Prey Capture Events Research Articles

Identifying prey capture events of a free-ranging marine predator using bio-logger data and deep learning.

Marine predators are integral to the functioning of marine ecosystems, and their consumption requirements should be integrated into ecosystem-based management policies. However, estimating prey consumption in diving marine predators requires innovative methods as predator-prey interactions are rarely observable. We developed a novel method, validated by animal-borne video, that uses tri-axial acceleration and depth data to quantify prey capture rates in chinstrap penguins (Pygoscelis antarctica). These penguins are important consumers of Antarctic krill (Euphausia superba), a commercially harvested crustacean central to the Southern Ocean food web. We collected a large data set (n = 41 individuals) comprising overlapping video, accelerometer and depth data from foraging penguins. Prey captures were manually identified in videos, and those observations were used in supervised training of two deep learning neural networks (convolutional neural network (CNN) and V-Net). Although the CNN and V-Net architectures and input data pipelines differed, both trained models were able to predict prey captures from new acceleration and depth data (linear regression slope of predictions against video-observed prey captures = 1.13; R 2 ≈ 0.86). Our results illustrate that deep learning algorithms offer a means to process the large quantities of data generated by contemporary bio-logging sensors to robustly estimate prey capture events in diving marine predators.

Hunting tactics of southern elephant seals Mirounga leonina and anti-predatory behaviours of their prey

For air-breathing marine predators that must save energy during dives, the ability to adopt hunting tactics that minimise the risk of triggering an escape reaction from the prey is crucial for efficient foraging. Female southern elephant seals (SES) Mirounga leonina forage on small mesopelagic prey that they must hunt almost continuously to meet their high-energy requirements. Here we aimed at understanding how these large time-limited deep-divers can efficiently exploit their small-sized prey. To do so, we used data recorded by a new biologger, the sonar tag, deployed on SES during their post-breeding foraging trips. This tag combines an active acoustic sensor with ultra-high-resolution movement and bioluminescence sensors. This combination of sensors offers a unique opportunity to simultaneously describe SES hunting tactics and defence mechanisms of their prey. We analysed more than 5800 prey capture events in 9 SES and show that they adopt a ‘stalking’ hunting behaviour, allowing them to get as close as possible to their prey before attacking. The ability of SES to adopt stealth approach tactics, minimising the risk of initiating a flight reaction from their prey, might be a key factor in the success of this far-ranging generalist predator.

Sealing the deal - Antarctic fur seals' active hunting tactics to capture small evasive prey revealed by miniature sonar tags.

The ability of predators to adopt hunting tactics that minimise escape reactions from prey is crucial for efficient foraging, and depends on detection capabilities and locomotor performance of both predators and prey. Here, we investigated the efficiency of a small pinniped, the Antarctic fur seal (Arctocephalus gazella) at exploiting their small prey by describing for the first time their fine-scale predator-prey interactions. We compared these with those from another diving predator, the southern elephant seal (Mirounga leonina) that forage on the same prey type. We used data recorded by a newly developed sonar tag that combines active acoustics with ultrahigh-resolution movement sensors to study simultaneously the fine-scale behaviour of both Antarctic fur seals and prey during predator-prey interactions in more than 1200 prey capture events for eight female Antarctic fur seals. Our results showed that Antarctic fur seals and their prey detect each other at the same time, i.e. 1-2 s before the strike, forcing Antarctic fur seals to display reactive fast-moving chases to capture their prey. In contrast, southern elephant seals detect their prey up to 10 s before the strike, allowing them to approach their prey stealthily without triggering an escape reaction. The active hunting tactics used by Antarctic fur seals is probably very energy consuming compared with the stalking tactics used by southern elephant seals but might be compensated for by the consumption of faster-moving larger prey. We suggest that differences in manoeuvrability, locomotor performance and detection capacities and in pace of life between Antarctic fur seals and southern elephant seals might explain these differences in hunting styles.

Safe Passage or Hunting Ground? A Test of the Prey-Trap Hypothesis at Wildlife Crossing Structures on NH 44, Pench Tiger Reserve, Maharashtra, India

Crossing structures are widely accepted mitigation measures used to offset the impacts of roads in ecologically sensitive areas that serve as important animal corridors. However, altered interspecies interactions at crossing structures may reduce the potency of these structures for some species and groups. Anecdotes of predation events at crossing structures have necessitated the assessment of predator–prey interactions at crossing structures. We investigated the ‘prey-trap’ hypothesis at nine crossing structures on a highway in central India adjacent to a tiger reserve by comparing the geometric mean latencies between successive prey, predator and free-ranging dog camera trap capture events at the crossing structures. Among all interactions, prey–predator latencies were the shortest, and significantly lower than prey–dog and predator–prey latencies. Prey–predator sequences involving wild dogs had the shortest average latencies (65.6 ± 9.7 min). Prey–predator latencies decreased with increasing crossing structure width; however, these crossing structures are also the sites that are most frequently used by wildlife. Results indicate that the crossing structures presently do not act as ‘prey-traps’ from wild predators or free-ranging dogs. However, measures used to alleviate such prospects, such as heterogeneity in structure design and increase in vegetation cover near crossing structures, are recommended.

Feeding characteristics and prey profitability in five herring‐feeding killer whales (Orcinus orca) in northern Norway

AbstractThis study used multisensory data tags with integrated 4K video to investigate feeding behavior and prey profitability in five adult male killer whales (Orcinus orca) at herring wintering grounds in Norway, in 2016 and 2017. Video recorded two killer whales engaged in carousel feeding, and two engaged in seiner feeding (i.e., feeding on herring discards around purse seiner vessels). The feeding behaviors identified from the video data allowed for determination of associated kinematic signatures, which were used to further identify and characterize carousel feeding and deep feeding dives over the entire logger duration. Prey consumption during on camera feeding bouts was also measured to calculate profitability of feeding bouts for the different behaviors. Average number of prey consumed per minute was 1.08 ± 0.43 for carousel feeding and 0.43 ± 0.07 for seiner feeding (n = 122 prey capture events). Using kinematic data, a total of 18 carousel feeding bouts and 206 deep feeding dives were identified. Whales spent at least 37%–65% of time over 24 hr feeding. Using field metabolic rate estimates from the literature and the energetic content of herring caught locally, killer whales required an estimated 285–578 herring/day to balance daily energy requirements.

Effect of interactive threats on foraging effort by endangered killer whales (Orcinus orca)

Killer whales face many anthropogenic threats including vessel traffic, noise, and reduced prey availability. Here, we use high-resolution suction-cup Dtags to study the behavior of endangered Southern Resident killer whales that rely on biosonar to hunt salmon, and investigate how proximate vessels affect foraging behavior. From tag data, we identified subsurface behavior, including foraging and prey capture events. We then tested several vessel and associated sound, demographic, and environmental variables on behavioral state occurrence, time spent within each state, and foraging effort involving prey capture. Whales made fewer prey capture dives and spent less time in these dives when vessels had an average distance <400 yard (366 m). Lower prey abundance and higher vessel speed reduced prey capture probability, documenting the interplay between these effects. Finally, whales dove to depth more slowly while increasing dive duration to capture prey in the presence of vessel-emitted sonar, but descended more quickly with higher noise levels and closer vessels. Current efforts investigating foraging behavior and noise exposure over the diel cycle aim to better quantify foraging rates and activity budgets in this endangered population. These findings advance awareness of vessels and noise consequences on killer whales to inform conservation and management actions.

Sound and movement data from two overlapping populations of killer whales reveal noise effects on foraging behavior

Understanding ambient noise effects on foraging ecology of endangered killer whales is a necessary step toward predicting population-level consequences of acoustic disturbance. However, this has been limited by the difficulty of identifying prey capture which typically occurs out of sight, and by the challenge of obtaining a sufficient range of noise conditions. We addressed these problems using sound and movement data from suction cup-attached archival tags deployed on 52 Northern resident and endangered Southern resident killer whales sampled during overlapping field efforts. We measured broadband ambient noise levels above low-frequency cut-offs that minimized flow noise, and quantified foraging behavior using established acoustic and movement signatures to detect prey capture events. Statistical models revealed significant effects of noise level on searching, pursuit and capture outcomes, including the likelihood that a prey pursuit dive was aborted. We discuss the utility of broadening the range of noise level conditions available by sampling from multiple populations within overlapping geographic regions, and highlight the implications of noise-induced foraging interference on two killer whale populations with different population trajectories. Finally, we suggest future applications of this comparative approach to foster a greater understanding of the population consequences of acoustic disturbance in killer whales.

Accelerometry predicts prey-capture rates in the deep-diving king penguin Aptenodytes patagonicus

Remotely estimating prey-capture rates in wild animals is key to assess foraging success. In diving animals, accelerometers have been particularly useful to remotely detect prey captures and have been shown to be more precise than traditional estimates relying on depth-derived measures (e.g., wiggles). However, validations of the accelerometry technique using a gold standard (i.e., with supervision) have been mostly restricted to shallow diving species, which can be equipped with camera-loggers for visual validation of prey-capture events. In species diving near the euphotic limit (150–200 m), accelerometers remain mostly untested due to the difficulty of validating such methods in darkness at extreme depth in the wild. In addition, prey-pursuits in low-light conditions might not result in intense and long-duration acceleration signatures, as predator–prey perception likely occurs at close-range in the dark (i.e., the “visual-interactions hypothesis”). We combined accelerometers with beak-opening sensors (for validation) and depth recorders on a wild deep-diving seabird, the king penguin Aptenodytes patagonicus, to describe prey captures at depth and create predictive models using accelerometers. Surprisingly, prey pursuits and captures were similar in duration (3.9 ± 3.5 s) and intensity (0.78 ± 0.31 g) as shallow-diving species reported by similar studies. As accelerometry signatures were distinct, accelerometry-derived variables were almost twice as accurate (Mean-squared error = 8.6) at predicting prey-capture events as depth-derived variables (“wiggles”, Mean-squared error = 16.0). As in the shallow-diving species, accelerometry outperforms traditional depth-derived models at measuring the foraging intake in deep-diving animals, highlighting the usefulness of accelerometers for measuring animal behavior.

Fine-scale body and head movements allow to determine prey capture events in the Magellanic Penguin (Spheniscus magellanicus)

The identification of when, how and where animals feed is essential to estimate the amount of energy they obtain and to study the processes associated with prey search and consumption. We combined the use of animal-borne video cameras and accelerometers to characterise the body and head movements associated to four types of prey capture behaviours in the Magellanic Penguin (Spheniscus magellanicus). In addition, we evaluated how the K-Nearest Neighbour (K-NN) algorithm recognized these behaviours from acceleration data. Finally, we compared the total capture and the capture per unit time (CPUT) derived by identifying prey capture events using the K-NN algorithm to that derived by counting undulations in the dive profile (“wiggles”). During captures, body and head movements were highly variable in the tridimensional space. Energy expenditure (i.e., VeDBA values) during diving periods with prey captures was from three to four times higher than during controls diving periods (i.e., with no capture events). The K-NN classification resulted effective and showed accuracy scores above 90% when considering both head and body related features. In addition, when captures were estimated using the K-NN method, the CPUT was similar or higher to that estimated by counting wiggles. Our study contributes to the knowledge of the trophic ecology of this species and provides an alternative method for estimating prey consumption in the Magellanic Penguin and other diving seabirds.

Foraging behavior and association with mixed flocks by the Critically Endangered Alagoas Tyrannulet Phylloscartes ceciliae (Aves: Passeriformes: Tyrannidae)

The Alagoas Tyrannulet Phylloscartes ceciliae is a Critically Endangered species with a restricted distribution to the Atlantic Forest of the Pernambuco Endemism Center. It seems to specialize in using the sally-strike maneuver (68.85%) to catch prey on green foliage (50.82%) and in the air (34.42%). We calculated a catch rate of 2.93 catches/minute, identified a new food item for this species (caterpillars), and included 12 species in the list of birds recorded in mixed flocks with the species. Our results show that majority of the prey capture events by Alagoas Tyrannulet occurred inside tree canopies, thus we can infer that this species needs a more advanced stage of forest sucession, with higher trees, emphasizing the urgency for restoration programs in the region.

Diving Behavior of Mirounga leonina: A Functional Data Analysis Approach

The diving behavior of southern elephant seals, Mirounga leonina, is investigated through the analysis of time-depth dive profiles. The originality of this work is to consider dive profiles as continuous curves. For this purpose, a Functional Data Analysis (FDA) approach is proposed for the shape analysis of a collection of dive profiles. Complexity of dive shapes is characterized by a mixture of three main shape variations accounting for about $80\%$ of the entire variability: U or V shape, vertical depth variability during the bottom time, and skewed left or right. Model-based clustering allows the identification of eight dive shape clusters in a quick and automated way. Connection between shape patterns and classical descriptors, as well as the number of prey capture events, is achieved, showing that the clusters are coherent to specific foraging behaviors previously identified in the literature labeled as drift, exploratory and active dives. Finally, FDA is compared to classical methods relying on the computation of discrete dive descriptors. Results show that taking the shape of the dive as a whole is more resilient to corrupted or incomplete sampled data. FDA is, therefore, an efficient tool adapted for processing and comparing dive data with different sampling frequencies and for improving the quality and the accuracy of transmitted data.

Acceleration-triggered animal-borne videos show a dominance of fish in the diet of female northern elephant seals.

Knowledge of the diet of marine mammals is fundamental to understanding their role in marine ecosystems and response to environmental change. Recently, animal-borne video cameras have revealed the diet of marine mammals that make short foraging trips. However, novel approaches that allocate video time to target prey capture events is required to obtain diet information for species that make long foraging trips over great distances. We combined satellite telemetry and depth recorders with newly developed date-/time-, depth- and acceleration-triggered animal-borne video cameras to examine the diet of female northern elephant seals during their foraging migrations across the eastern North Pacific. We obtained 48.2 h of underwater video, from cameras mounted on the head (n=12) and jaw (n=3) of seals. Fish dominated the diet (78% of 697 prey items recorded) across all foraging locations (range: 37-55°N, 122-152°W), diving depths (range: 238-1167 m) and water temperatures (range: 3.2-7.4°C), while squid comprised only 7% of the diet. Identified prey included fish such as myctophids, Merluccius sp. and Icosteus aenigmaticus, and squid such as Histioteuthis sp., Octopoteuthis sp. and Taningia danae Our results corroborate fatty acid analysis, which also found that fish are more important in the diet, and are in contrast to stomach content analyses that found cephalopods to be the most important component of the diet. Our work shows that in situ video observation is a useful method for studying the at-sea diet of long-ranging marine predators.

Repeatability of between-group differences in collective foraging is shaped by group composition in social spiders

Stable between-group variation in collective behavior has been observed in a variety of taxa. We examine here whether climate of origin (arid/wet), colony personality composition (shy/bold/mixed), and group size determine the repeatability of collective foraging behavior in Stegodyphus dumicola Pocock, 1898 (Eresidae). Experimental colonies were created with contrasting ratios of bold/shy group members and run through 20 simulated prey capture events in a greenhouse. We found that (i) larger colonies and colonies composed of bold spiders were more repeatable in how many attackers they deployed to prey stimuli, (ii) colonies composed of shy spiders were more repeatable in their latency of attack, and (iii) climate of origin had no effect on the repeatability of colony behavior. Colony bold/shy composition had no effect on within-group variation in foraging behavior. Thus, differences in repeatability were the result of increases in between-group differences in foraging behavior, and not shifts in the behavioral flexibility of individual colonies. These results indicate that changes to colony composition and group size can alter the extent to which colonies exhibit characteristic behavioral differences.

Selective consumption of sacoglossan sea slugs (Mollusca: Gastropoda) by scleractinian corals (Cnidaria: Anthozoa).

Recent studies revealed that reef corals can eat large-sized pelagic and benthic animals in addition to small planktonic prey. As follow-up, we document natural ingestion of sea slugs by corals and investigate the role of sacoglossan sea slugs as possible prey items of scleractinian corals. Feeding trials were carried out using six sacoglossan species as prey, two each from the genera Costasiella, Elysia and Plakobranchus, and four free-living solitary corals (Danafungia scruposa, Fungia fungites, Pleuractis paumotensis and Heteropsammia cochlea) as predators. Trials were carried out under both in-situ and ex-situ conditions with the aim to observe ingestion and assess signs of prey consumption based on tissue loss of prey individuals over time. Significant differences were observed in both ingestion time and consumption state of prey between prey species, with three of them being ingested more rapidly and preferentially consumed over the others. Additionally, prey size was found to be a significant factor with larger prey (>12 mm) being ingested more slowly and rarely than smaller ones (<6 mm and 6–12 mm). Comparisons of consumption capability among predators showed no significant difference with all coral species showing similar preferences for prey species. While no specific mechanism of prey capture is proposed, we also document instances of kleptoparisitism and resuspension of prey items by wrasses. This study highlights the important distinction between opportunistic prey capture and true predation events.

Behavioral responses vary with prey species in the social spider, Stegodyphus sarasinorum

Abstract Predators living in social groups often show consistent interindividual differences in prey capture behavior that may be linked to personality. Though personality predisposes individuals for certain behaviors, responses can also be influenced by context. Studies examining personality-dependent participation in prey capture have largely employed only one prey species, offering the predator no choice. In nature, predators encounter a range of prey species, therefore participation in or leading a prey capture event must also depend on prey attributes (e.g., size and risk). In the social spider Stegodyphus sarasinorum, collective prey capture is mediated by personality types as a consequence of which some individuals are consistently more likely to attack. Here, we examined if an individual’s consistency to attack persisted within and between the 2 prey species (honeybees and grasshoppers) and if the same individuals attacked first with both prey species. Our results showed that interindividual differences in attacking persisted within and between the 2 prey species. Spiders showed greater participation in attacking grasshoppers relative to bees. Identities of the first attackers were not the same for bees and grasshoppers. Spiders showed greater consistency over time in attacking bees relative to grasshoppers. Bees attracted fewer attackers than size-matched grasshoppers. These results suggest that greater task specialization may be necessary to successfully subdue bees. Spiders handled bees more cautiously, which is likely to explain the observed plasticity in attacking the 2 prey species. Thus, participation in prey capture in social spiders is influenced by the attributes of prey species.

Hydrodynamics of frontal striking in aquatic snakes: drag, added mass, and the possible consequences for prey capture success

Transient locomotion under water is highly constrained by drag and added mass, yet some aquatic snakes catch their prey using a fast forward acceleration, with the mouth opened. These aquatic snakes show a convergence of their head shape in comparison with closely related species that do not forage under water. As both drag and added mass are related to some extent to the shape of the moving object, we explored how shape impacts the hydrodynamic forces applied to the head of a snake during a prey capture event. We compared two 3D-printed heads representing typical shapes of aquatically-foraging and non-aquatically-foraging snakes, and frontal strike kinematics based on in vivo observations. By using direct force measurements, we calculated the drag and added mass coefficient of the two models. Our results show that both drag and added mass are reduced in aquatic snakes. The drag coefficient of the aquatic model is 0.24, which is almost two times smaller than the non-aquatic model. The added mass coefficient of the aquatic model is 0.15 versus 0.24 for the non-aquatic model, showing that the convergence of head shape in aquatically foraging snakes is associated with a hydrodynamic advantage during frontal striking. The vorticity field measurements with particle image velocimetry show that a less intense recirculation bubble behind the jaw of the aquatic model, compared to the non-aquatic model, might be the basis of this advantage.

Kinematic signatures of prey capture from archival tags reveal sex differences in killer whale foraging activity.

Studies of odontocete foraging ecology have been limited by the challenges of observing prey capture events and outcomes underwater. We sought to determine whether subsurface movement behavior recorded from archival tags could accurately identify foraging events by fish-eating killer whales. We used multisensor bio-logging tags attached by suction cups to Southern Resident killer whales (Orcinus orca) to: (1) identify a stereotyped movement signature that co-occurred with visually confirmed prey capture dives; (2) construct a prey capture dive detector and validate it against acoustically confirmed prey capture dives; and (3) demonstrate the utility of the detector by testing hypotheses about foraging ecology. Predation events were significantly predicted by peaks in the rate of change of acceleration ('jerk peak'), roll angle and heading variance. Detection of prey capture dives by movement signatures enabled substantially more dives to be included in subsequent analyses compared with previous surface or acoustic detection methods. Males made significantly more prey capture dives than females and more dives to the depth of their preferred prey, Chinook salmon. Additionally, only half of the tag deployments on females (5 out of 10) included a prey capture dive, whereas all tag deployments on males exhibited at least one prey capture dive (12 out of 12). This dual approach of kinematic detection of prey capture coupled with hypothesis testing can be applied across odontocetes and other marine predators to investigate the impacts of social, environmental and anthropogenic factors on foraging ecology.

Comparative Prey Spectra Analyses on the Endangered Aquatic Carnivorous Waterwheel Plant (Aldrovanda vesiculosa, Droseraceae) at Several Naturalized Microsites in the Czech Republic and Germany.

SynopsisThe critically endangered carnivorous waterwheel plant (Aldrovanda vesiculosa, Droseraceae) possesses underwater snap traps for capturing small aquatic animals, but knowledge on the exact prey species is limited. Such information would be essential for continuing ecological research, drawing conclusions regarding trapping efficiency and trap evolution, and eventually, for conservation. Therefore, we performed comparative trap size measurements and snapshot prey analyses at seven Czech and one German naturalized microsites on plants originating from at least two different populations. One Czech site was sampled twice during 2017. We recorded seven main prey taxonomic groups, that is, Cladocera, Copepoda, Ostracoda, Ephemeroptera, Nematocera, Hydrachnidia, and Pulmonata. In total, we recorded 43 different prey taxa in 445 prey-filled traps, containing in sum 461 prey items. With one exception, prey spectra did not correlate with site conditions (e.g. water depth) or trap size. Our data indicate that A. vesiculosa shows no prey specificity but catches opportunistically, independent of prey species, prey mobility mode (swimming or substrate-bound), and speed of movement. Even in cases where the prey size exceeded trap size, successful capture was accomplished by clamping the animal between the traps’ lobes. As we found a wide prey range that was attracted, it appears unlikely that the capture is enhanced by specialized chemical- or mimicry-based attraction mechanisms. However, for animals seeking shelter, a place to rest, or a substrate to graze on, A. vesiculosa may indirectly attract prey organisms in the vicinity, whereas other prey capture events (like that of comparably large notonectids) may also be purely coincidental.

Search and foraging behaviors from movement data: A comparison of methods.

Search behavior is often used as a proxy for foraging effort within studies of animal movement, despite it being only one part of the foraging process, which also includes prey capture. While methods for validating prey capture exist, many studies rely solely on behavioral annotation of animal movement data to identify search and infer prey capture attempts. However, the degree to which search correlates with prey capture is largely untested. This study applied seven behavioral annotation methods to identify search behavior from GPS tracks of northern gannets (Morus bassanus), and compared outputs to the occurrence of dives recorded by simultaneously deployed time–depth recorders. We tested how behavioral annotation methods vary in their ability to identify search behavior leading to dive events. There was considerable variation in the number of dives occurring within search areas across methods. Hidden Markov models proved to be the most successful, with 81% of all dives occurring within areas identified as search. k‐Means clustering and first passage time had the highest rates of dives occurring outside identified search behavior. First passage time and hidden Markov models had the lowest rates of false positives, identifying fewer search areas with no dives. All behavioral annotation methods had advantages and drawbacks in terms of the complexity of analysis and ability to reflect prey capture events while minimizing the number of false positives and false negatives. We used these results, with consideration of analytical difficulty, to provide advice on the most appropriate methods for use where prey capture behavior is not available. This study highlights a need to critically assess and carefully choose a behavioral annotation method suitable for the research question being addressed, or resulting species management frameworks established.

A Comparative Analysis of SVM and IDNN for Identifying Penguin Activities

ABSTRACTWhere, when and how much animals eat provide valuable insights into their ecology. In this paper, we present a comparative analysis between Support Vector Machine (SVM) and Input Delay Neural Network (IDNN) models to identify prey capture events from penguin accelerometry data. A pre-classified dataset of 3D time-series data from back-mounted accelerometers was used. We trained both the models to classify the penguins’ behavior at intervals as either ‘prey handling’ or ‘swimming’. The aim was to determine whether IDNN could achieve the same level of classification accuracy as SVM, but with reduced memory demands. This would enable the IDNN model to be embedded on the accelerometer micro-system itself, and hence reduce the magnitude of the output data to be uploaded. Based on the classification results, this paper provides an analysis of the two models from both an accuracy and applicability point of view. The experimental results show that both models achieve an equivalent accuracy of approx. 85% using the featured data, with a memory demand of 0.5 kB for IDNN and 0.7 Mb for SVM. The raw accelerometer data let us improve the generalizability of the models with a slightly lower accuracy to around 80%. This indicates that the IDNN model can embed on the accelerometer itself, reducing problems associated with raw time-series data retrieval and loss.