Uncovering the secrets of polar bear dens

There’s still plenty to learn about the most vulnerable days of a polar bear’s life

One of the many remarkable aspects of polar bear behaviour is the fact that these huge carnivores begin life weighing no more than two pounds, born inside dens that their mothers have carved out of snow.

Almost as remarkable is the fact that, despite being created by fully grown, heavily pregnant bears, most dens are — from a human perspective — “claustrophobic,” says Geoff York, Polar Bears International's senior director of research and policy.

York examined a number of dens in Alaska in the 2000s when working as a researcher with the U.S. Geological Survey with Dr. Steven Amstrup, who later led the Polar Bears International (PBI) research team for more than a decade and is now PBI’s chief scientist emeritus.

Early research in Alaska

During those early days of den research, Amstrup, York and colleagues would use data from bears that had been fitted with radio collars to infer both the locations of dens and when they had been vacated. Then they would visit the sites, locate the dens and — after thoroughly checking to confirm they were empty — crawl inside for measurements. Narrow tunnels, several feet in length, would lead to small den areas “just large enough for a 300- or 400-pound polar bear to turn around,” says York. It was, he muses, always quite the sight when a female that had not yet fully vacated her den would return and “jog along and then dive nose-first into this tiny hole, her butt sticking out until she wriggled inside.”

No matter how cramped the experience, being inside a den is “pretty special,” says York. “All the walls have claw marks, intermixed with little pieces of the female’s hair that have come off through abrasion.” Throughout the winter, the female is not only nursing her cubs and keeping them warm, but also actively maintaining the den. “In all dens I've entered, the female has clearly spent time scraping ice buildup from the walls and ceiling of the den — presumably to regulate airflow.”

The Alaska den studies led by Amstrup followed on the heels of pioneering work by Dr. Thor Larsen in Norway, Dr. Ian Stirling in Canada and Dr. Stanislav Belikov in Russia. They focused initially on understanding fundamental basics, such as den location and structure. “We were trying to see, for example, are the dens always on the east side of a slope, or is there some other characteristic that would increase the predictability of where they might be?” explains York. 

Amstrup was the first scientist to document that as many as half the bears in the Southern Beaufort Sea denned on the sea ice, the only population known to do so. By the early 2000s, however, Amstrup and team had confirmed that they largely no longer did so: The sea ice in the region was now too thin, too active and broke up too early in the year for denning on the ice to be anything other than a foolhardy option.

The shift from denning on the ice to denning on land shows the rapid pace of change in parts of the polar bears’ range, and in the Southern Beaufort Sea, it presented another issue. With the entirety of the population now denning along the coast of northern Alaska, how vulnerable were pregnant and nursing mothers to disturbance from the oil and gas industry that dominates areas of Alaska’s North Slope?

Monitoring dens with cameras

Studies to determine the vulnerability of denning polar bears to disturbance in Alaska began by placing recording instruments in artificial dens to gain a sense of how strongly different sounds and vibrations were transmitted into them. That was then superseded by observers in tents watching dens and emerging bears directly — although it soon became clear that the observers were being observed by the bears as much as the other way around. Finally, says York, the need to monitor the bears with as little possible disturbance, “led to the development of what we called cooler cameras, where we put a mini cam into an insulated cooler with a bunch of car batteries, and we'd go set it up and let it go to work.”

The den-cam monitoring work in Alaska lasted nearly a decade and yielded valuable data. It was led by Dr. Tom Smith of Brigham Young University, with funding and field support from PBI. In 2020, York, Smith and Wesley Larson published a study which showed that bears in Alaska seemed highly reluctant to leave their dens even when disturbed by human activity. They found that, during the approximately two-week period in spring when polar bear mothers and cubs begin to emerge from the dens, the mothers in particular showed overt responses to most human activities, particularly low-flying aircraft; their reactions were, however, less than expected, suggesting that a mandated one-mile buffer zone around the dens was proving somewhat effective. However, they noted that they could only document overt responses, and that they didn’t have data to assess whether or not bears were stressed in ways that did not obviously manifest.

While the study confirmed that protective buffer zones are important, the research team realized that, in order to protect dens, managers first need to know where they are located, not an easy task for dens hidden under the snow. For more than a decade, the oil and gas industry had been relying on forward-looking infrared radar (FLIR) to locate dens, a technology also developed by York and Amstrup while at USGS. But a 2020 study led by Smith and coauthored by Amstrup and York showed that those surveys failed to find up to 55 percent of known dens — which, given bears’ seeming reticence to abandon those dens, may make them especially vulnerable to heavy equipment or high-intensity disturbance. 

In response to FLIR’s high failure rate in finding dens, PBI scientists began studying higher resolution synthetic signature aperture radar (SAR) and wide-band radar as more accurate tools, conducting tests in Alaska and Svalbard, Norway. Research on this promising technology will hopefully be finalized in the coming years.

Den research in Svalbard

In 2016, PBI began conducting an annual den-emergence study in Svalbard, using far more advanced versions of the cameras that Smith and team had used years ago in Alaska. Powered by solar panels and with digital hard drives rather than tapes that must be swapped out periodically, the cameras could be set up while bears were in their dens, left unattended and retrieved long after the bears had left, ensuring minimal to no disturbance.

PBI’s Svalbard Maternal Den Study, conducted in cooperation with the Norwegian Polar Institute and the San Diego Zoo Wildlife Alliance, ran through 2025. It yielded rare footage of moms and cubs emerging from dens, and revealed a great deal of variation in how long they spent outside the den before popping back in for safety. The study also measured how quickly the den was abandoned altogether as the family left for the sea ice.

While not yet a quantifiable trend, one thing the PBI teams conducting the studies noticed is that bears in Svalbard may be leaving dens earlier and earlier in response to warming conditions.

“I think the last two years, we arrived in Norway a full week earlier, and we still struggled to get out and stay ahead of the females exiting their dens,” recalls York. “We also saw record warm temperatures and even rain in February and early March. So to some extent, it's not surprising bears might be drawn out of their dens earlier in those conditions. It will be interesting to see what the long-term research shows.”

Further possible areas of study include denning areas in Manitoba or Ontario, where conditions are rapidly changing and where little is documented about den emergence behaviour. In a world where both climate warming and human activities are increasing in the Arctic, it is clear that ongoing den studies are necessary to understand and protect polar bears during this fascinating but vulnerable period in their lives.