Telemetry without collars: performance of fur- and ear-mounted satellite tags for evaluating the movement and behaviour of polar bears

doi:10.21203/rs.3.rs-3848682/v1

Download PDF

Research Article

Telemetry without collars: performance of fur- and ear-mounted satellite tags for evaluating the movement and behaviour of polar bears

https://doi.org/10.21203/rs.3.rs-3848682/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 15 Jul, 2024

Read the published version in Animal Biotelemetry →

You are reading this latest preprint version

The study of animal movement provides insights into underlying ecological processes and informs analyses of behaviour and resource use, which have implications for species management and conservation. The tools used to study animal movement have evolved over the past decades, allowing for data collection from a variety of species, including those living in remote environments. Satellite-linked radio and GPS collars have been used to study polar bear (Ursus maritimus) ecology and movements throughout the circumpolar Arctic for over 50 years. However, due to morphology and growth constraints, only adult female polar bears can be reliably collared for long durations. Further, collars have proven to be safe and reliable but there has been opposition to their use, resulting in a deficiency in data across much of the species’ range. To bolster knowledge of movement characteristics and behaviours for polar bears other than adult females, while also providing an alternative to collars, we tested the use of fur- and ear-mounted telemetry tags that can be affixed to polar bears of any sex and age. We also used data collected from the tags to quantify the amount of time subadult and adult males spent resting versus traveling while on land. Our results show fur tags remained functional for shorter durations than ear tags, but had comparable positional error estimates and provided sufficient data to model different behavioural states. Further, as hypothesized, subadult and adult male polar bears spent the majority of their time resting while on land, likely as a means of conserving energy until the sea ice reforms in early winter. Fur tags provide promise as a shorter-term means of collecting movement data from free-ranging polar bears.

telemetry

fur tag

hair tag

polar bear

Ursus maritimus

The movement of organisms is an essential component of life, shaped by ecological and biological processes often acting across multiple spatial and temporal scales [1–4]. While humans have tracked the movements of animals for millennia, in recent decades, the study of where and how animals travel through their environment has broadened our understanding of the factors influencing habitat selection and species distributions [5–8]. Analyses of animal movements provide opportunities to better understand these species-habitat relationships and improve predictions of space use and, ultimately, population dynamics, which depend on the spatial distribution of individuals.

Methods used to record animal movements have evolved over the past 50 years. Progressively smaller satellite-linked transmitters, along with advances in battery technology, have allowed for remote collection of data from an increasing variety of organisms, including cryptic and migratory species, and those living in remote environments where direct observation is impractical [9, 10]. Associated increases in the spatial and temporal resolution of location data also afford refined insights into biological and environmental factors affecting animal movement. These advances have provided benefits for species conservation and management through the identification of critical habitat and elucidation of shifting movement patterns, such as changes in migration phenology and distribution in response to climate change [11–14].

For some high latitude species, the effects of climate change are an increasing concern due to the higher rates of warming in the Arctic than areas at lower latitude [15–18]. Polar bears (Ursus maritimus) are marine carnivores that travel thousands of kilometers over large seasonal home ranges [19–21]. Their movements are linked to sea-ice dynamics, as bears select for areas of high sea-ice concentration over the continental shelf where ocean productivity and prey availability are high [22, 23]. Due to the dynamic nature of Arctic sea ice, polar bears change their movements in response to sea-ice drift and seasonal fragmentation [24–27]. Over the past several decades, climate-related reductions in sea ice have altered polar bear movements, resulting in shifts in home range sizes and distribution [28–30], increased use of terrestrial habitats [31–33], and observations of long-distance swimming [34–36]. In the southern portion of polar bear range, sea ice undergoes an annual freeze-thaw cycle and bears are forced ashore during the ice-free season [37–39]. During this time, some bears remain close to the coast while others move inland to maternal denning habitat or refugia to avoid disturbance from other bears [40–43]. Thus, understanding the movements of polar bears in light of changes to sea ice is a critical conservation topic.

Since the late 1970s, satellite-linked radio and global positioning system (GPS) collars fitted around the necks of adult female bears have been the primary means of studying polar bear movements, distribution, behaviour, and habitat selection [44, 45]. Generally, subadult bears are not collared due to the potential for growth-related injury, whereas adult males are rarely collared because the circumference of their neck exceeds that of their head, making collars likely to slip off [44, 46, 47]. Although studies of polar bear behaviour and habitat selection have used movement data from collared subadult bears (see [48] and [49]), both sample size and collar functional duration were limited in comparison to the numerous studies examining adult female movements. Other attachment methods, including harnesses [50, 51], ear tags [52, 53], subcutaneous implants [46, 54], and adhesives [52, 55] have been used to temporarily affix transmitters to subadult or adult male polar bears with limited success [44]. For instance, ear-mounted transmitters deployed between 2007–2013 were functional for short durations, averaging approximately 70 days, whereas collars often transmit for several years [44, 53, 56]. Wiig et al. (2017) also noted several instances of injury from infection and forced removal of ear-mounted transmitters. The authors speculated that bears may have torn the transmitters from their ear, or the transmitters may have become caught on objects in their surroundings, resulting in observations of split ears. Thus, there is a need to refine attachment methods to enhance animal welfare and provide location data for subadult and adult male bears.

Although collars are the most reliable means of collecting multi-year polar bear movement data, in addition to the limits on which sex and age classes can be monitored with this method, there has been public concern, particularly from Indigenous communities, about possible negative physical and behavioural impacts. Specifically, concerns of possible injuries caused by collars becoming too tight, and impairment of a bear’s ability to successfully hunt seals have been raised [57–59]. Studies have demonstrated collars have no appreciable impact on polar bear movement rates, body condition, recruitment, or survival [52, 60], and their use has been critical for polar bear management [44]. However, there is an ongoing desire to refine polar bear research techniques to ensure they are effective, less invasive, and respect the views of all stakeholders [57].

The limited amount of movement data collected from subadult and adult male polar bears suggest there may be sex and age-class related differences in movement rates, behaviour, dispersal, and habitat selection [46, 49, 53, 61]; however, additional data are needed to better assess these differences. Thus, to increase knowledge of movements and associated behaviours of polar bears other than adult females, while providing an alternative to collars, we tested the use of fur- and ear-mounted telemetry tags that can be affixed to polar bears of all sex and age-classes. Specifically, we tested three different designs of fur tags and compared their performance to ear tags in terms of both retention time and error resolution. Further, using both fur and ear tags, we collected location data from subadult and adult male polar bears to examine their movement patterns and associated behavioural states during the ice-free season along the Hudson Bay coast in Canada. We hypothesized that subadult and adult male polar bears would spend the majority of the ice-free season resting, gradually increasing their time spent traveling as temperatures decreased and sea ice in Hudson Bay began to reform in early winter.

Study Area

Our study occurred along the southwestern coast of Hudson Bay, Canada, bounded roughly by Fort Severn, Ontario in the southeast and Churchill, Manitoba in the northwest (Fig. 1).

Most of the area falls within the Hudson Bay Lowlands ecoregion, which includes extensive marine beaches and coastal mudflats that become exposed during low tide [62, 63]. Further inland, beach sediments and coastal vegetation transition to areas dominated by mosses and lichen interspersed with large patches of willow (Salix spp.), alder (Alnus spp.), and dwarf birch (Betula glandulosa). In the southern portion of the study area (i.e., below treeline), tundra and wetland vegetation transition to black spruce (Picea mariana) and white spruce (P. glauca) boreal forest [62–64]. Hudson Bay remains frozen for most of the year, but becomes ice-free between August and November. Sea ice typically begins reforming along the western coast in December and reaches its maximum extent and thickness in March or April [65, 66].

Tag Design and Application

We tested three different fur tag designs. The first was made of a ballistic mesh (n = 3; 2021) or rubber (n = 3; 2022) and strips of semi-rigid plastic backing cut into a pentagonal shape with holes punched into each of the five corners (Fig. 2a).

Using a cable puller, polar bear hairs were ensnared and pulled through each hole. A copper ferrule was then placed around the hairs and slid to the base of the tag where it was crimped twice in orthogonal directions using pliers. Herein referred to as a ‘pentagon tag’, this tag was equipped with a 3.9 x 2.0 x 1.9 cm, 26 g Argos Eartag Transmitter (ETA-2620; Telonics, AZ, USA), secured to an accessory bolt on the tag using a nut and locking washer. The second tag design, commercially available as the SeaTrkr GPS/Iridium tag (Telonics, AZ, USA), was similar in design; however, it included an oval-shaped rigid plastic baseplate with 10 equally spaced holes around its outside edge (Fig. 2b). It was attached to hairs in the same manner as the pentagon tag. The SeaTrkr tag was equipped with an Iridium-linked Telonics GPS SeaTrkr-4370 transmitter, which measured 10.3 x 4.5 x 3.6 cm and weighed 190 g. The third tag design, herein referred to as a ‘tribrush tag’, was a rigid plastic triangle with perforated tubes spanning the length of each edge (Fig. 2c). After placing the tag firmly against a bear’s fur, a 12 mm x 75 mm nylon pipe brush was then inserted into each of the three tubes and twisted clockwise, ensnaring the bear’s hair in the brushes’ plastic bristles. The brushes were twisted until they collected enough hair that they could no longer be turned by hand, and the metal handles were cut flush with the base of the bristles. The same ETA-2620 transmitters used on the pentagon tags were attached to the tribrush tags, but were subsequently covered with a protective radome cover that clipped securely to the aforementioned perforated tubes.

The ear tags (ETA-2620; Telonics, AZ, USA) consisted of an Argos transmitter (ST-26; Telonics, AZ, USA) housed in a plastic casing with an integrated washer. Using specialized pliers, the tag and a separate friction-fit pin (Allflex, Rahway, NJ, USA) placed on the ventral side of the bear’s ear were clamped together, secured through a hole made in the ear using a 6-mm punch.

Polar Bear Capture and Tag Deployment

During August and September 2021–2022, subadult (3–4 year) and adult (≥ 5 year) male polar bears were opportunistically located from a helicopter along the Hudson Bay coast near the Ontario-Manitoba border (Fig. 1). Bears were chemically immobilized via remote injection using zolazepam-tiletamine (Zoletil®; Virbac, France) following [67]. Ages were determined using counts of cementum growth layers from a vestigial premolar extracted during handling or records from previous capture [68]. Each bear received one of the three fur tags, all of which were secured to hairs above the thoracic spine, immediately posterior the scapulae (Fig. 2). For two of the tribrush tags, a fast-cure two-part epoxy (3M Scotch-Weld Epoxy Adhesive DP100) was applied to the ensnared hairs to assess the influence of supplementary adhesive on tag retention time. Ear tags were deployed on polar bears immobilized on the Hudson Bay sea ice during spring 2017–2019 and 2022 using the same capture and handling protocols.

Additional ear tags were deployed during fall 2016–2021 on polar bears captured within or near Churchill, Manitoba as part of ongoing operations of the Polar Bear Alert Program [69, 70]. Polar bears deemed a threat to human life or property were captured by Manitoba government staff via remote injection using zolazepam-tiletamine. Sex and reproductive status were determined during handling. Age was verified using the same procedures noted above and included examination of tooth eruption patterns for dependent offspring [68]. Argos-linked tags, including pentagon and tribrush fur tags were programmed to record locations every 4 h, whereas GPS/Iridium SeaTrkr tags were programmed to record locations every 2 h.

Research protocols were reviewed and approved annually by the animal care committees of Environment and Climate Change Canada (Prairie and Northern Region), the Ontario Ministry of Natural Resources and Forestry, University of Alberta BioSciences Animal Care and Use Committee, and York University, and followed the guidelines of the American Society of Mammalogists for the use of wild mammals in research [71] and the Canadian Council on Animal Care (www.ccac.ca).

Data Preparation & Statistical Analyses

We used an automated filtering routine in the argosfilter R package [72] to remove biologically implausible locations that implied movement rates ≥ 40 km h^− 1 between successive locations [73]. Polar bears have been observed running at speeds of 30–40 km h^− 1 for short durations, but generally sustain speeds of approximately 4 km h^− 1 over longer periods [74, 75]; therefore, this threshold represents a conservative upper limit intended to identify only extreme outlier observations [76].

We evaluated two performance metrics for each tag design. First, we measured functional duration (i.e., the number of days transmitters remained active while attached to a bear). Second, we measured horizontal error estimates and error classes associated with each point location recorded using GPS- and Argos-linked transmitters, respectively. Horizontal error estimates for GPS/Iridium transmitters are expressed in meters and represent the radius of a circle surrounding each reported GPS position within which the transmitter was likely located. Each Argos location was assigned one of six possible error classes (3, 2, 1, 0, A, B; in order of decreasing accuracy) based on the number of messages transmitted to passing satellites. Locations derived from ≥4 messages have relatively small horizontal error (< 250 m for Class 3; 500 m for Class 2; 1500 m for Class 1; and > 1500 m for Class 0), while those derived from < 4 messages cannot be estimated [9, 77].

After determining the tags’ functional duration and horizontal error using the full suite of filtered data, locations were subset to include only those observations recorded on land. Using hidden Markov models (HMM), we used these terrestrial locations to examine movements and associated behavioural states of subadult and adult male polar bears during the ice-free period in Hudson Bay. Forays into Hudson Bay (< 50 consecutive locations), whether during the open water season or as the sea ice was forming in early winter, were included as long as the bear returned to land during the same season before moving onto the sea ice for the winter. The remaining locations had sporadic temporal gaps longer than the programmed fix rates for both Argos- (4 h) and GPS/Iridium-linked tags (2 h). Considering HMMs require telemetry data to be provided at consistent intervals [78, 79], we standardized the data to consistent 4 h time intervals by interpolating missing Argos locations and rarifying GPS/Iridium locations using the R package crawl [80, 81]. There were also instances of longer gaps (> 12 h) in the data due to consecutive failed fix attempts. Interpolating telemetry data across longer gaps is problematic because it produces uncertainty in missing location estimates [82, 83]. Therefore, we segmented polar bear movement paths when temporal gaps between successive locations were longer than 24 h, thereby removing these intervals of missing data from analyses. Each track segment was treated as an independent time series arising from the same underlying statistical model, and because fitted parameters are common to all tracks, the same behavioural states influence movements both before and after the gaps, and any existing correlation was accounted for [83, 84]. Lastly, individual segments with < 100 locations were omitted from analyses because they provided little information about the transition between, and relative time spent in, different behavioural states [83].

We estimated the proportion of time subadult and adult male bears spent resting vs. traveling using the R package momentuHMM, which uses discrete-time HMMs to infer latent behavioural states and associated transition probabilities from animal telemetry data [79]. Fundamentally, HMMs estimate distinct, unobserved behavioural states based on attendant movement characteristics, often using derived quantities such as step lengths (i.e., distance between successive point locations) and turn angles (i.e., a measure of directional change between subsequent steps) between consecutive telemetry locations [78, 85]. For instance, ‘foraging’ behaviour may be associated with movement paths characterized by short step lengths and large, turn angles, indicating tortuous movements, whereas ‘transiting’ behaviour may be characterized by long step lengths and near-linear movement paths (i.e., near-zero turn angles). Here, we modelled polar bear step lengths and turn angles using gamma and von Mises distributions, respectively. Due to the sensitivity of HMMs to initial values when calculating maximum likelihood estimates for model parameters, we refit each model 50 times using different starting values for step length mean, step length standard deviation, turn angle mean, and turn angle concentration for the two behavioural states. Standardized Argos and GPS/Iridium data (i.e., regularized to consistent 4-h fixes) were pooled to increase sample size and inferential power. We also modelled state transition probabilities as a function of ambient temperature (temp and temp²) using hourly temperature data collected from the nearest permanent weather station, located in either Churchill, Manitoba or Fort Severn, Ontario (https://climate.weather.gc.ca/; accessed October 2022). We used Akaike’s information criterion with a correction for small sample sizes (AIC_c) to compare fitted models, selecting formulations with the lowest AIC_c values as the best or most parsimonious model upon which further inferences were based [86]. Lastly, for the selected model(s), we estimated the proportion of time bears spent resting vs. traveling using the Viterbi algorithm, which provides the most likely sequence of unobserved behavioural states given observed data and fitted models [83, 85].

We obtained 33,840 locations from 58 individuals equipped with ear or fur tags between 2016 and 2022. Two adult males were recaptured during the study period and equipped with a second ear tag after their initial tag had been removed or stopped transmitting. Fur tags accounted for 6,232 locations. Filtering with the argosfilter R package removed 2,149 spurious locations, which accounted for ca. 6% of the combined data collected using Argos-linked ear and fur tags. An additional 19,093 on-ice locations were removed. After segmenting tracks with temporal gaps > 24 h and omitting those with < 100 locations, missing locations were interpolated, resulting in a total of 3,104 locations across 13 tracks.

Fur tags were affixed to 1 subadult and 15 adult male polar bears; 6 bears were equipped with pentagon tags, 6 were equipped with SeaTrkr tags, and the remaining 4 were equipped with tribrush tags. Functional duration varied considerably both within and among the three different designs (Fig. 3).

Pentagon tags remained active for a mean of 22 d (range: 6–66 d; SE: 9.2). SeaTrkr tags had the longest mean functional duration of the three designs, averaging 58 d (range: 29–100 d; SE: 9.8), whereas tribrush tags remained active for a mean of 47 d (range: 2-114 d; SE: 27.3). The two longest-lasting tribrush tags (69 and 114 d) were the only tags applied with supplementary adhesive. Ear tags, which were affixed to 10 subadult and 32 adult male polar bears, remained active for the longest mean duration of 137 d (range: 10–364 d; SE: 12.7). Ten ear tags lasted > 200 d.

Pentagon tags had a high proportion of fixes (39%) with Class 0 (> 1500 m) Argos location error, the largest of the Argos error classes. The remaining fixes had horizontal error estimates that were either < 1500 m (i.e., Class 1–3: 14%) or could not be estimated (i.e., Class A and B: 47%; see Fig. 4).

Tribrush tags had the highest proportion of fixes for which horizontal error could not be estimated (Class A and B: 75%). Ear tags had fairly uniform proportions of Class 0–2 (Class 0: 13%; Class 1: 14%; Class 2: 14%) and A (18%) horizontal errors, with slightly smaller and larger proportions of Class 3 (6%) and B errors (36%), respectively. Locations recorded using the six GPS/Iridium-equipped SeaTrkr tags had considerably smaller horizontal error values (mean = 11 m, range 4-102 m) than the Argos-linked tags.

While influenced by differences in tag retention times, bears in this study exhibited considerable variation in movement characteristics, traveling cumulative distances ranging from 7 to 1,064 km (mean: 309 km; SE: 40 km). Similarly, straight-line distances traveled from tag application sites varied among individuals, ranging from 0 km to 343 km (mean: 73 km; SE: 13 km). Lastly, path tortuosity, a measure of directional persistence among successive locations where a value of 1 indicates a perfectly linear path and values approaching zero indicate a more convoluted path, varied from 0.01 to 0.93 (mean: 0.25; SE: 0.04). Despite individual variation in movement characteristics, two-state HMMs reliably differentiated among two behavioural states. The presumed resting state was characterized by short step lengths (mean: 0.06 km; 0.01 km/h; Fig. 5) and turn angles with greater concentration around 𝜋 and – 𝜋, whereas the presumed traveling state was characterized by longer mean step lengths (mean: 1.14 km; 0.29 km/h) and turn angles concentrated closer to 0, suggesting greater directional persistence among successive locations.

Based on model selection using AIC_c, the two-state HMM that included linear effects of ambient temperature on state transition probabilities was deemed more parsimonious than the remaining models that included a polynomial term for ambient temperature and no temperature covariates (Table 1), suggesting bears travelled more as temperatures cooled (Fig. 6).

Table 1

Model selection of two-state hidden Markov models fit to subadult and adult male polar bears equipped with Argos- and GPS/Iridium-linked ear and fur tags on the coast of Hudson Bay between 2016–2022. Models are defined in the *Methods* section; K is the number of parameters in the model, and ΔAIC_c is the Akaike information criterion of each model relative to the best fitting or most parsimonious model with the lowest AIC_c score. The base model included no effect of temperature on state transition probabilities.
Model	K	AIC_c	ΔAIC_c	Log Likelihood
Temp	13	3883.136	0	4802.445
Temp + Temp²	15	3887.772	4.636	3951.084
⁻	17	3931.367	48.231	2597.298

Table 1

Model selection of hidden Markov models.
Model	K	AIC_c	ΔAIC_c	Log Likelihood
Temp	13	3883.136	0	4802.445
Temp + Temp²	15	3887.772	4.636	3951.084
⁻	17	3931.367	48.231	2597.298
Model selection of two-state hidden Markov models fit to subadult and adult male polar bears equipped with Argos- and GPS/Iridium-linked ear and fur tags on the coast of Hudson Bay during between 2016–2022. Models are defined in the Methods section; K is the number of parameters in the model, and ΔAIC_c is the Akaike information criterion of each model relative to the best fitting or most parsimonious model with the lowest AIC_c score. The base model included no effect of temperature on state transition probabilities.

Lastly, results from the Viterbi algorithm used to estimate the most likely sequence of unobserved behavioural states given the top-ranking model suggested bears spent 72% of their time resting and 28% of their time traveling while on land.

For decades, telemetry collars have remained the primary means of collecting long-term, high-resolution movement data from adult female polar bears [44]. However, the need to collect movement data from other age- and sex-classes of polar bears, together with a desire to provide alternatives to collars, particularly for certain shorter-term applications, led to the development of novel telemetry devices, including the fur tags we described and tested. Although fur tags had shorter mean functional durations than ear tags, they had similar horizontal error estimates and provided sufficient data to quantify behavioural states of free-ranging subadult and adult male polar bears.

Although the specific causes of tag detachment are unknown, there are several aspects of polar bear behaviour that may have contributed to the short functional durations observed for fur tags. While onshore, subadult and adult male polar bears generally remain close to the coast where they form aggregations and rest in shallow earthen pits [64, 87, 88]. Bears are frequently observed lying in a prone position, but routinely rest in lateral or supine positions as well. Between bouts of resting, bears occasionally swim in Hudson Bay or travel further inland where coastal tundra transitions to areas dominated by taller vegetation, including willow and black spruce [39, 64, 89]. Thus, fur-mounted tags may be subjected to shear forces from ocean waves and abrasion against both sand and occasionally dense terrestrial vegetation. Male polar bears also engage in social play [90], including grappling and wrestling, which involve bears biting and wrapping their forelegs around the neck and/or shoulder region of their partner. Therefore, fur tags may also be susceptible to detachment during social play because unlike ear tags, they were not permanently secured through an appendage but rather affixed to a more exposed part of the body.

Compared to ear tags and collars, fur tags were designed to remain affixed to polar bears for a relatively short period, as polar bear fur is replaced annually during a gradual moult between May and August [91–93]. Therefore, the maximum duration any fur-mounted tag can remain affixed to a free-ranging polar bear is approximately one year, after which it will be shed. Given the timing of our study (September to December), moulting likely did not contribute to premature detachment; however, it remains a consideration for future deployments, particularly during the spring and summer.

Differences in mean functional duration among the three fur tag designs are likely attributable to the method of attachment. For instance, SeaTrkr and pentagon tags were both attached using copper ferules crimped around multiple tufts of hair; however, SeaTrkr tags were secured to ten separate tufts of hair while pentagon tags were only secured to five. Considering SeaTrkr tags are approximately double the size and nearly three times the weight of pentagon tags, it appears doubling the number of attachment points contributed to the difference in mean retention times. Also, SeaTrkr tags include a smooth outer casing, designed to reduce drag, whereas the Argos transmitters attached to pentagon tags remained exposed, secured to an accessory bolt on the tags using a nut and locking washer. Wiig et al. (2017) noted several observations of polar bears with pieces of discarded fishing net caught on plastic identification ear tags, which are smaller than the Argos transmitters. The authors also speculate that instances of lost ear tag transmitters may be the result of similar entanglements. Thus, the more streamlined design of the SeaTrkr tags may have rendered them less likely than the irregularly shaped, multi-part pentagon tags to become caught in surrounding objects, or inadvertently removed by conspecifics during sparring.

Tribrush tags were attached by entangling hairs in three nylon-bristle pipe brushes that were twisted repeatedly inside perforated tubes spanning the length of the tags’ edges. Although care was taken during application to ensnare as much hair as possible, the relatively short coats of bears in late summer may have contributed to the tags’ short functional duration. Without adequate contact between the bristles and hair, tags may have loosened as the bears moved, leaving them susceptible to detachment. Application of supplementary adhesive appears to have enhanced retention time, as tribrush tags applied using two-part epoxy lasted 69 and 114 days, the longest duration of all the fur tags. The remaining two tribrush tags, which were applied without epoxy, lasted only 2 and 3 days. Although the exothermic reaction of the two-part epoxy caused concerns (i.e., damage to the hair and/or skin), a more targeted, lower volume application likely could ameliorate these issues. Unlike fur tags, which are designed to remain affixed for a short period, ear tags may remain attached indefinitely because they are mounted through a hole in the ear and do not include a drop-off mechanism. While there have been reports of detachments [53], their size, means of attachment, and peripheral location likely make ear tags less prone to incidental detachment than fur tags, which are only secured to hair that may be shed or break on an exposed part of the bears’ torso. Fur tags are less likely to cause injury to the bear if they become entangled in debris or are pulled by another bear during social interactions, whereas ear tags, if entangled or pulled could cause injury.

Animal location data recorded using GPS receivers are usually accurate to < 20 m, whereas horizontal error estimates associated with data collected using Argos transmitters can only be specified to within < 250 m [9, 77]. Accordingly, differences between GPS and Argos systems in terms of how location data are recorded is likely responsible for the higher resolution horizontal error estimates associated with the SeaTrkr GPS/Iridium tags compared to the Argos-linked tags.

Among the Argos-linked transmitters, tribrush tags had the largest proportion of fixes for which horizontal error could not be estimated (i.e., Class A and B). Transmitters were placed in the same position on each of the six bears fitted with the pentagon and tribrush tags. However, in addition to slightly different means of attachment, tribrush tags were equipped with plastic radome covers, which were intended to help protect the transmitters from wind, precipitation, and abrasion. Pentagon tags did not include a protective covering because the base was made of a flexible material that was less amenable to a cover. While radome covers are meant to have a minimal effect on the attenuation of electromagnetic signals, research using stationary GPS arrays has shown they may reduce the accuracy of GPS position estimates, particularly along the vertical axis [94]. The magnitude of signal attenuation also depends on the antenna design, along with the composition and thickness of the cover [94, 95]. Thus, the radome covers may have degraded the transmitters’ ability to reliably communicate with satellites, perhaps contributing to the higher proportion of inestimable horizontal error values associated with tribrush tags, a trend not observed in the nearly identical pentagon tags. Both pentagon tags and ear tags had similar proportions of fixes with Class 0–3 error estimates despite differences in attachment. Wiig et al. (2017) reported similar proportions of fixes with high resolution error estimates for Argos-equipped SPOT-227 and − 305A ear tags (mean: 64.68% and 53.10%, respectively) deployed on subadult and adult polar bears of both sexes. While male polar bears may spend extended periods during the ice-free season lying in shallow earthen pits, often excavated in coastal ridges [88, 96], it appears fur and ear tags remained comparably unobscured, resulting in similar proportions of successful transmissions to Argos satellites. The consistently low horizontal error estimates associated with GPS/Iridium-linked SeaTrkr tags further suggests that fur tags positioned on a bear’s back provide adequate communication with satellites and comparable error resolution to conventional ear tags and GPS collars. The high positional accuracy of the SeaTrkr tags, along with their longer mean retention time, suggest these tags may provide the best option among the three fur-mounted tag designs for tracking polar bear movements over short time periods.

Our results demonstrate subadult and adult male polar bears limit their movements while ashore, corroborating previous observational studies that showed bears spent approximately 70–90% of their time resting during the ice-free period in Hudson Bay [87, 97, 98]. Because polar bears are prone to hyperthermia, searching for and consuming terrestrial sources of food, which provide limited nutritional contributions [99], is likely more energetically costly than resting and fasting until the ice reforms [88, 98, 100–102]. Indeed, the top-ranking two-state HMM, which included a linear effect of ambient temperature on stationary state probabilities, suggested bears travelled less during warmer weather, increasing their time spent traveling as temperatures cooled. Colder temperatures coincide with sea ice formation during the late fall and early winter in Hudson Bay when all polar bears, with the exception of pregnant females, begin a seasonal migration towards newly forming sea ice [27, 103, 104]. Accordingly, the estimated proportion of time subadult and adult male bears spent resting during our study period further supports the notion that it is likely more metabolically efficient for polar bears to rest rather than expending energy for often unpredictable opportunities to consume terrestrial foods that provide limited energetic returns [88, 98, 100–102].

We only considered two behavioural states due to the limited temporal resolution and number of locations recorded using our tags. Given these limitations, along with research suggesting bears spend only ca. 3% of their overall time budget foraging while on land, it is unlikely HMMs could reliably distinguish between three or more distinct states [48, 97, 105]. Others have demonstrated polar bears engage in additional behaviours, particularly during winter and spring while on the sea ice [48, 97, 106, 107]. For instance, Togunov et al. (2022), using a similar HMM approach, showed adult female bears alternated between three distinct movement-related behavioural states (drifting, area restricted search, and olfactory search) while on the sea ice between January and June. Similarly, Pagano et al. (2017) showed video-linked accelerometer data collected at 2-second intervals could be used to distinguish between three behaviours (i.e., resting, walking, and swimming), and were capable of identifying up to five behaviours while bears where on land (i.e., resting, walking, eating, grooming, and head shaking). Future behavioural studies using fur tags may consider increasing fix-rates to identify more intricate behaviours and attendant habitat associations.

While collars remain the best option for collecting long-term, high-resolution movement data [44], fur-mounted tags offer promise for shorter-term applications. For instance, fur tags may be used to study the movements and behaviours of polar bears during particularly important periods, such as the spring hyperphagia and mating seasons, transition on and off the sea ice, and during the ice-free season. Short-term monitoring of subadult and adult males may further clarify sex- and age-class-related differences in movement characteristics, including home range size and habitat selection. Further, fur tags may be well suited for use in mitigation of human-bear conflicts. Bears captured near human settlements could be equipped with the temporary tags to monitor their proximity to people and infrastructure, allowing conservation staff to intercept the bears and prevent recidivist encounters. GPS collars are poorly suited for this singular task because most bears involved in conflicts are subadult and adult males [69, 70, 108, 109]. Fur-mounted tags are also less expensive than GPS collars, thereby allowing for monitoring of more bears for the same cost. Fur tags offer promise as a safe, less expensive, shorter-term means of monitoring the movements of free-ranging polar bears for purposes of both applied scientific research and mitigating human-bear conflicts.

Further refinement and testing of fur tag designs may improve their reliability. Our results demonstrate that increasing the number of attachment points, along with use of suitable supplementary adhesive, ought to increase mean retention times. Further tests may also be used to evaluate their suitability for use on other age classes. Tracking bears other than adult females is important for broadening understanding of critical aspects of the species’ habitat use and behaviour, particularly as the Arctic warms in response to ongoing climate change. Current estimates suggest climate-mediated changes to Arctic environments are likely to cause shifts in polar bear distribution and habitat selection, and result in higher rates of human-bear conflicts [23, 70, 110, 111]. Therefore, along with other remote tracking technologies, including ear tags, fur-mounted tags offer a means of collecting data that will enable managers and other stakeholders to make informed decisions vital for the ongoing management and conservation of polar bears.

Ethical Approval

Funding

Funding for this work was provided the Banrock Station Environmental Trust, Canadian Association of Zoos and Aquariums, the Churchill Northern Studies Centre, Canadian Wildlife Federation, Environment and Climate Change Canada, Earth Rangers, Hauser Bears, the Isdell Family Foundation, Kansas City Zoo, Manitoba Department of Agriculture and Resource Development, Manitoba Sustainable Development, Natural Sciences and Engineering Research Council of Canada, Pittsburgh Zoo Conservation Fund, Polar Bears International, Polar Continental Shelf Project, Polar Knowledge Canada, Quark Expeditions, San Diego Zoo Wildlife Alliance, the University of Alberta, Wildlife Media, Inc, and the Weston Family Foundation.

Author Contribution

BJ.K., J.K. and G.Y. developed the fur-mounted tags. A.D., D.M., G.T., J.N., N.L., T.R. and V.T. deployed the ear- and fur-mounted tags on free-ranging polar bears in Hudson Bay. A.J. and T.R. compiled and formatted the data. T.R. analyzed the data and wrote the manuscript, incorporating feedback from all authors. All authors reviewed the final manuscript.

Acknowledgments

This project was supported by the Banrock Station Environmental Trust, Canadian Association of Zoos and Aquariums, the Churchill Northern Studies Centre, Canadian Wildlife Federation, Environment and Climate Change Canada, Earth Rangers, Hauser Bears, the Isdell Family Foundation, Kansas City Zoo, Manitoba Department of Agriculture and Resource Development, Manitoba Sustainable Development, Natural Sciences and Engineering Research Council of Canada, Pittsburgh Zoo Conservation Fund, Polar Bears International, Polar Continental Shelf Project, Polar Knowledge Canada, Quark Expeditions, San Diego Zoo Wildlife Alliance, the University of Alberta, Wildlife Media, Inc, and the Weston Family Foundation. We are grateful to our partners who conducted initial testing on captive polar bears, all of whom played a critical role in early tag design: Point Defiance Zoo and Aquarium; Kansas City Zoo; Columbus Zoo and Aquarium; San Diego Zoo, Como Park Zoo; Oregon Zoo; Louisville Zoo; Maryland Zoo; Hogle Zoo; Assiniboine Park Zoo, and Skandinavisk Dyrepark.

Availability of Data and Materials

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Johnson AR, Wiens JA, Milne BT, Crist TO. Animal movements and population dynamics in heterogeneous landscapes. Landsc Ecol. 1992;7:63–75.
Turchin P. Quantitative analysis of movement: measuring and modeling population redistribution in animals and plants. Sunderland: Sinauer Associates; 1998.
Nathan R, Getz WM, Revilla E, Holyoak M, Kadmon R, Saltz D, Smouse PE. A movement ecology paradigm for unifying organismal movement research. Proc Natl Acad Sci U S A. 2008;105:19052–9.
Hooten MB, Johnson DS, McClintock BT, Morales JM. Animal movement: statistical models for telemetry data. Boca Raton: CRC Press; 2017.
Morales JM, Moorcroft PR, Matthiopoulos J, Frair JL, Kie JG, Powell RA, Merrill, EH, Haydon, DT. Building the bridge between animal movement and population dynamics. Philos Trans R Soc Lond, B, Biol Sci. 2010;365:2289–301.
Avgar T, Potts JR, Lewis MA, Boyce MS. Integrated step selection analysis: bridging the gap between resource selection and animal movement. Methods Ecol Evol. 2016;7:619–30.
Guisan A, Thuiller W, Zimmermann NE. Habitat suitability and distribution models: with applications in R. Cambridge: Cambridge University Press; 2017.
Turchin P. Translating foraging movements in heterogeneous environments into the spatial distribution of foragers. Ecology. 1991;72:1253–66.
Tomkiewicz SM, Fuller MR, Kie JG, Bates KK. Global positioning system and associated technologies in animal behaviour and ecological research. Philos Trans R Soc Lond, B, Biol Sci. 2010;365:2163–76.
Kays R, Crofoot MC, Jetz W, Wikelski M. Terrestrial animal tracking as an eye on life and planet. Science. 2015;348:1–9.
Allen AM, Singh NJ. Linking movement ecology with wildlife management and conservation. Front Ecol Evol. 2016;3:155.
Hays GC, Bailey H, Bograd SJ, Bowen WD, Campagna C, Carmichael RH, et al. Translating marine animal tracking data into conservation policy and management. Trends Ecol Evol. 2019;34:459–73.
Bastille-Rousseau G, Wittemyer G. Characterizing the landscape of movement to identify critical wildlife habitat and corridors. Conserv. Biol. 2021;35:346–59.
Davidson S, Bohrer G, Gurarie E, Lapoint S, Mahoney P, Boelman N, et al. Ecological insights from three decades of animal movement tracking across a changing Arctic. Science. 2020;370:712–5.
Stirling I, Derocher AE. Possible impacts of climatic warming on polar bears. Arctic. 1993;46:240–5.
Tynan CT, Demaster DP. Observations and predictions of Arctic climatic change: potential effects on marine mammals. Arctic. 1997;50:308–22.
Post E, Bhatt US, Bitz CM, Brodie JF, Fulton TL, Hebblewhite M, et al. Ecological consequences of sea-ice decline. Science. 2013;341:519–24.
IPCC. Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. 2021.
Durner GM, Amstrup SC. Movements of a polar bear from northern Alaska to northern Greenland. Arctic. 1995;48:338–41.
Parks EK, Derocher AE, Lunn NJ. Seasonal and annual movement patterns of polar bears on the sea ice of Hudson Bay. Can J Zool. 2006;84:1281–94.
Johnson AC, Pongracz JD, Derocher AE. Long-distance movement of a female polar bear from Canada to Russia. Arctic. 2017;70:121–8.
Mauritzen M, Derocher AE, Wiig Ø, Belikov SE, Boltunov AN, Hansen Edmond, et al. Using satellite telemetry to define spatial population structure in polar bears in the Norwegian and western Russian Arctic. J Appl Ecol. 2002;39:79–90.
Stirling I, Derocher AE. Effects of climate warming on polar bears: a review of the evidence. Glob Chang Biol. 2012;18:2694–706.
Sahanatien V, Derocher AE. Monitoring sea ice habitat fragmentation for polar bear conservation. Anim Conserv. 2012;15:397–406.
Auger-Méthé M, Lewis MA, Derocher AE. Home ranges in moving habitats: polar bears and sea ice. Ecography. 2016;39:26–35.
Klappstein NJ, Togunov RR, Reimer JR, Lunn NJ, Derocher AE. Patterns of sea ice drift and polar bear (Ursus maritimus) movement in Hudson Bay. Mar Ecol Prog Ser. 2020;641:227–40.
Biddlecombe BA, Bayne EM, Lunn NJ, McGeachy D, Derocher AE. Effects of sea ice fragmentation on polar bear migratory movement in Hudson Bay. Mar Ecol Prog Ser. 2021;666:231–41.
Durner GM, Douglas DC, Albeke SE, Whiteman JP, Amstrup SC, Richardson E, et al. Increased Arctic sea ice drift alters adult female polar bear movements and energetics. Glob Chang Biol. 2017;23:3460–73.
Laidre KL, Stern H, Born EW, Heagerty P, Atkinson S, Wiig Ø, et al. Changes in winter and spring resource selection by polar bears Ursus maritimus in Baffin Bay over two decades of sea-ice loss. Endanger Species Res. 2018;36:1–14.
Laidre KL, Born EW, Atkinson SN, Wiig Ø, Andersen LW, Lunn NJ, et al. Range contraction and increasing isolation of a polar bear subpopulation in an era of sea-ice loss. Ecol Evol. 2018;8:2062–75.
Rode KD, Wilson RR, Regehr EV, Martin M St., Douglas DC, Olson J. Increased land use by Chukchi Sea polar bears in relation to changing sea ice conditions. PLoS One. 2015;10:e0142213.
Pagano AM, Durner GM, Atwood TC, Douglas DC. Effects of sea ice decline and summer land use on polar bear home range size in the Beaufort Sea. Ecosphere. 2021;12:p.e03768.
Rode KD, Douglas DC, Atwood TC, Durner GM, Wilson RR, Pagano AM. Observed and forecasted changes in land use by polar bears in the Beaufort and Chukchi Seas, 1985–2040. Glob Ecol Conserv. 2022;40:e02319.
Durner GM, Whiteman JP, Harlow HJ, Amstrup SC, Regehr EV, Ben-David M. Consequences of long-distance swimming and travel over deep-water pack ice for a female polar bear during a year of extreme sea ice retreat. Polar Biol. 2011;34:975–84.
Pagano AM, Durner GM, Amstrup SC, Simac KS, York GS. Long-distance swimming by polar bears (Ursus maritimus) of the southern Beaufort Sea during years of extensive open water. Can J Zool. 2012;90:663–76.
Pilfold NW, McCall A, Derocher AE, Lunn NJ, Richardson E. Migratory response of polar bears to sea ice loss: to swim or not to swim. Ecography. 2017;40:189–99.
Russell RH. The food habits of polar bears of James Bay and southwest Hudson Bay in summer and autumn. Arctic. 1975;28:117–29.
Jonkel C, Smith P, Stirling I, Kolenosky GB. The present status of the polar bear in the James Bay and Belcher Islands area. Canadian Wildlife Service, Occasional Paper No. 26. Ottawa: Canadian Wildlife Service. 1976: 42 pp.
Stirling I, Jonkel C, Smith P, Robertson R, Cross R. The ecology of the polar bear (Ursus maritimus) along the western coast of Hudson Bay. Canadian Wildlife Service, Occasional Paper No. 33. Ottawa: Canadian Wildlife Service. 1977: 64 pp.
Kolenosky G, Prevett J. Productivity and maternity denning of polar bears in Ontario. International Conference on Bear Research and Management. 1983;5:238–45.
Derocher AE, Stirling I. Distribution of polar bears (Ursus maritimus) during the ice-free period in western Hudson Bay. Can J Zool. 1990;68:1395–403.
Amstrup SC, Gardner C. Polar bear maternity denning in the Beaufort Sea. J Wildl Manage. 1994;58:1–10.
Obbard ME, Walton LR. The importance of polar bear provincial park to the southern Hudson Bay polar bear population in the context of future climate change. Parks and protected areas research in Ontario, 2004: Planning northern parks and protected areas. Proceedings of the Parks Research Forum of Ontario General Meeting, May 4–6, 2004. 2004;105–16.
Laidre KL, Durner GM, Lunn NJ, Regehr EV, Atwood TC, Rode KD, et al. The role of satellite telemetry data in 21st century conservation of polar bears (Ursus maritimus). Front Mar Sci. 2022;9: 816666
Togunov RR, Derocher AE, Lunn NJ, Auger-Méthé M. Drivers of polar bear behavior and the possible effects of prey availability on foraging strategy. Mov Ecol. 2022;10:p.50.
Amstrup SC, Durner GM, McDonald TL, Mulcahy DM, Garner GW. Comparing movement patterns of satellite-tagged male and female polar bears. Can J Zool. 2001;79:2147–58.
Taylor MK, Akeeagok S, Andriashek D, Barbour W, Born EW, Calvert W, et al. Delineating Canadian and Greenland polar bear (Ursus maritimus) populations by cluster analysis of movements. Can J Zool. 2001;79:690–709.
Pagano AM, Rode KD, Cutting A, Owen MA, Jensen S, Ware J V., et al. Using tri-axial accelerometers to identify wild polar bear behaviors. Endanger Species Res. 2017;32:19–33.
Johnson AC, Derocher AE. Variation in habitat use of Beaufort Sea polar bears. Polar Biol. 2020;43:1247–60.
Schweinsburg RE, Lee LJ. Movement of four satellite-monitored polar bears in Lancaster Sound, Northwest Territories. Arctic. 1982;35:504–11.
Taylor MK. The effect of radio transmitter harnesses on free-ranging polar bears. International Conference on Bear Research and Management. 1986;6:219–21.
Rode KD, Pagano AM, Bromaghin JF, Atwood TC, Durner GM, Simac KS, et al. Effects of capturing and collaring on polar bears: findings from long-term research on the southern Beaufort Sea population. Wildl Res. 2014;41:311–22.
Wiig Ø, Born EW, Laidre KL, Dietz R, Jensen MV, Durner GM, et al. Performance and retention of lightweight satellite radio tags applied to the ears of polar bears (Ursus maritimus). Anim Biotelemetry. 2017;5:1–11.
Mulcahy DM, Garner G. Subcutaneous implantation of satellite transmitters with percutaneous antennae into male polar bears (Ursus maritimus). J Zoo Wildl Med. 1999;30:510–5.
Wilson RR, Regehr EV, Rode KD, St. Martin M. Invariant polar bear habitat selection during a period of sea ice loss. Proc Royal Soc B, Biol Sci. 2016;283:p.20160380.
Amstrup SC, McDonald TL, Durner GM. Using satellite radiotelemetry data to delineate and manage wildlife populations. Wildl Soc Bull. 2004;32:661–79.
Henri D, Gilchrist HG, Peacock E. Understanding and managing wildlife in Hudson Bay under a changing climate: some recent contributions from Inuit and Cree Ecological Knowledge. A Little Less Arctic: Top Predators in the World’s Largest Northern Inland Sea, Hudson Bay. Netherlands: Springer. 2010:267–89.
Joint Secretariat. Inuvialuit Settlement Region Polar Bear Joint Management Plan Joint Secretariat, Inuvialuit Settlement Region. 2017: 66 pp.
Wong PBY, Dyck MG, Arviat Hunters and Trappers, Ikajutit Hunters and Trappers, Mayukalik Hunters and Trappers, Murphy RW. Inuit perspectives of polar bear research: lessons for community-based collaborations. Polar Rec. 2017;53(3):257–70.
Thiemann GW, Derocher AE, Cherry SG, Lunn NJ, Peacock E, Sahanatien V. Effects of chemical immobilization on the movement rates of free-ranging polar bears. J Mammal. 2013;94:386–97.
Laidre KL, Born EW, Gurarie E, Wiig Ø, Dietz R, Stern H. Females roam while males patrol: divergence in breeding season movements of pack-ice polar bears (Ursus maritimus). Proc Royal Soc B, Biol Sci. 2013;280:p20122371.
Sjors H. Bogs and fens in the Hudson Bay Lowlands. Arctic. 1959;12:2–19.
Ritchie JC. The vegetation of northern Manitoba V. Establishing the major zonation. Arctic. 1960;13:210–29.
Clark DA, Stirling I. Habitat preferences of polar bears in the Hudson Bay lowlands during late summer and fall. Ursus. 1998;10:243–50.
Hochheim K, Barber DG, Lukovich J V. Changing sea ice conditions in Hudson Bay, 1980–2005. A Little Less Arctic: Top Predators in the World’s Largest Northern Inland Sea, Hudson Bay. Netherlands: Springer. 2010:39–51.
Stewart DB, Barber DG. The ocean-sea ice-atmosphere system of the Hudson Bay complex. A Little Less Arctic: Top Predators in the World’s Largest Northern Inland Sea, Hudson Bay. Netherlands: Springer. 2010:1–37.
Stirling I, Spencer C, Andriashek D. Immobilization of polar bears (Ursus maritimus) with Telazol in the Canadian Arctic. J Wildl Dis. 1989;25:159–68.
Calvert W, Ramsay MA. Evaluation of age determination of polar bears by counts of cementum growth layer groups. Ursus. 1998;10:449–53.
Towns L, Derocher AE, Stirling I, Lunn NJ, Hedman D. Spatial and temporal patterns of problem polar bears in Churchill, Manitoba. Polar Biol. 2009;32:1529–37.
Heemskerk S, Johnson AC, Hedman D, Trim V, Lunn NJ, McGeachy D, et al. Temporal dynamics of human-polar bear conflicts in Churchill, Manitoba. Glob Ecol Conserv. 2020;24:e01320.
Sikes RS. 2016 Guidelines of the American Society of Mammalogists for the use of wild mammals in research and education. J Mammal. 2016;97:663–88.
R Core Team. R: A language and environment for statistical computing. Vienna, Austria.: R Foundation for Statistical Computing; 2021.
Freitas C, Lydersen C, Fedak MA, Kovacs KM. A simple new algorithm to filter marine mammal Argos locations. Mar Mamm Sci. 2008;24:315–25.
Derocher AE, Wiig Ø, Bangjord G. Predation of Svalbard reindeer by polar bears. Polar Biol. 2000;23:675–8.
Pagano AM, Carnahan AM, Robbins CT, Owen MA, Batson T, Wagner N, et al. Energetic costs of locomotion in bears: is plantigrade locomotion energetically economical? J Exp Biol. 2018;221:jeb175372.
Jonsen ID, Patterson TA, Costa DP, Doherty PD, Godley BJ, Grecian WJ, et al. A continuous-time state-space model for rapid quality control of Argos locations from animal-borne tags. Mov Ecol. 2020;8:1–13.
Boyd JD, Brightsmith DJ. Error properties of Argos satellite telemetry locations using least squares and Kalman filtering. PLoS One. 2013;8:e63051.
Langrock R, King R, Matthiopoulos J, Thomas L, Fortin D, Morales JM. Flexible and practical modeling of animal telemetry data: hidden Markov models and extensions. Ecology. 2012;93:2336–42.
McClintock BT, Michelot T. momentuHMM: R package for generalized hidden Markov models of animal movement. Methods Ecol Evol. 2018;9:1518–30.
Johnson DS, London JM, Lea M-A, Durban JW. Continuous-time correlated random walk model for animal telemetry data. Ecology. 2008;89:1208–15.
Johnson DS, London JM. crawl: an R package for fitting continuous-time correlated random walk models to animal movement data. R package version 2.3.0.2022. http://CRAN.Rproject.org/package=crawl.
McClintock BT. Incorporating telemetry error into hidden Markov models of animal movement using multiple imputation. J Agric Biol Environ Stat. 2017;22:249–69.
Bacheler NM, Michelot T, Cheshire RT, Shertzer KW. Fine-scale movement patterns and behavioral states of gray triggerfish Balistes capriscus determined from acoustic telemetry and hidden Markov models. Fish Res. 2019;215:76–89.
Conners MG, Michelot T, Heywood EI, Orben RA, Phillips RA, Vyssotski AL, et al. Hidden Markov models identify major movement modes in accelerometer and magnetometer data from four albatross species. Mov Ecol. 2021;9:1–16.
Zucchini W, MacDonald IL, Langrock R. Hidden Markov Models for Time Series. 2nd ed. Boca Raton: CRC; 2017.
Burnham KP, Anderson DR. Model Selection and Multimodel Inference. 2nd ed. New York: Springer; 2002.
Latour PB. Spatial relationships and behavior of polar bears (Ursus maritimus Phipps) concentrated on land during the ice-free season of Hudson Bay. Can J Zool. 1981;59:1763–74.
Derocher AE, Stirling I. Observations of aggregating behaviour in adult male polar bears (Ursus maritimus). Can J Zool. 1990;68:1390–4.
Lunn NJ. The ecological significance of supplemental food to polar bears on land during the ice-free period in western Hudson Bay. Master’s thesis, University of Alberta, Edmonton, Canada. 1985.
Latour PB. Interactions between free-ranging, adult male polar bears (Ursus maritimus Phipps): a case of adult social play. Can J Zool. 1981;59:1775–83.
St. Louis VL, Derocher AE, Stirling I, Graydon JA, Lee C, Jocksch E, et al. Differences in mercury bioaccumulation between polar bears (Ursus maritimus) from the Canadian high- and sub-Arctic. Environ Sci Technol. 2011;45:5922–8.
Rogers MC, Peacock E, Simac K, O’Dell MB, Welker JM. Diet of female polar bears in the southern Beaufort Sea of Alaska: evidence for an emerging alternative foraging strategy in response to environmental change. Polar Biol. 2015;38:1035–47.
Bechshoft T, Luo Y, Bohart AM, Derocher AE, Richardson ES, Lunn NJ, et al. Monitoring spatially resolved trace elements in polar bear hair using single spot laser ablation ICP-MS. Ecol Indic. 2020;119:106822.
Kaniuth K, Huber S. An assessment of radome effects on height estimates in the EUREF network. Mitt des Bundesamtes für Kartographie und Geodäsie. 2003;29:97–102.
Képeši V, Labun J. Radar signal attenuation due to finite radome thickness. NAŠE MORE: znanstveni časopis za more i pomorstvo. 2015;62:200–3.
Jonkel CJ, Kolenosky GB, Robertson RJ, Russell RH. Further notes on polar bear denning habits. Bears: Their Biology and Management 2. 1972;142–58.
Knudsen B. Time budgets of polar bears (Ursus maritimus) on North Twin Island, James Bay, during summer. Can J Zool. 1978;56:1627–8.
Lunn NJ, Stirling I. The significance of supplemental food to polar bears during the ice-free period of Hudson Bay. Can J Zool. 1985;63:2291–7.
Rode KD, Robbins CT, Nelson L, Amstrup SC. Can polar bears use terrestrial foods to offset lost ice-based hunting opportunities? Front Ecol Environ. 2015;13(3):138–45.
Øritsland NA. Temperature regulation of the polar bear (Thalarctos maritimus). Comp Biochem Physiol. 1970;37:225–33.
Hurst RJ, Øritsland NA, Watts PD. Body mass, temperature and cost of walking in polar bears. Acta Physiol Scand. 1982;115:391–5.
Hurst RJ, Leonard ML, Watts PD, Beckerton P, Øritsland NA. Polar bear locomotion: body temperature and energetic cost. Can J Zool. 1982;60:40–4.
Cherry SG, Derocher AE, Thiemann GW, Lunn NJ. Migration phenology and seasonal fidelity of an Arctic marine predator in relation to sea ice dynamics. J Anim Ecol. 2013;82:912–21.
Yee M, Reimer J, Lunn NJ, Togunov RR, Pilfold NW, McCall AG, et al. Polar bear (Ursus maritimus) migration from maternal dens in western Hudson Bay. Arctic. 2017;70:319–27.
Togunov RR, Derocher AE, Lunn NJ, Auger-Méthé M. Drivers of polar bear behavior and the possible effects of prey availability on foraging strategy. Mov Ecol. 2022;10(1):50.
Stirling I. Midsummer observations on the behavior of wild polar bears (Ursus maritimus). Can J Zool. 1974;52:1191–8.
Stirling I, Latour PB. Comparative hunting abilities of polar bear cubs of different ages. Can J Zool. 1978;56:1768–72.
Dyck MG. Characteristics of polar bears killed in defense of life and property in Nunavut, Canada, 1970–2000. Ursus. 2006;17:52–62.
Wilder JM, Vongraven D, Atwood T, Hansen B, Jessen A, Kochnev A, et al. Polar bear attacks on humans: implications of a changing climate. Wildl Soc Bull. 2017;41:537–47.
Durner GM, Douglas DC, Nielson RM, Amstrup SC, McDonald TL, Stirling I, et al. Predicting 21st-century polar bear habitat distribution from global climate models. Ecol Monogr. 2009;79:25–58.
Molnár PK, Derocher AE, Thiemann GW, Lewis MA. Predicting survival, reproduction and abundance of polar bears under climate change. Biol Conserv. 2010;143:1612–22.

No competing interests reported.

Download PDF

Journal Publication

published 15 Jul, 2024

Read the published version in Animal Biotelemetry →

Editorial decision: Revision requested
21 Feb, 2024
Reviews received at journal
21 Feb, 2024
Reviews received at journal
29 Jan, 2024
Reviewers agreed at journal
15 Jan, 2024
Reviewers invited by journal
15 Jan, 2024
Editor assigned by journal
11 Jan, 2024
Submission checks completed at journal
11 Jan, 2024
First submitted to journal
09 Jan, 2024

You are reading this latest preprint version

Telemetry without collars: performance of fur- and ear-mounted satellite tags for evaluating the movement and behaviour of polar bears

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Materials and Methods

Study Area

Tag Design and Application

Polar Bear Capture and Tag Deployment

Data Preparation & Statistical Analyses

Results

Discussion

Declarations

Funding

Author Contribution

Acknowledgments

Availability of Data and Materials

References

Additional Declarations

Status:

Journal Publication

Version 1