affected space use and foraging behavior, and competition affected foraging behavior, our study shows that food quality and quantity more strongly affected space use and foraging behavior than predation risk or competition. It seems that we need to reconsider the relative importance of the landscape of food in a world of fear and competition.

Decision-making by a soaring bird: time, energy and risk considerations at different spatio-temporal scales

Natural selection theory suggests that mobile animals trade off time, energy and risk costs with food, safety and other pay-offs obtained by movement. We examined how birds make movement decisions by integrating aspects of flight biomechanics, movement ecology and behaviour in a hierarchical framework investigating flight track variation across several spatio-temporal scales. Using extensive global positioning system and accelerometer data from Eurasian griffon vultures (Gyps fulvus) in Israel and France, we examined soaring-gliding decision-making by comparing inbound versus outbound flights (to or from a central roost, respectively), and these (and other) home-range foraging movements (up to 300 km) versus long-range movements (longer than 300 km). We found that long-range movements and inbound flights have similar features compared with their counterparts: individuals reduced journey time by performing more efficient soaring-gliding flight, reduced energy expenditure by flapping less and were more risk-prone by gliding more steeply between thermals. Age, breeding status, wind conditions and flight altitude (but not sex) affected time and energy prioritization during flights. We therefore suggest that individuals facing time, energy and risk trade-offs during movements make similar decisions across a broad range of ecological contexts and spatial scales, presumably owing to similarity in the uncertainty about movement outcomes.This article is part of the themed issue 'Moving in a moving medium: new perspectives on flight'.

Improved supervised classification of accelerometry data to distinguish behaviors of soaring birds

Soaring birds can balance the energetic costs of movement by switching between flapping, soaring and gliding flight. Accelerometers can allow quantification of flight behavior and thus a context to interpret these energetic costs. However, models to interpret accelerometry data are still being developed, rarely trained with supervised datasets, and difficult to apply. We collected accelerometry data at 140Hz from a trained golden eagle (Aquila chrysaetos) whose flight we recorded with video that we used to characterize behavior. We applied two forms of supervised classifications, random forest (RF) models and K-nearest neighbor (KNN) models. The KNN model was substantially easier to implement than the RF approach but both were highly accurate in classifying basic behaviors such as flapping (85.5% and 83.6% accurate, respectively), soaring (92.8% and 87.6%) and sitting (84.1% and 88.9%) with overall accuracies of 86.6% and 92.3% respectively. More detailed classification schemes, with specific behaviors such as banking and straight flights were well classified only by the KNN model (91.24% accurate; RF = 61.64% accurate). The RF model maintained its accuracy of classifying basic behavior classification accuracy of basic behaviors at sampling frequencies as low as 10Hz, the KNN at sampling frequencies as low as 20Hz. Classification of accelerometer data collected from free ranging birds demonstrated a strong dependence of predicted behavior on the type of classification model used. Our analyses demonstrate the consequence of different approaches to classification of accelerometry data, the potential to optimize classification algorithms with validated flight behaviors to improve classification accuracy, ideal sampling frequencies for different classification algorithms, and a number of ways to improve commonly used analytical techniques and best practices for classification of accelerometry data.

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.

Expectation-Maximization Binary Clustering for Behavioural Annotation

The growing capacity to process and store animal tracks has spurred the development of new methods to segment animal trajectories into elementary units of movement. Key challenges for movement trajectory segmentation are to (i) minimize the need of supervision, (ii) reduce computational costs, (iii) minimize the need of prior assumptions (e.g. simple parametrizations), and (iv) capture biologically meaningful semantics, useful across a broad range of species. We introduce the Expectation-Maximization binary Clustering (EMbC), a general purpose, unsupervised approach to multivariate data clustering. The EMbC is a variant of the Expectation-Maximization Clustering (EMC), a clustering algorithm based on the maximum likelihood estimation of a Gaussian mixture model. This is an iterative algorithm with a closed form step solution and hence a reasonable computational cost. The method looks for a good compromise between statistical soundness and ease and generality of use (by minimizing prior assumptions and favouring the semantic interpretation of the final clustering). Here we focus on the suitability of the EMbC algorithm for behavioural annotation of movement data. We show and discuss the EMbC outputs in both simulated trajectories and empirical movement trajectories including different species and different tracking methodologies. We use synthetic trajectories to assess the performance of EMbC compared to classic EMC and Hidden Markov Models. Empirical trajectories allow us to explore the robustness of the EMbC to data loss and data inaccuracies, and assess the relationship between EMbC output and expert label assignments. Additionally, we suggest a smoothing procedure to account for temporal correlations among labels, and a proper visualization of the output for movement trajectories. Our algorithm is available as an R-package with a set of complementary functions to ease the analysis.

AcceleRater: a web application for supervised learning of behavioral modes from acceleration measurements

BACKGROUND

The study of animal movement is experiencing rapid progress in recent years, forcefully driven by technological advancement. Biologgers with Acceleration (ACC) recordings are becoming increasingly popular in the fields of animal behavior and movement ecology, for estimating energy expenditure and identifying behavior, with prospects for other potential uses as well. Supervised learning of behavioral modes from acceleration data has shown promising results in many species, and for a diverse range of behaviors. However, broad implementation of this technique in movement ecology research has been limited due to technical difficulties and complicated analysis, deterring many practitioners from applying this approach. This highlights the need to develop a broadly applicable tool for classifying behavior from acceleration data.

DESCRIPTION

Here we present a free-access python-based web application called AcceleRater, for rapidly training, visualizing and using models for supervised learning of behavioral modes from ACC measurements. We introduce AcceleRater, and illustrate its successful application for classifying vulture behavioral modes from acceleration data obtained from free-ranging vultures. The seven models offered in the AcceleRater application achieved overall accuracy of between 77.68% (Decision Tree) and 84.84% (Artificial Neural Network), with a mean overall accuracy of 81.51% and standard deviation of 3.95%. Notably, variation in performance was larger between behavioral modes than between models.

CONCLUSIONS

AcceleRater provides the means to identify animal behavior, offering a user-friendly tool for ACC-based behavioral annotation, which will be dynamically upgraded and maintained.