Polar Biology

, Volume 38, Issue 8, pp 1301–1304

Longest recorded underwater dive by a polar bear


    • Department of Biological Sciences and Wildlife Research Division, Environment Canada, Department of Biological SciencesUniversity of Alberta
  • Rinie van Meurs
Short Note

DOI: 10.1007/s00300-015-1684-1

Cite this article as:
Stirling, I. & van Meurs, R. Polar Biol (2015) 38: 1301. doi:10.1007/s00300-015-1684-1


The maximum dive duration for a wild polar bear (Ursus maritimus) of any age is unknown, and opportunities to document long dives by undisturbed bears are rare. We describe the longest dive reported to date, by a wild undisturbed adult male polar bear. This dive was made during an aquatic stalk of three bearded seals (Erignathus barbatus) lying several meters from each other at the edge of an annual ice floe. The bear dove for a total duration of 3 min 10 s and swam 45–50 m without surfacing to breathe or to reorient itself to the locations of the seals. The duration of this dive may be approaching its maximum capability. Polar bears diverged from brown bears (Ursus arctos) about 4–500,000 years ago, which is recent in evolutionary terms. Thus, it is possible that the ability to hold its breath for so long may indicate the initial development of a significant adaptation for living and hunting in its marine environment. However, increased diving ability cannot evolve rapidly enough to compensate for the increasing difficulty of hunting seals because of the rapidly declining availability of sea ice during the open-water period resulting from climate warming.


Polar bearUrsus maritimusDive durationHuntingAquatic stalk


There are few published records of dive durations for undisturbed polar bears. Russell (1975) documented incidents of polar bears apparently hunting swimming sea ducks (Somateria spp.) by diving and attempting to seize them on the surface of the water after approaching briefly beneath the surface. Actual durations under water were not recorded but were reported to be brief. Stirling (1974) reported polar bears diving for 34, 40, 55, and a maximum of 72 s to feed on kelp. Lønø (1970) reported polar bears diving to depths of 3–4 m to retrieve “seaweed,” but no dive durations were given. Dyck and Romberg (2007) observed a subadult polar bear diving for fish in a river near its mouth and staying under water for between 3 and 29 s on a total of 17 dives. Over the past 10 years, we have observed both adult male and female polar bears in Svalbard, diving to feed on the remains of whales in a few meters of water, for untimed durations estimated to range between about 15 and 30 s. In this paper, we document the longest reported underwater dive of a wild undisturbed polar bear, which far exceeds anything previously observed.


Observations were made from the bridge of the MV Lars in the drifting pack ice to the north of the island of Spitsbergen in the Svalbard Archipelago while searching for polar bears to photograph. Viewing was done with 10 × 50 binoculars and recorded on two different video cameras.


In mid-afternoon, August 19, 2014, one of us (RVM) observed a large adult male polar bear (Ursus maritimus) lying and looking around on a large flat annual ice floe several hundred meters across at 80′ 23.46° N, 013′ 13.09° E, 20 km north of the north coast of Spitsbergen, the main island in the Svalbard Archipelago. The adult male bear was exceptionally thin (Fig. 1): body condition one out of five, the lowest category (Stirling et al. 2008). Shortly after first approaching the ship, the bear noticed three adult bearded seals (Erignathus barbatus) lying several meters apart about 0.5 m from the edge of a smooth annual ice floe, about 150–200 m away across open water. When he reached the edge of the floe he was on, he slipped into the water and began to swim toward the seals on the surface before diving and swimming under water in an aquatic stalk (Stirling 1974; 2011 pp. 174–175).
Fig. 1

Adult male in poor body condition that made the long duration dive during aquatic stalk (© R. van Meurs)

From shortly after the bear began to swim toward the seals, to the end of the hunting sequence, two video records were made, one of the entire sequence and one of the first two-thirds. From these videos, distances between the bear and the first seal, distances between seals, and the duration of his dive could be determined. Adult bearded seals are 2–2.5 m in body length (Kovacs 2009), and the ship was approximately at right angles from the edge of the floe the seals were on so that fairly accurate measurements could be taken of the distance between the bear and the closest seal when he first dove, as well as the distances between all three seals. In addition, the complete hunting event was continuously observed by RVM, the ship’s captain, and some passengers using 8–10 X binoculars. The distance from the bear was estimated to be 80–120 m, in almost calm water, under a thinly overcast sky but with bright light conditions.

When the bear was about 35–40 m from the first seal, it dove with the rump clearly visible as it disappeared beneath the water. The closest seal was lying with its head down, facing in the direction from which the bear was approaching, across the wind. Six to seven seconds later, the seal raised its head and became vigilant and, 20 s after the bear first dove, continued to look from side to side. Thirty-two seconds after the bear first submerged, the seal suddenly dove into the water. The bear was not seen though a small swirl in the water, possibly from his body or rear end being close to the surface, was noted momentarily. No sign of the nose or head, indicating that the bear might have taken a quick breath before continuing, was noted by any of the observers. Subsequent careful inspection of the shorter video, which included the location where the bear was submerged, confirmed that the bear did not surface. During close observations of 27 aquatic stalks, made through a telescope under good viewing conditions, Stirling (1974) noted that when a bear undertook an aquatic stalk, it could surface, breathe, raise its head slightly to look if necessary, and submerge without significant disturbance in the water surface. However, such behavior was always clearly visible through the telescope, even at distance. Thus, although the bear was close to the surface at the time the first seal dove, it was clear that it did not surface to breathe.

The second seal was approximately 12–13 m further along the edge of the floe from the first. The second and the third seal had their heads raised after the first seal went into the water, but neither appeared vigilant while the bear was continuing to approach under water. They appeared to be exhibiting the normal periodic head-up position of a seal on the ice looking about to simply check for potential predators on the ice. After being under water for 3 min 10 s, the bear exploded out of the water at the floe edge and propelled itself halfway onto the ice immediately in front of the second seal. The seal spun rapidly, thus preventing the bear from getting a grip on its head or front quarters which might have enabled him to prevent the seal from escaping. The bear appeared to seize a hind flipper of the seal briefly as it dove into the water, but could not restrain it. He then rested his left foreleg on the surface of the ice and panted rapidly at a rate of about 2 breaths/s for about 20 s, after which he began slow normal swimming along the floe edge and looking around but no longer appeared to be breathing heavily.

However, the 3-min 10-s duration of this long underwater swim to stalk a bearded seal far exceeds anything previously reported.


Besides documentation of the long duration of the dive, other interesting behavioral and possibly evolutionary aspects of interest arise from these observations. The fact that the bear did not surface when the first seal fled into the water, but continued 12–13 m further to exactly where the second seal was, suggests he had a good spatial sense of his location in relation to at least the first two seals at the beginning of the dive. As well, he showed the ability, while remaining below the surface, to extend his underwater stalk to the location of the second seal without needing to breathe again or reorient himself.

Taxonomically, Ursidae (bears) as a group are considered terrestrial carnivores that largely feed on vegetation (Ewer 1998). However, the polar bear is considered a marine mammal because it depends almost completely on the arctic marine ecosystem to survive and spends as much of its life as possible on the sea ice where it hunts for food (Stirling 2009). Estimates of when polar bears diverged from brown bears (Ursus arctos) range up to 4–5 mya (e.g., Miller et al. 2012), while more recent analyses indicate that separation of these species occurred much more recently, about 343–478 thousand years BP (Liu et al. 2014). The ease with which polar and brown bears interbreed and that their offspring are fertile are supportive of the more recent estimates of the time differentiation. However, even in that evolutionarily short period, selection pressure has been sufficiently intense to facilitate rapid evolution of several morphological traits for hunting seals, including white pelage, large feet for swimming and walking on thin ice, longer neck and skull, teeth and skull more suited to a carnivorous diet rather one consisting largely of vegetation, a deeper rostrum with more upwardly positioned eye orbits that appear to improve forward vision while swimming, and genes that have been positively selected to favor adaptation to a high-lipid diet (Figueirido et al. 2009; Slater et al. 2010; Cahill et al. 2014; Liu et al. 2014).

The extent to which selection for survival in the marine environment may have also improved the ability of the polar bear to prolong dive duration as an adaptation for hunting involving diving is unknown. However, like cetaceans, pinnipeds, and sea otters, polar bears have a reniculate (lobulated) kidney (Ewer 1998), although with far fewer lobules than either cetaceans or pinnipeds (Makita et al. 1998; Berta et al. 2008). This type of kidney morphology, absent in most terrestrial mammals, increases the surface area of the filtering mechanisms so that metabolic byproducts can be removed from the body more efficiently than by a non-lobed kidney (Berta et al. 2008). The fact that the polar bear we observed was able to remain under water so long and then apparently physiologically recover from the dive in little more than 20 s, before continuing to swim and hunt, suggests that besides external morphological adaptations to its marine environment, the species may show evolutionary adaptations to facilitate extended diving. In comparison, the sea otter (Enhydra lutris), which evolved from terrestrial otter stock about 4 mya (Koepfli et al. 2008), also has far fewer lobules than cetaceans or pinnipeds (Berta et al. 2008) but, despite its small size (50+ kg), maximum dive times of 31 individuals in the wild ranged between 162 and 422 s (Estes et al. 2009).


The maximum possible dive time for polar bears of any size is unknown. However, we speculate that an adult male in better physical condition, with more overall muscle mass than the individual we observed, might be able to store more oxygen before submerging and thus be able to remain under water for a longer period than the dive we report in this paper. In comparison with our unpublished observations of fatter, possibly more buoyant bears engaged in shorter but untimed dives in past years, the thin bear described here showed no apparent difference in the ease with which it was able to remain under water. Dive times of smaller bears would likely be less than those of adult animals. As the climate continues to warm and the amount of sea ice in the circumpolar Arctic continues to decline (Stroeve et al. 2012), access to seals on ice floes may become progressively more difficult owing to the presence of more water between floes and earlier seasonal disappearance of ice altogether. Although our observations demonstrate polar bears possess strong subsurface swimming and navigating abilities, previous observations confirm that the success rate of aquatic stalks of seals haul out on the ice surface where they are vulnerable to predation from the water is low (Stirling 1974). Therefore, we suggest that polar bears cannot evolve increased diving ability rapidly enough to compensate for greater difficulty of hunting seals because of the rapidly declining availability of sea ice during the open-water period resulting from climate warming.


We thank Derryn Cooper and Petrina Steains for permission to use their video recordings of the aquatic stalk observed and Captain Mike Donnithorne for skillful piloting of M. V Lars to facilitate clear observations of this event. We thank Oceanwide Expeditions, Quark Expeditions, and One Ocean Expeditions for their support of our participation on arctic ecotourism expeditions where we have been able to observe undisturbed polar bears in their natural environment in Svalbard. We thank Dr. Andrew E. Derocher for constructive criticism of an earlier draft of this paper.

Copyright information

© Springer-Verlag Berlin Heidelberg 2015