Search

Abstract Animal tracking data are indispensable for understanding the ecology, behaviour and physiology of mobile or cryptic species. Meaningful signals in these data can be obscured by noise due to imperfect measurement technologies, requiring rigorous quality control as part of any comprehensive analysis. State–space models are powerful tools that separate signal from noise. These tools are ideal for quality control of error‐prone location data and for inferring where animals are and what they are doing when they record or transmit other information. However, these statistical models can be challenging and time‐consuming to fit to diverse animal tracking data sets. The R package aniMotum eases the tasks of conducting quality control on and inference of changes in movement from animal tracking data. This is achieved via: (1) a simple but extensible workflow that accommodates both novice and experienced users; (2) automated processes that alleviate complexity from data processing and model specification/fitting steps; (3) simple movement models coupled with a powerful numerical optimization approach for rapid and reliable model fitting. We highlight aniMotum's capabilities through three applications to real animal tracking data. Full R code for these and additional applications is included as Supporting Information, so users can gain a deeper understanding of how to use aniMotum for their own analyses.

Abstract Background State-space models are important tools for quality control and analysis of error-prone animal movement data. The near real-time (within 24 h) capability of the Argos satellite system can aid dynamic ocean management of human activities by informing when animals enter wind farms, shipping lanes, and other intensive use zones. This capability also facilitates the use of ocean observations from animal-borne sensors in operational ocean forecasting models. Such near real-time data provision requires rapid, reliable quality control to deal with error-prone Argos locations. Methods We formulate a continuous-time state-space model to filter the three types of Argos location data (Least-Squares, Kalman filter, and Kalman smoother), accounting for irregular timing of observations. Our model is deliberately simple to ensure speed and reliability for automated, near real-time quality control of Argos location data. We validate the model by fitting to Argos locations collected from 61 individuals across 7 marine vertebrates and compare model-estimated locations to contemporaneous GPS locations. We then test assumptions that Argos Kalman filter/smoother error ellipses are unbiased, and that Argos Kalman smoother location accuracy cannot be improved by subsequent state-space modelling. Results Estimation accuracy varied among species with Root Mean Squared Errors usually

Subsea infrastructure of the oil and gas industry attracts commercial fish species as well as megafauna including sea lions, turtles, sharks and whales. Potential impacts of this attraction, whether positive or negative, are unknown. As part of a pilot study, we deployed acoustic telemetry equipment around offshore infrastructure to assess its effectiveness in detecting tagged marine animals and to gain insights into patterns of megafauna occurrence around these structures. Acoustic receivers were placed around four oil and gas platforms and on two remotely operated vehicles (ROVs) on Australia’s North West Shelf. Two whale sharks (Rhincodon typus) tagged in the World Heritage Ningaloo Reef Marine Park were detected at two platforms, North Rankin A and Pluto, located up to 340 km to the northeast. The shark at North Rankin A was detected infrequently and only 15 times over ∼6 weeks. The shark at Pluto was detected each day of the 24-day deployment, in total 4,894 times. Detections at Pluto platform were highest during the day, with peaks at dusk and dawn. Our study indicates that acoustic telemetry around platforms may be an effective method for understanding how marine megafauna utilise these structures. We recommend collaborating with industry to undertake receiver detection range testing to understand the effectiveness of the method. Furthermore, future studies should co-occur with tagging programs at sites like Ningaloo Reef and around the structures themselves to maximise the probability of detecting animals at these sites, thereby improving our understanding of how marine megafauna interact with these structures.

Abstract Background Over the past 15 years, the integration of localised passive telemetry networks into centralised data repositories has greatly enhanced our ability to monitor the presence and movements of highly mobile and migratory species. These large-scale networks are now generating big data, allowing meta-analyses across multiple species, locations, and temporal scales. Broad-scale comparisons of animal movement metrics are constrained by the use of diverse analytical techniques among researchers. Accordingly, there is a need for a tool-set to assist in calculating animal movement metrics that can be easily applied to datasets from local studies to large-scale cooperative networks. Results We present a standardised framework and an associated analysis tool-set that facilitates the calculation of a range of activity space and movement metrics for passive telemetry datasets. Application of the tool-set is demonstrated using data from the Integrated Marine Observing System continental-scale network of underwater acoustic receivers. We show how the metrics can: (1) be directly compared among multiple species monitored at multiple sites; (2) be compared among multiple species tagged at a single study site; and (3) assess changes in activity space metrics over time. Conclusions Establishing a framework and tool-set to analyse data from large-scale networks progresses the field of passive telemetry beyond the traditional individual-, species-, or location-centric approaches to facilitate national- or international-scale outputs that better address important questions in the field of movement ecology.

Despite the low chance of a person being bitten by a shark, there are serious associated costs. Electronic deterrents are currently the only types of personal deterrent with empirical evidence of a substantial reduction in the probability of being bitten by a shark. We aimed to predict the number of people who could potentially avoid being bitten by sharks in Australia if they wear personal electronic deterrents. We used the Australian Shark Attack File from 1900 to 2020 to develop sinusoidal time-series models of per capita incidents, and then stochastically projected these to 2066. We predicted that up to 1063 people (range: 185–2118) could potentially avoid being bitten across Australia by 2066 if all people used the devices. Avoiding death and injury of people over the next half-century is of course highly desirable, especially when considering the additional costs associated with the loss of recreational, commercial and tourism revenue potentially in the tens to hundreds of millions of dollars following clusters of shark-bite events.

The energy expenditure of free-living fur seals and sea lions is difficult to measure directly, but may be indirectly derived from flipper stroke rate. We filmed 10 captive otariids swimming with accelerometers either attached to a harness (Daily Diary: sampling frequency 32 Hz, N=4) or taped to the fur (G6a+: 25 Hz, N=6). We used down sampling to derive four recording rates from each accelerometer (Daily Diary: 32, 16, 8, 4 Hz; G6a+: 25, 20, 10, 5 Hz). For each of these sampling frequencies, we derived 20 combinations of two parameters (RMW, the window size used to calculate the running mean; and m, the minimum number of points smaller than a local maxima used to detect a peak) from the dynamic acceleration of x, z and x+z, to estimate stroke rate from the accelerometers. These estimates differed by up to ∼20% in comparison to the actual number of foreflipper strokes counted from videos. RMW and the choice of axis used to make the calculations (x, z or x+z) had little effect on the overall differences, though the variability was reduced when using x+z. The best m varied depending on the axis used and the sampling frequency; a larger m was needed for higher sampling frequencies. This study demonstrates that when parameters are appropriately tuned, accelerometers are a simple yet valid tool for estimating the stroke rates of swimming otariids.

Abstract The growing number of large databases of animal tracking provides an opportunity for analyses of movement patterns at the scales of populations and even species. We used analytical approaches, developed to cope with “big data”, that require no ‘a priori’ assumptions about the behaviour of the target agents, to analyse a pooled tracking dataset of 272 elephant seals (Mirounga leonina) in the Southern Ocean, that was comprised of >500,000 location estimates collected over more than a decade. Our analyses showed that the displacements of these seals were described by a truncated power law distribution across several spatial and temporal scales, with a clear signature of directed movement. This pattern was evident when analysing the aggregated tracks despite a wide diversity of individual trajectories. We also identified marine provinces that described the migratory and foraging habitats of these seals. Our analysis provides evidence for the presence of intrinsic drivers of movement, such as memory, that cannot be detected using common models of movement behaviour. These results highlight the potential for “big data” techniques to provide new insights into movement behaviour when applied to large datasets of animal tracking.

Abstract We trialed the collection of blow samples using a waterproof electric multirotor (quadcopter) drone from two free‐ranging dolphin species, the abundant and approachable bottlenose dolphin (Tursiops aduncus) and the less common and boat shy humpback dolphin (Sousa sahulensis). This drone was fast, maneuverable, and quiet compared to other drones commonly used in studies of cetaceans and relative to their hearing thresholds. We were successful in collecting blow samples from four individual dolphins (three bottlenose dolphins and one humpback dolphin) in two groups. The success of obtaining samples was dependent on the individual dolphin's activity. We were successful in sampling when dolphins were resting and socializing but found that socializing dolphins were not predictable in their surfacing and direction and therefore do not recommend drone sampling socializing dolphins. The suitability and preference of the sampling technique over biopsy sampling is highly dependent on the dolphin activity. We also attempted to extract DNA from the blow samples with the aim of assessing the feasibility of using blow sampling by drone for population genetic studies. We were unsuccessful in extracting DNA and recommend that others attempting to sample dolphin blow with a drone should prioritize collecting a larger volume of blow that may yield adequate concentrations of DNA to be amplified. Blow sample volume could potentially be increased by sampling with more absorbent materials.

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.

Giardia and Cryptosporidium are amongst the most common protozoan parasites identified as causing enteric disease in pinnipeds. A number of Giardia assemblages and Cryptosporidium species and genotypes are common in humans and terrestrial mammals and have also been identified in marine mammals. To investigate the occurrence of these parasites in an endangered marine mammal, the Australian sea lion (Neophoca cinerea), genomic DNA was extracted from faecal samples collected from wild populations (n = 271) in Southern and Western Australia and three Australian captive populations (n = 19). These were screened using PCR targeting the 18S rRNA of Giardia and Cryptosporidium. Giardia duodenalis was detected in 28 wild sea lions and in seven captive individuals. Successful sequencing of the 18S rRNA gene assigned 27 Giardia isolates to assemblage B and one to assemblage A, both assemblages commonly found in humans. Subsequent screening at the gdh and β-giardin loci resulted in amplification of only one of the 35 18S rRNA positive samples at the β-giardin locus. Sequencing at the β-giardin locus assigned the assemblage B 18S rRNA confirmed isolate to assemblage AI. The geographic distribution of sea lion populations sampled in relation to human settlements indicated that Giardia presence in sea lions was highest in populations less than 25 km from humans. Cryptosporidium was not detected by PCR screening in either wild colonies or captive sea lion populations. These data suggest that the presence of G. duodenalis in the endangered Australian sea lion is likely the result of dispersal from human sources. Multilocus molecular analyses are essential for the determination of G. duodenalis assemblages and subsequent inferences on transmission routes to endangered marine mammal populations.

Blue whale (Balaenoptera musculus) movements are often driven by the availability of their prey in space and time. While globally blue whale populations undertake long-range migrations between feeding and breeding grounds, those in the northern Indian Ocean remain in low latitude waters throughout the year with the implication that the productivity of these waters is sufficient to support their energy needs. A part of this population remains around Sri Lanka where they are usually recorded close to the southern coast during the Northeast Monsoon. To investigate inter-annual variability in sighting locations, we conducted systematic Conductivity-Temperature-Depth (CTD) and visual surveys between January–March 2011 and January–March 2012. In 2011, there was a notable decrease in inshore sightings compared to 2009 and 2012 (p < 0.001). CTD data revealed that in 2011 there was increased freshwater in the upper water column accompanied by deeper upwelling than in 2012. We hypothesise that anomalous rainfall, along with higher turbidity resulting from river discharge, affected the productivity of the inshore waters and caused a shift in blue whale prey and, consequently, the distribution of the whales themselves. An understanding of how predators and their prey respond to environmental variability is important for predicting how these species will respond to long-term changes. This is especially important given the rapid temperature increases predicted for the semi-enclosed northern Indian Ocean.

The analysis of animal movement from telemetry data provides insights into how and why animals move. While traditional approaches to such analysis mostly focus on predicting animal states during movement, we describe an approach that allows us to identify representative movement patterns of different animal groups. To do this, we propose a carefully designed recurrent neural network and combine it with telemetry data for automatic feature extraction and identification of non-predefined representative patterns. In the experiment, we consider a particular marine predator species, the southern elephant seal, as an example. With our approach, we identify that the male seals in our data set share similar movement patterns when they are close to land. We identify this pattern recurring in a number of distant locations, consistent with alternative approaches from previous research.

Abstract In the Southern Ocean, wide‐ranging predators offer the opportunity to quantify how animals respond to differences in the environment because their behavior and population trends are an integrated signal of prevailing conditions within multiple marine habitats. Southern elephant seals in particular, can provide useful insights due to their circumpolar distribution, their long and distant migrations and their performance of extended bouts of deep diving. Furthermore, across their range, elephant seal populations have very different population trends. In this study, we present a data set from the International Polar Year project; Marine Mammals Exploring the Oceans Pole to Pole for southern elephant seals, in which a large number of instruments (N = 287) deployed on animals, encompassing a broad circum‐Antarctic geographic extent, collected in situ ocean data and at‐sea foraging metrics that explicitly link foraging behavior and habitat structure in time and space. Broadly speaking, the seals foraged in two habitats, the relatively shallow waters of the Antarctic continental shelf and the Kerguelen Plateau and deep open water regions. Animals of both sexes were more likely to exhibit area‐restricted search (ARS) behavior rather than transit in shelf habitats. While Antarctic shelf waters can be regarded as prime habitat for both sexes, female seals tend to move northwards with the advance of sea ice in the late autumn or early winter. The water masses used by the seals also influenced their behavioral mode, with female ARS behavior being most likely in modified Circumpolar Deepwater or northerly Modified Shelf Water, both of which tend to be associated with the outer reaches of the Antarctic Continental Shelf. The combined effects of (1) the differing habitat quality, (2) differing responses to encroaching ice as the winter progresses among colonies, (3) differing distances between breeding and haul‐out sites and high quality habitats, and (4) differing long‐term regional trends in sea ice extent can explain the differing population trends observed among elephant seal colonies.

Reproductive parameters of three closely-related sympatric species of wobbegong (Orectolobus ornatus, O. maculatus and O. halei) off New South Wales (NSW) were investigated to contribute to a biological basis for the management of a commercial fishery targeting wobbegongs. Estimates of the onset of maturity based on ovary condition were 729 mm, 1117 mm, and 1591 mm total length (TL) for O. ornatus, O. maculatus and O. halei respectively. The length at which 50% of the male population is mature based on clasper calcification was 803 mm, 1282 mm and 1784 mm TL for O. ornatus, O. maculatus and O. halei, respectively, and was similar to female onset of maturity based on uterus condition and to TL-at-maternity. These species of wobbegong had synchronous, triennial reproductive cycles. Follicles took 2 years to enlarge before ovulation. During the first year, follicles remained small, and then grew rapidly during the second year before ovulation during November. Gestation lasted ~10?11 months and parturition occurred during September?October. Mean litter sizes were nine (s.e. 0.5) and 21 (s.e. 1.5) for O. ornatus and O. maculatus, respectively, and increased with female total length in O. ornatus. Pregnant O. ornatus and O. maculatus were frequently caught in northern NSW and no pregnant wobbegongs, or females with large, yolky follicles were captured south of Sydney. Differences in the reproductive conditions of wobbegongs caught in northern and central NSW suggested geographically dependent reproductive behaviour. Knowledge of the reproductive parameters provided in this paper is necessary for adequate fisheries management and species conservation assessments. [ABSTRACT FROM AUTHOR]

Abstract Background Semi-automating the analyses of accelerometry data makes it possible to synthesize large data sets. However, when constructing activity budgets from accelerometry data, there are many methods to extract, analyse and report data and results. For instance, machine learning is a robust approach to classifying data. We used a new method, super learning, that combines base learners (different machine learning methods) in an optimal manner to achieve overall improved accuracy. Other facets of super learning include the number of behavioural categories to predict, the number of epochs (sample window size) used to split data for training and testing and the parameters on which to train the models. Results The super learner accurately classified behaviour categories with higher accuracy and lower variance than comparative models. For all models tested, using four behaviours, in comparison with six, achieved higher rates of accuracy. The number of epochs chosen also affected the accuracy with smaller epochs (7 and 13) performing better than longer epochs (25 and 75). Conclusions Correct model selection, training and testing are imperative to creating reliable and valid classification models. To do so means model fitting must use a wide array of selection criteria. We evaluated a number of these including model, number of behaviours to classify and epoch length and then used a parameter grid search to implement the models. We found that all criteria tested contributed to the models’ overall accuracies. Fewer behaviour categories and shorter epoch length improved the performance of all models tested. The super learner classified behaviours with higher accuracy and lower variance than other models tested. However, when using this model, users need to consider the additional human and computational time required for implementation. Machine learning is a powerful method for classifying the behaviour of animals from accelerometers. Care and consideration of the modelling parameters evaluated in this study are essential when using this type of statistical analysis.

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.

The southern coastline of Australia forms part of the worlds' only northern boundary current system. The Bonney Upwelling occurs every austral summer along the south-eastern South Australian coastline, a region that hosts over 80% of the worlds population of an endangered endemic otariid, the Australian sea lion. We present the first data on the movement characteristics and foraging behaviour of adult male Australian sea lions across their South Australian range. Synthesizing telemetric, oceanographic and isotopic datasets collected from seven individuals enabled us to characterise individual foraging behaviour over an approximate two year time period. Data suggested seasonal variability in stable carbon and nitrogen isotopes that could not be otherwise explained by changes in animal movement patterns. Similarly, animals did not change their foraging patterns despite fine-scale spatial and temporal variability of the upwelling event. Individual males tended to return to the same colony at which they were tagged and utilized the same at-sea regions for foraging irrespective of oceanographic conditions or time of year. Our study contrasts current general assumptions that male otariid life history strategies should result in greater dispersal, with adult male Australian sea lions displaying central place foraging behaviour similar to males of other otariid species in the region.

Advances in molecular techniques have enabled the study of genetic diversity and population structure in many different contexts. Studies that assess the genetic structure of cetacean populations often use biopsy samples from free-ranging individuals and tissue samples from stranded animals or individuals that became entangled in fishery or aquaculture equipment. This leads to the question of how representative the location of a stranded or entangled animal is with respect to its natural range, and whether similar results would be obtained when comparing carcass samples with samples from free-ranging individuals in studies of population structure. Here we use tissue samples from carcasses of dolphins that stranded or died as a result of bycatch in South Australia to investigate spatial population structure in two species: coastal bottlenose (Tursiops sp.) and short-beaked common dolphins (Delphinus delphis). We compare these results with those previously obtained from biopsy sampled free-ranging dolphins in the same area to test whether carcass samples yield similar patterns of genetic variability and population structure. Data from dolphin carcasses were gathered using seven microsatellite markers and a fragment of the mitochondrial DNA control region. Analyses based on carcass samples alone failed to detect genetic structure in Tursiops sp., a species previously shown to exhibit restricted dispersal and moderate genetic differentiation across a small spatial scale in this region. However, genetic structure was correctly inferred in D. delphis, a species previously shown to have reduced genetic structure over a similar geographic area. We propose that in the absence of corroborating data, and when population structure is assessed over relatively small spatial scales, the sole use of carcasses may lead to an underestimate of genetic differentiation. This can lead to a failure in identifying management units for conservation. Therefore, this risk should be carefully assessed when planning population genetic studies of cetaceans.

BACKGROUND: In many gregarious mammals, mothers and offspring have developed the abilities to recognise each other using acoustic signals. Such capacity may develop at different rates after birth/parturition, varying between species and between the participants, i.e., mothers and young. Differences in selective pressures between species, and between mothers and offspring, are likely to drive the timing of the onset of mother-young recognition. We tested the ability of Australian sea lion mothers to identify their offspring by vocalisation, and examined the onset of this behaviour in these females. We hypothesise that a rapid onset of recognition may reflect an adaptation to a colonial lifestyle. PRINCIPAL FINDINGS: In a playback study maternal responses to own pup and non-filial vocalisations were compared at 12, 24 and every subsequent 24 hours until the females' first departure post-partum. Mothers showed a clear ability to recognise their pup's voice by 48 hours of age. At 24 hours mothers called more, at 48 hours they called sooner and at 72 hours they looked sooner in response to their own pup's vocalisations compared to those of non-filial pups. CONCLUSIONS: We demonstrate that Australian sea lion females can vocally identify offspring within two days of birth and before mothers leave to forage post-partum. We suggest that this rapid onset is a result of selection pressures imposed by a colonial lifestyle and may be seen in other colonial vertebrates. This is the first demonstration of the timing of the onset of maternal vocal recognition in a pinniped species.