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

The urban lives of green sea turtles: Insights into behavior in an industrialized habitat using an animal-borne camera

The cryptic and aquatic life histories of sea turtles have made them a challenging group to directly observe, leaving significant knowledge gaps regarding social behavior and fine-scale elements of habitat use. Using a custom-designed animal-borne camera, we observed previously undocumented behaviors by green turtles (Chelonia mydas) at a foraging area in San Diego Bay, a highly urbanized ecosystem in California, USA. We deployed a suction-cup-attached pop-off camera (manufactured by Customized Animal Tracking Solutions) on 11 turtles (mean straight carapace length = 84.0 ± 11.2 cm) for between 1 and 30.8 h. Video recordings, limited to sunlit hours, provided 73 h of total observation time between May 2022 and June 2023. We observed 32 conspecific interactions; we classified 18 as active, entailing clear social behaviors, as compared with 14 passive interactions representing brief, chance encounters. There was no evidence for agonistic interactions. The camera additionally revealed that green turtles consistently use metal structures within urban San Diego Bay. In seven instances, turtles exhibited rubbing behavior against metal structures, and we observed two examples of turtles congregating at these structures. High rates of intraspecific interaction exhibited relatively consistently among individuals provide a compelling case for sociality for green turtles in San Diego Bay, adding to a growing research base updating their historical label of "non-social." The frequent use of metal structures by the population, in particular the rubbing of exposed skin, has implications for behavioral adaptations to urban environments. Our study exemplifies the promise of technological advances (e.g., underwater and animal-borne cameras) for updating natural history paradigms, even for well-studied populations.

Predictive model of sperm whale prey capture attempts from time-depth data

BACKGROUND

High-resolution sound and movement recording tags offer unprecedented insights into the fine-scale foraging behaviour of cetaceans, especially echolocating odontocetes, enabling the estimation of a series of foraging metrics. However, these tags are expensive, making them inaccessible to most researchers. Time-Depth Recorders (TDRs), which have been widely used to study diving and foraging behaviour of marine mammals, offer a more affordable alternative. Unfortunately, data collected by TDRs are bi-dimensional (time and depth only), so quantifying foraging effort from those data is challenging.

METHODS

A predictive model of the foraging effort of sperm whales (Physeter macrocephalus) was developed to identify prey capture attempts (PCAs) from time-depth data. Data from high-resolution acoustic and movement recording tags deployed on 12 sperm whales were downsampled to 1 Hz to match the typical TDR sampling resolution and used to predict the number of buzzes (i.e., rapid series of echolocation clicks indicative of PCAs). Generalized linear mixed models were built for dive segments of different durations (30, 60, 180 and 300 s) using multiple dive metrics as potential predictors of PCAs.

RESULTS

Average depth, variance of depth and variance of vertical velocity were the best predictors of the number of buzzes. Sensitivity analysis showed that models with segments of 180 s had the best overall predictive performance, with a good area under the curve value (0.78 ± 0.05), high sensitivity (0.93 ± 0.06) and high specificity (0.64 ± 0.14). Models using 180 s segments had a small difference between observed and predicted number of buzzes per dive, with a median of 4 buzzes, representing a difference in predicted buzzes of 30%.

CONCLUSIONS

These results demonstrate that it is possible to obtain a fine-scale, accurate index of sperm whale PCAs from time-depth data alone. This work helps leveraging the potential of time-depth data for studying the foraging ecology of sperm whales and the possibility of applying this approach to a wide range of echolocating cetaceans. The development of accurate foraging indices from low-cost, easily accessible TDR data would contribute to democratize this type of research, promote long-term studies of various species in several locations, and enable analyses of historical datasets to investigate changes in cetacean foraging activity.

Burmese pythons in Florida: A synthesis of biology, impacts, and management tools

Burmese pythons (Python molurus bivittatus) are native to southeastern Asia, however, there is an established invasive population inhabiting much of southern Florida throughout the Greater Everglades Ecosystem. Pythons have severely impacted native species and ecosystems in Florida and represent one of the most intractable invasive-species management issues across the globe. The difficulty stems from a unique combination of inaccessible habitat and the cryptic and resilient nature of pythons that thrive in the subtropical environment of southern Florida, rendering them extremely challenging to detect. Here we provide a comprehensive review and synthesis of the science relevant to managing invasive Burmese pythons. We describe existing control tools and review challenges to productive research, identifying key knowledge gaps that would improve future research and decision making for python control.

The role of individual variability on the predictive performance of machine learning applied to large bio-logging datasets

Animal-borne tagging (bio-logging) generates large and complex datasets. In particular, accelerometer tags, which provide information on behaviour and energy expenditure of wild animals, produce high-resolution multi-dimensional data, and can be challenging to analyse. We tested the performance of commonly used artificial intelligence tools on datasets of increasing volume and dimensionality. By collecting bio-logging data across several sampling seasons, datasets are inherently characterized by inter-individual variability. Such information should be considered when predicting behaviour. We integrated both unsupervised and supervised machine learning approaches to predict behaviours in two penguin species. The classified behaviours obtained from the unsupervised approach Expectation Maximisation were used to train the supervised approach Random Forest. We assessed agreement between the approaches, the performance of Random Forest on unknown data and the implications for the calculation of energy expenditure. Consideration of behavioural variability resulted in high agreement (> 80%) in behavioural classifications and minimal differences in energy expenditure estimates. However, some outliers with < 70% of agreement, highlighted how behaviours characterized by signal similarity are confused. We advise the broad bio-logging community, approaching these large datasets, to be cautious when upscaling predictions, as this might lead to less accurate estimates of behaviour and energy expenditure.

Multivariate analysis of biologging data reveals the environmental determinants of diving behaviour in a marine reptile

Diving behaviour of 'surfacers' such as sea snakes, cetaceans and turtles is complex and multi-dimensional, thus may be better captured by multi-sensor biologging data. However, analysing these large multi-faceted datasets remains challenging, though a high priority. We used high-resolution multi-sensor biologging data to provide the first detailed description of the environmental influences on flatback turtle (Natator depressus) diving behaviour, during its foraging life-history stage. We developed an analytical method to investigate seasonal, diel and tidal effects on diving behaviour for 24 adult flatback turtles tagged with biologgers. We extracted 16 dive variables associated with three-dimensional and kinematic characteristics for 4128 dives. K-means and hierarchical cluster analyses failed to identify distinct dive types. Instead, principal component analysis objectively condensed the dive variables, removing collinearity and highlighting the main features of diving behaviour. Generalized additive mixed models of the main principal components identified significant seasonal, diel and tidal effects on flatback turtle diving behaviour. Flatback turtles altered their diving behaviour in response to extreme tidal and water temperature ranges, displaying thermoregulation and predator avoidance strategies while likely optimizing foraging in this challenging environment. This study demonstrates an alternative statistical technique for objectively interpreting diving behaviour from multivariate collinear data derived from biologgers.

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.

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

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

Equine Activity Time Budgets: The Effect of Housing and Management Conditions on Geriatric Horses and Horses with Chronic Orthopaedic Disease

Housing and management conditions strongly influence the health, welfare and behaviour of horses. Consequently, objective and quantifiable comparisons between domestic environments and their influence on different equine demographics are needed to establish evidence-based criteria to assess and optimize horse welfare. Therefore, the present study aimed to measure and compare the time budgets (=percentage of time spent on specific activities) of horses with chronic orthopaedic disease and geriatric (≥20 years) horses living in different husbandry systems using an automated tracking device. Horses spent 42% (range 38.3-44.8%) of their day eating, 39% (range 36.87-44.9%) resting, and 19% (range 17-20.4%) in movement, demonstrating that geriatric horses and horses suffering from chronic orthopaedic disease can exhibit behaviour time budgets equivalent to healthy controls. Time budget analysis revealed significant differences between farms, turn-out conditions and time of day, and could identify potential areas for improvement. Horses living in open-air group housing on a paddock had a more uniform temporal distribution of feeding and movement activities with less pronounced peaks compared to horses living in more restricted husbandry systems.

Activity of loggerhead turtles during the U-shaped dive: insights using angular velocity metrics

Understanding the behavioural ecology of endangered taxa can inform conservation strategies. The activity budgets of the loggerhead turtle Caretta caretta are still poorly understood because many tracking methods show only horizontal displacement and ignore dives and associated behaviours. However, time-depth recorders have enabled researchers to identify flat, U-shaped dives (or type 1a dives) and these are conventionally labelled as resting dives on the seabed because they involve no vertical displacement of the animal. Video- and acceleration-based studies have demonstrated this is not always true. Focusing on sea turtles nesting on the Cabo Verde archipelago, we describe a new metric derived from magnetometer data, absolute angular velocity, that integrates indices of angular rotation in the horizontal plane to infer activity. Using this metric, we evaluated the variation in putative resting behaviours during the bottom phase of type 1a dives for 5 individuals over 13 to 17 d at sea during a single inter-nesting interval (over 75 turtle d in total). We defined absolute resting within the bottom phase of type 1a dives as periods with no discernible acceleration or angular movement. Whilst absolute resting constituted a significant proportion of each turtle’s time budget for this 1a dive type, turtles allocated 16-38% of their bottom time to activity, with many dives being episodic, comprised of intermittent bouts of rest and rotational activity. This implies that previously considered resting behaviours are complex and need to be accounted for in energy budgets, particularly since energy budgets may impact conservation strategies.

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.