Measurement error causes scale-dependent threshold erosion of biological signals in animal movement data.

Recent advances in telemetry technology have created a wealth of tracking data available for many animal species moving over spatial scales from tens of meters to tens of thousands of kilometers. Increasingly, such data sets are being used for quantitative movement analyses aimed at extracting fundamental biological signals such as optimal searching behavior and scale-dependent foraging decisions. We show here that the location error inherent in various tracking technologies reduces the ability to detect patterns of behavior within movements. Our analyses endeavored to set out a series of initial ground rules for ecologists to help ensure that sampling noise is not misinterpreted as a real biological signal. We simulated animal movement tracks using specialized random walks known as Lévy flights at three spatial scales of investigation: 100-km, 10-km, and 1-km maximum daily step lengths. The locations generated in the simulations were then blurred using known error distributions associated with commonly applied tracking methods: the Global Positioning System (GPS), Argos polar-orbiting satellites, and light-level geolocation. Deviations from the idealized Lévy flight pattern were assessed for each track after incrementing levels of location error were applied at each spatial scale, with additional assessments of the effect of error on scale-dependent movement patterns measured using fractal mean dimension and first-passage time (FPT) analyses. The accuracy of parameter estimation (Lévy mu, fractal mean D, and variance in FPT) declined precipitously at threshold errors relative to each spatial scale. At 100-km maximum daily step lengths, error standard deviations of > or = 10 km seriously eroded the biological patterns evident in the simulated tracks, with analogous thresholds at the 10-km and 1-km scales (error SD > or = 1.3 km and 0.07 km, respectively). Temporal subsampling of the simulated tracks maintained some elements of the biological signals depending on error level and spatial scale. Failure to account for large errors relative to the scale of movement can produce substantial biases in the interpretation of movement patterns. This study provides researchers with a framework for understanding the limitations of their data and identifies how temporal subsampling can help to reduce the influence of spatial error on their conclusions.     

Spatial and temporal patterns of habitat use by Eurasian oystercatchers (<Emphasis Type="Italic">Haematopus ostralegus</Emphasis>) in the eastern Wadden Sea revealed using GPS data loggers

Detailed knowledge on species–habitat relationships is of crucial importance for the understanding of processes in marine ecosystems. Being top-predators, birds are important bio-indicators for marine systems. The aim of this study was to elucidate precise information on foraging habitat use and foraging times of oystercatchers (Haematopus ostralegus) on wide tidal flats using global positioning system (GPS) data loggers. The study was conducted to collect hints for the negative population trends in oystercatchers in the Wadden Sea. It is the first time that GPS technique has been used in a shorebird species. Although oystercatchers are known to exhibit foraging site fidelity, a number of individuals visited multiple sites. Foraging trips at night were longer, and the targeted sites were further away than those used during the day. These patterns were likely to be caused by higher risks of clutch predation by avian predators during the day that led adults to reduce their absence to defend their clutches. Our methodological approach enabled the subtle spatio-temporal patterns of habitat use to be determined on a very fine spatio-temporal scale. We suggest further potential studies using GPS data loggers that may help to reveal the reasons for the current declines in oystercatcher populations in the German Wadden Sea.     

Highly directional swimming by scalloped hammerhead sharks,Sphyrna lewini,and subsurface irradiance,temperature, bathymetry,and geomagnetic field

The homing behavior of scalloped hammerhead sharks (Sphyrna lewini) to and fro between Espiritu Santo Seamount and Las Animas Island and the surrounding pelagic environment was studied to reveal their mechanisms of navigation in the oceanic environment. Four sharks were tagged with ultrasonic transmitters and tracked at the former location and one shark at the latter site during July, August, or September between 1981 and 1988. Hammerhead swimming movements were highly oriented: the mean coefficient of concentration (r) for sets of ten consecutive swimming directions recorded during eight homing movements by three hammerhead sharks ranged from 0.885 to 0.996. Drift within a current could not explain this directionality, since highly variable directions were recorded from a transmitter floating at the sea surface after becoming detached from a shark. Forward swimming momentum was an unlikely explanation, since highly directional swimming was maintained for a period of 32 min with only a gradual change in course. To maintain directionality over this period, an environmental property should be necessary for guidance. The hammerheads swam at night, with repeated vertical excursions ranging from 100 to 450 m deep, out of view of either the sea surface or the sea floor. The sharks' vertical diving movements were compared to distributions of spectral irradiance (relative to elasmobranch scotopic and photopic visual sensitivities), temperature, and current-flow directions in the water column. No relationships were evident between these properties and the sharks' oriented swimming movements. Movements of scalloped hammerhead sharks to and from a seamount were compared to topographic features in bathymetry and geomagnetic field leading away from the seamount. Sharks swam repeatedly over fixed geographic paths, and these paths occurred less often along submarine ridges and valleys than maxima and minima in the geomagnetic field. No significant difference existed between the degree of association of points from the sharks' tracks and points from track simulations and 20° changes in the slope of the depth record. On the other hand, significantly more points from the sharks' tracks were associated with slope changes in the magnetic intensity record than points from track simulations. A magnetic intensity gradient of 0.037 nanoteslas/m (nT/m) existed at 175 m depth, where a shark swam directionally, and this gradient was three times steeper than that measured at the sea surface and exceeded that recorded at a depth of 200 m. The hammerheads are hypothesized to find the seamount using geomagnetic topotaxis. The shark could be attracted to and move back and forth along ridges and valleys, features in the relief of magnetic field intensities occurring over a geographical area.     

Home range of immature green turtles tracked at an offshore tropical reef using automated passive acoustic technology

Seventeen immature green turtles Chelonia mydas were tracked concurrently by automated ultrasonic receivers at a coral reef off North-Eastern Australia (September–December 2010, 16.4°S, 145.6°E). The majority (n = 11) were tracked for the entire 100-day study, the remainder for 23–85 days. Detection data aggregated at 30-min intervals produced median 6.5–35 daily locations for individual turtles. Home range areas (95 % utilisation distribution) were ≤1 km², $ {\bar{\text{x}}} $  ± SD = 0.74 km² ± 0.159. To the best of our knowledge, these are the first home range estimates for C. mydas foraging at offshore tropical reefs. The findings are important for conservation in revealing near-continuous presence of the same individuals within a small geographic area. Time between detections was very short (median <3 min) demonstrating passive ultrasonic technology can track multiple turtles in a foraging environment with higher temporal resolution than typically achieved by satellite tracking.     

Factors affecting the foraging behaviour of the European shag: implications for seabird tracking studies

Seabird tracking has become an ever more popular tool to aid environmental procedures such as the designation of marine protected areas and environmental impact assessments. However, samples used are usually small and little consideration is given to experimental design and sampling protocol. European shags Phalacrocorax aristotelis were tracked using GPS technology over three breeding seasons and the following foraging trip characteristics: trip duration, trip distance, maximum distance travelled from the colony, size of area used and direction travelled from colony were determined for each foraging trip. The effect of sex, year of study, breeding site, number and age of chicks and the timing of tracking on foraging behaviour were investigated using a General Estimation Equation model. A range of sampling scenarios reflecting likely field sampling were also tested to compare how foraging behaviour differed depending on composition of the sample of birds tracked. Trip distance, trip duration, maximum distance travelled and size of area used were all significantly affected by the breeding site, and the number of chicks a tracked adult was raising. The effect of sex was also seen when examining trip distance, trip duration and the maximum distance travelled. The direction travelled on a foraging trip was also significantly affected by breeding site. This study highlights the importance of sampling regime and the influence that year, sex, age, number of chicks and breeding site can have on the foraging trip characteristics for this coastal feeding seabird. Given the logistical and financial constraints in tracking large numbers of individuals, this study identifies the need for researchers to consider the composition of their study sample to ensure any identified foraging areas are as representative as possible of the whole colony’s foraging area.     

On the integration of individual foraging strategies with colony ergonomics in social insects: nectar-collection in honeybees

Summary We experimentally tested whether foraging strategies of nectar-collecting workers of the honeybee (Apis mellifera) vary with colony state. In particular, we tested the prediction that bees from small, fast growing colonies should adopt higher workloads than those from large, mature colonies. Queenright small colonies were set up by assembling 10 000 worker bees with approximately 4100 brood cells. Queenright large colonies contained 35 000 bees and some 14 500 brood cells. Thus, treatments differed in colony size but not in worker/brood ratios. Differences in workload were tested in the context of single foraging cycles. Individuals could forage on a patch of artificial flowers offering given quantities and qualities of nectar rewards. Workers of small colonies took significantly less nectar in an average foraging excursion (small: 40.1 ± 1.1 SE flowers; large: 44.8 ± 1.1), but spent significantly more time handling a flower (small: 7.3 ± 0.4 s ; large: 5.8 ± 0.4 s). When the energy budgets for an average foraging trip were calculated, individuals from all colonies showed a behavior close to maximization of net energetic efficiency (i.e., the ratio of net energetic gains to energetic costs). However, bees from small colonies, while incurring only marginally smaller costs, gained less net energy per foraging trip than those from large colonies, primarily as a result of prolonged handling times. The differences between treatments were largest during the initial phases of the experimental period when also colony development was maximally different. Our results are at variance with simple models that assume natural selection to have shaped behavior in a single foraging trip only so as to maximize colony growth. Offprint requests to: P. Schmid-Hempel     

Vibrational communication between hitchhikers and foragers in leaf-cutting ants (Atta cephalotes)

Abstract In a foraging column of the leaf-cutting ant Atta cephalotes, minim workers (the smallest worker subcaste) hitchhike on leaf fragments carried by larger workers. It has been demonstrated that they defend leaf carriers against parasitic phorid flies. The present study examines the cues used by the potential hitchhikers to locate leaf carriers. As recently reported, foraging workers stridulate while cutting a leaf fragment, and the stridulatory vibrations serve as closerange recruitment signals. We tested the hypothesis that these plant-borne stridulatory vibrations are used by the potential hitchhikers to locate workers engaged in cutting. Three different lines of evidence support this view. Firstly, the repetition rate of the stridulations produced by foraging workers increases significantly as foragers maneuver the leaf fragment into the carrying position and walk loaded to the nest. This is the moment when hitchhikers usually climb on the leaf. Although the leaf-borne stridulatory vibrations are considerably attenuated when transmitted through the workers' legs, they can nevertheless be detected at short distances by minims. This subcaste is several times more sensitive to substrate-borne vibrations than larger workers. Secondly, when a stridulating and a silent leaf were simultaneously presented at the foraging site, minim workers spent significantly more time on the stridulating than on the silent leaf. Thirdly, hitchhiking was more frequent in leaf carriers which cut fragments out of the stridulating leaf than in those cutting the silent leaf.Abstract In a foraging column of the leaf-cutting ant Atta cephalotes, minim workers (the smallest worker subcaste) hitchhike on leaf fragments carried by larger workers. It has been demonstrated that they defend leaf carriers against parasitic phorid flies. The present study examines the cues used by the potential hitchhikers to locate leaf carriers. As recently reported, foraging workers stridulate while cutting a leaf fragment, and the stridulatory vibrations serve as closerange recruitment signals. We tested the hypothesis that these plant-borne stridulatory vibrations are used by the potential hitchhikers to locate workers engaged in cutting. Three different lines of evidence support this view. Firstly, the repetition rate of the stridulations produced by foraging workers increases significantly as foragers maneuver the leaf fragment into the carrying position and walk loaded to the nest. This is the moment when hitchhikers usually climb on the leaf. Although the leaf-borne stridulatory vibrations are considerably attenuated when transmitted through the workers' legs, they can nevertheless be detected at short distances by minims. This subcaste is several times more sensitive to substrate-borne vibrations than larger workers. Secondly, when a stridulating and a silent leaf were simultaneously presented at the foraging site, minim workers spent significantly more time on the stridulating than on the silent leaf. Thirdly, hitchhiking was more frequent in leaf carriers which cut fragments out of the stridulating leaf than in those cutting the silent leaf.Communicated by P. Pamilo     

Trunk trails and the searching strategy of a leaf-cutter ant,Atta colombica

Summary Atta colombica uses chemical mass recruitment that allows the rapid exploitation of resources. Most foragers thus search only within patches. Accumulation of extra foragers at patches results in sampling of alternate food items and area-restricted search as patch resources are depleted.Individual workers have a higher probability of removing a leaf fragment the earlier they arrive at a bait. Workers that arrive when much of the resource is gone travel further on the bait (within the patch) but do not spend significantly more time at the patch. They give up after 50–80s.Foraging effort is centered on the extensive trail system, not on the nest a predicted by time and energy foraging models. Search effort is also trail centered. The probability that an item will be discovered decreases with distance from the trail and increasing litter depth. Trail traffic and trail quality together mave no significant effect although this may be because they act antagonistically.Economic considerations predict that trials should be built to high quality and very productive sites. If trails are built as a result of recruitment and recruitment reflects patch quality and productivity, characteristics of forage sites are physically embodied in the trail system.Leaf cutter foraging is better understood as a long term optimization that effectively exploits resources over the lifetime of the colony than as prudent predation that husbands resources.     

GPS and time-depth loggers reveal underwater foraging plasticity in a flying diver,the Cape Cormorant

Knowledge on how divers exploit the water column vertically in relation to water depth is crucial to our understanding of their ecology and to their subsequent conservation. However, information is still lacking for the smaller-bodied species, due mostly to size constraints of data-loggers. Here, we report the diving behaviour of a flying diving seabird, the Cape Cormorant Phalacrocorax capensis, weighing 1.0–1.4 kg. Results were obtained by simultaneously deploying small, high resolution and high sampling frequency GPS and time-depth loggers on birds breeding on islands off Western South Africa (34°S, 18°E) in 2008. In all, dive category was assigned to all dives performed by 29 birds. Pelagic dives occurred almost as frequently as benthic dives. Pelagic dives were shallow (mean: 5 m) and took place over seafloors 5–100 m deep. Benthic dives were deeper, occurring on seafloors mainly 10–30 m deep. Dive shape was linked to dive category in only 60% of dives, while the descent rate, ascent rate and bottom duration/dive duration ratio of a dive best explained its dive category. This shows that only the concomitant use of tracking and depth tags can adequately classify diving strategies in a diver like the Cape Cormorant. Diet was mainly Cape Anchovy Engraulis encrasicolis, suggesting that birds probably displayed two contrasted strategies for capturing the same prey. Flexible foraging techniques represent an important key to survival inside the highly productive but heterogeneous Benguela upwelling ecosystem.     

The choice of two prey types that minimises the probability of starvation

Summary In this paper we investigate the optimal diet of a forager faced with two prey types. Classical optimal foraging theory, based on the maximization of the mean net rate of energetic gain , predicts that the optimal policy is either to take only the more profitable prey type or to take both prey types. The decision between these policies does not depend on the forager's energy reserves or the time available for foraging. We develop two alternative models, based on the minimization of the probability of starvation. In the first model, foraging occurs continuously, and it is optimal to take a prey type if and only if it increases the forager's energy reserves. In the second model foraging stops at dusk, and the forager dies during the night if its reserves at dusk are too low. The optimal policy, which has to be found numerically by dynamic programming, depends on the forager's reserves and the time left till dusk. In general the optimal policy is either to take both types or to take only the more profitable type. Taking both types is optimal when reserves are low, and there is some evidence that this occurs. The models show that factors that have been ignored in classical models may be of importance.     

Does interaction between bumblebees (Bombus ignitus) reduce their foraging area?: bee-removal experiments in a net cage

To examine whether the interaction between bumblebees, Bombus ignitus, reduces their foraging area, we conducted bee-removal experiments in a net cage. In the cage, we set potted Salvia farinacea plants, allowed bumblebees to forage freely on those plants, and followed their plant-to-plant movements to identify a bee with a relatively small foraging area. We then removed all the other foraging bees, except for the bee with a small foraging area, and observed the change of the foraging area of the focal bee under conditions of no interaction with other bees. After the removal of the other bees, all five bees tested enlarged their foraging areas, suggesting that the interaction between bees is an important determinant of their foraging areas. The result also means that bumblebees are able to adjust their foraging areas in response to other foragers, indicating the necessity for future studies to clarify what cues bees use to interact with other bees. Moreover, after the removal treatments, all five bees showed temporary increases in the number of flower probes per plant. This can be explained by their optimal foraging according to the old average intake rate for the plant population and by the delayed changes in response to the new high average energy intake rate after the bee-removal treatments.Communicated by M. Giurfa     

Information transfer and gain in flocks: the effects of quality and quantity of social information at different neighbour distances

We assessed experimentally how the quality and quantity of social information affected foraging decisions of starlings (Sturnus vulgaris) at different neighbour distances, and how individuals gained social information as a function of head position. Our experimental set up comprised three bottomless enclosures, each housing one individual placed on a line at different distances. The birds in the extreme enclosures were labelled senders and the one in the centre receiver. We manipulated the foraging opportunities of senders (enhanced, natural, no-foraging), and recorded the behaviour of the receiver. In the first experiment, receivers responded to the condition of senders. Their searching rate and food intake increased when senders foraged in enhanced conditions, and decreased in no-foraging conditions, in relation to natural conditions. Scanning was oriented more in the direction of conspecifics when senders behaviour departed from normal. In the second experiment, responses were dose dependent: receivers increased their searching rate and orientated their gaze more towards conspecifics with the number of senders foraging in enhanced food conditions. In no-foraging conditions, receivers decreased their searching and intake rates with the number of senders, but no variation was found in scanning towards conspecifics. Differences in foraging and scanning behaviour between enhanced and no-foraging conditions were much lower when neighbours were separated farther. Overall, information transfer within starling flocks affects individual foraging and scanning behaviour, with receivers monitoring and copying senders behaviour mainly when neighbours are close. Information transfer may be related to predation information (responding to the vigilance of conspecifics) and foraging information (responding to the feeding success of conspecifics). Both sources of information, balanced by neighbour distance, may simultaneously affect the behaviour of individuals in natural conditions.Communicated by H. Kokko     

Short-term effects of data-loggers on Cory’s shearwater (<Emphasis Type="Italic">Calonectris diomedea</Emphasis>)

We investigated the possible effects of a 12-g data-logger attached to a darvic ring on the performance of Corys shearwater (Calonectris diomedea, 600–850 g) from two different colonies in the western Mediterranean Sea. We compared return rates, current breeding success and body condition between equipped and unequipped birds. Effects on feeding ecology during winter and breeding period was also evaluated through the measurement of stable isotopes of carbon (¹³C) and nitrogen (¹⁵N) in one of the colonies. We found no evidence of negative effects of loggers on demographic parameters analysed or in feeding ecology. Power analyses suggested a high power to detect medium effect size, but a low power to detect small changes. Despite the non-significant results we could only exclude medium to strong effects of the devices on one of these parameters. We detected some short-term negative effects on body condition for the equipped birds, but these were unlikely to have had important consequences. Results suggest that the use of loggers is an adequate methodology to obtain information from seabirds at sea, but data should be carefully interpreted with regard to potential biases during severe environmental conditions.Communicated by S.A. Poulet, Roscoff     

Multi-scale foraging variability in Northern gannet (Morus bassanus) fuels potential foraging plasticity

The survival of marine predators depends on behavioural plasticity to cope with changes in prey distribution. Variability in behaviour might predict plasticity and is easier to assess than plasticity. Using miniaturized GPS loggers over several breeding seasons in two Norwegian Northern gannet (Morus bassanus) colonies, we investigated if and how the variability within and between individuals, but also between colonies and years, affected foraging strategies. Results revealed strong individual variability (foraging trip durations, foraging effort and different foraging areas). Individuals from both colonies showed preferred commuting routes, flight bearings and feeding hotspots. Individuals from the largest colony used larger and more foraging areas than individuals from the small colony. Feeding hotspots and foraging ranges varied amongst years in the largest colony only. Our study demonstrated that gannets show flexibility by changing prey fields that are driven by shifting oceanographic conditions.     

Worker size variability and foraging efficiency in Veromessor pergandei (Hymenoptera: Formicidae)

Summary As foragers of the harvester ant, Veromessor pergandei, travel further from their nest they spend significantly more time sampling seeds in experimental patches. Although accepted seeds are heavier than offered seeds, mass of accepted seed is not correlated with sampling time. Variably sized V. pergandei workers do not size-match; little, if any, variance in size of seed selected can be attributed to body size of forager. The lack of size-matching in V. pergandei suggests individual performance may be an inadequate measure of colony foraging success.     

The establishment of foraging flocks in house sparrows: risk of predation and daily temperature

Summary The foraging decisions of animals often reflect a trade-off between the risk of predation and efficient foraging. One way an animal may reduce the risk of predation, and hence exploit a resource patch in relative safety, is by foraging in a group. Solitary pioneer sparrows often recruit others to a food source by making chirrup calls in order to establish foraging flocks. This study describes the decisions of house sparrows that arrive at food resources of different risks of predation. Four feeding sites at different distances from a perching site and from an observer were presented to sparrows. When the feeder was adjacent to the perching site and far from the observer, the pioneers chirruped less frequently and were more likely to forage alone than when the feeder was in the other three positions. There were differences in the scanning behaviour of sparrows at these sites, suggesting that they were responding to different risks of predation. Furthermore, the chirrup rates of pioneer sparrows in this study and a previous study were found to be negatively correlated with maximum daily temperature. This is consistent with the hypothesis that energy requirements may affect the flock establishment decisions of sparrows, and that the benefits of foraging in flocks may be greater at lower temperatures.     

Swim speeds of free-ranging great cormorants

Information about foraging speeds is particularly valuable when the impact of a predator species upon a community of prey has to be defined, as in the case of great cormorants. We measured the swim speed of 12 (six males and six females) free-ranging great cormorants Phalacrocorax carbo, foraging off the Greenland coast during the summer of 2003, using miniaturized data-loggers. Although mean body mass of males was 27% greater than that of females, and mean swim speed of males were 29–57% higher than that of females during foraging phases (but not descent phases) of dives, these differences in speeds were not significant due to high variances. Birds descended to the mean maximum depth of 4.7 m at an average speed of 1.6±0.5 m s⁻¹, a speed similar to that measured in captive cormorants in previous studies. Although bursts of up to 4 m s⁻¹ were recorded, speed usually decreased during the deepest (foraging) phase of dives, being on average 0.8±0.6 m s⁻¹. Speeds measured here should be taken with caution, because the large propeller loggers used to measure speed directly decreased descent speeds by up to 0.5 m s⁻¹ when compared to smaller depth-only loggers. Cormorants in Greenland seem to combine two searching strategies, one requiring low speed to scan the water column or benthos, and one requiring high speed to pursue prey. These two strategies depend on the two main habitats of their prey: pelagic or demersal.     

Optimal foraging and territorial defence in the Great Tit (Parus major)

Summary We describe an experiment designed to investigate the trade-off between foraging and territorial vigilance in the Great Tit. Captive territorial male Great Tits were observed while foraging in a large indoor aviary. They obtained food from two operant patches in which the supply of food was gradually depleted during a visit. We predicted that during control sessions the birds would switch between patches in such a way as to maximise their overall feeding rate. In experimental sessions, we introduced briefly a rival male as an intruder at the start of the test. The foraging male could see the rival only when travelling between patches and not while feeding within patches. We predicted that during experimental sessions birds would switch between patches more often than in control tests, sacrificing food intake for territorial vigilance. Three of the four males tested behaved in approximately the predicted manner. We discuss the use of an inverse optimality argument to provide a calibration of feeding against the benefit resulting from territorial vigilance.     

Seasonal changes in movements and habitat preferences of the scalloped hammerhead shark (Sphyrna lewini) while refuging near an oceanic island

Movements and habitat preferences of sharks relative to a central location are widely documented for many species; however, the reasons for such behaviors are currently unknown. Do movements vary spatially or temporally or between individuals? Do sharks have seasonal habitat and environmental preferences or simply perform movements at random at any time of the year? To help understand requirements for the designation of critical habitats for an endangered top predator and to develop zoning and management plans for key habitats, we examined vertical and horizontal movements, and determined habitat and environmental preferences of scalloped hammerhead sharks (Sphyrna lewini). We tracked seven hammerheads for 19–96 h at Wolf Island (1.38ºN, 91.82ºW) between 2007 and 2009 using ultrasonic transmitters with depth and temperature sensors, and we profiled temperature through the water column. Movements of individual hammerheads fell in two classes: constrained (remaining near the island) and dispersive (moving offshore to pelagic environments). The central activity space or kernel off the southeast side of Wolf Island was small and common to most, but the area varied among individuals (mean ± SE 0.25 ± 0.2 km²), not exceeding 0.6 km² for any of the sharks, and not changing significantly between seasons. In general, hammerheads showed preference for the up-current habitat on the eastern side of Wolf Island in both the warm and cold seasons. However, the depth of sharks varied with season, apparently in response to seasonal changes in the vertical structure of temperature. Hammerheads performed frequent vertical excursions above the thermocline during offshore movements and, in general, were observed to prefer temperatures of 23–26 °C found above the thermocline. At times, though individuals moved into the thermocline and made brief dives below it. Our results provided evidence that hammerheads (1) are highly selective of location (i.e., habitat on up-current side of island) and depth (i.e., top of the thermocline) while refuging, where they may carry out essential activities such as cleaning and thermoregulation, and (2) perform exploratory vertical movements by diving the width of the mixed layer and occasionally diving below the thermocline while moving offshore, most likely for foraging.