Polar bear (Ursus maritimus)

Female polar bear with cub on sea ice. The cub is jumping from one ice floe to another.
Photo: Janne Schreuder / Norwegian Polar Institute

The polar bear in the Barents Sea lives on disappearing sea ice. Climate warming is the main threat to polar bears. The polar bear habitat is changing rapidly, and the Polar Basin could be ice-free in summer within a few years. Gaining access to preferred denning areas and their favourite prey, ringed seals, depends on good sea ice conditions at the right time and place. The population probably increased after hunting was banned in 1973, but little is known about the present population trend.

What is being monitored?

Polar bear dens and sea ice cover

Loading chart ...

The figure shows the number of days when the sea ice cover around the island of Kongsøya from 1 October31 December (left axis) exceeded 60% and the number of dens the same winter (right axis). Dens are counted by surveys on the ground (over several weeks) or from a helicopter (1–2 days). The coverage in the study area has varied from one year to another. Years without surveys (e.g. 2014) have no bar. The number of days when the sea ice cover exceeds 60% has a strong influence on the number of females that den on Kongsøya; few days with sufficient ice cover leads to few females entering dens.

Loading chart ...

The figure shows the number of days when the sea ice cover around the island of Hopen from 1 October31 December (left axis) exceeded 60% and the number of dens the same winter (right axis). Dens are counted from helicopter in spring the same winter. The number of survey days has varied from one year to another. Years without surveys (e.g. 2014) have no bar, except for 2012, when no dens were registered on Hopen during the count. The number of days when the sea ice cover exceeds 60% has a strong influence on the number of females that den on Hopen; few days with sufficient ice cover leads to few females entering dens.

Loading chart ...

The figure shows the number of days between 1 October and 31 December that had a sea ice cover exceeding 60% (within four 25×25 km pixels) in and around 5 major denning areas in Svalbard (Hopen, Kongsøya, Storfjorden, Sjuøyane and Nordaustlandet). The annual median value is represented by a line within each box, while box segments and lines over and under the median show the other data points for each year. There has been a decrease in the number of days with the given sea ice cover for all 5 areas over time (all p-values < 0.05). Fewer than 30 days with a 60% sea ice cover or more means that pregnant females will struggle to get to the denning area in time to breed, especially to the more distant islands such as Hopen, Kongsøya and Sjuøyane.

Recruitment

Loading chart ...

Average litter size (COYs, cubs of the year); data from the annual capture-recapture programme 1993–2015. There is a non-significant trend of decreasing litter size over time.

Loading chart ...

The figure shows the number of cubs of the year (COYs) per adult female based on data from the annual capture-recapture programme in 1993–2015. The red line shows the trend over time, and there is no significant change (p = 0.08). A major part of the interannual variation, however, is explained by variations in the Arctic Oscillation (AO) in spring (April–June) the preceding year. Higher values of AO correlate with lower cub production the year after (p < 0.01).

Loading chart ...

Proportion of females with cubs of the year (COYs) and yearlings, based on data from the annual capture-recapture programme in 1993–2015. The blue dotted line describes the trend in the proportion of females with COYs over time, and there is no statistically significant change (p = 0.09). A large part of the interannual variation can, however, be explained by variations in the climatic index Arctic Oscillation (AO) in spring (April–June) the preceding year. Higher values of AO correlate with a lower proportion of females with COYs (p < 0.01). The observed interannual variation in the proportion of females with yearlings cannot be explained by either time trend or variations in the AO. The error bars show 95% confidence intervals of the annual estimates.

Body condition

Loading chart ...

Body condition index of adult male polar bears caught in spring (March–May) in 1993–2015.

The lines in the middle of each box show the median value, and the box segments and lines above and below the median each cover ca 25% of the data points. Circles show extreme values. There is no statistical significant trend over time. However, a major part of the inter-annual variation can be explained by variations in the Arctic Oscillation (AO). High values of AO (milder weather) at the time of capture correlate with a lower body condition index (leaner animals) (statistically significant, p = 0.001).

 

Status and trend

A period of intensive hunting in and around Svalbard began about 1870, and on average some 300 bears were harvested annually in the following 100 years. The population was consequently at a low level when it was protected in 1973.

In 2004, it was estimated to number between 1900 and 3600, which means that it is not threatened by the effects smaller populations may be impacted by, such as loss of genetic diversity or random demographic processes. Furthermore, genetic studies show that exchange between neighboring subpopulations to the west and east is large.

Lack of earlier estimates makes it impossible to say anything certain about historical trends, but various types of data indicate that the subpopulation grew rapidly following its protection and into the 1980s. Using demographic data, it has been suggested that the subpopulation also grew up to the turn of the century. Even though the loss of sea ice has been marked around Svalbard in recent years, and is expected to continue in the coming decades, the size of the subpopulation may still be below the carrying capacity. It is therefore not possible to say whether the subpopulation is still growing, even though the availability of habitats has become poorer for much of the year.

The occurrence of dens on the islands of Hopen and Kongsøya clearly show that few females reach there in the autumn if the ice comes late, sometime after the first part of November. Whether this means that the proportion of females in the subpopulation having cubs is declining, or they den elsewhere, is not clear.

The number of days with sea ice around all the 5 most important denning areas has experienced a dramatic negative trend since 1979. In some years, the ice has come late in all the areas the same year, and such years can prove challenging for females who have spent the summer in the pack ice to reach one of the relevant areas.

Data from the annual tagging programme indicate that there is no significant decrease in the production of first-year cubs over time, but also that there is a link between reproduction and the Arctic Oscillation (AO) index. The results indicate that in years following mild spring seasons fewer cubs are born, compared to years after cold spring seasons. Mild winters typically have less sea ice than warmer winters. Polar bears feed intensely on ringed seals in the fast ice habitat in spring.

Production of one-year-olds may be a better parameter to judge reproduction because they have survived their first year and are more likely to reach weaning age. There was no trend over time or connection with the AO which succeeded in showing effects on the production of one-year-olds.

The links between the AO and body fitness in males show that milder weather in spring corresponds with poorer fitness, but there is no clear time trend.

Causal factors

The availability of sea ice has always varied a great deal from year to year and through the seasons. The changes in large parts of the Arctic towards the period with ice-free water becoming longer and longer are explained by a milder climate with especially large changes in the Arctic. With current levels of greenhouse gas emissions, and the forecasts for the coming decade, it is expected that the areas around Svalbard will experience particularly large changes.

The links between the AO, reproduction and fitness, as well as the availability of sea ice around denning areas in the autumn, show that a milder climate dictates where females can den and this may affect the fitness of the polar bears. Observations so far are, however, unable to document that changes in the climate have had clear effects on the subpopulation. Habitat availability may still be good enough to be able to maintain a population at the current level or higher. Declining trends in reproduction or fitness will not be explainable just by variations in climate impact or corresponding trends in sea ice distribution.

Density-dependent responses on reproduction are not an unlikely future scenario and may become prominent with time due to the growth in the subpopulation following its protection in 1973. Such density dependence will also be expected to increase if habitat availability decreases, but not before the population size approaches the carrying capacity.

Very high levels of several fat-soluble substances in the Barents Sea area make it likely that such effects make themselves felt at the population level, but it is very difficult to quantify how important they are.

The large yearly variations in the availability of sea ice make it possible to study the effects of habitat changes more directly, and may eventually enable the effects of density-dependence, habitat changes and other pressure factors to be separated from one another.

Consequences

The polar bear is an important species in the arctic ecosystem. Changes in their population size and the way bears use their habitat, which can be expected to be particularly large if changes in the availability of sea ice continue, will therefore be likely to have consequences beyond the changes for the species itself.

The polar bear is entirely dependent upon seals that live in the ice, and will have an impact on these species, at any rate locally. When the sea ice does not freeze in the fjords in Svalbard before late winter, ringed seals may, for lack of snow to make lairs, give birth to their pups directly on the ice. If, in addition, the areas of sea ice are limited, the pups will quickly be taken by bears, foxes and gulls. Over time, it is conceivable that poor seal recruitment leads to less food being available for polar bears.

Perhaps more importantly, less ice makes it difficult for the polar bear to hunt. More bears on land in years with little ice also increases predation on eggs and birds in several areas, and may thus also influence ecosystems on land.

It is impossible to say anything now about exactly how changes in the number of bears and available habitats will affect the ecosystems, but a total loss of sea ice will necessarily also lead to a loss of the ecosystems that depend upon it.

Changes in the distribution and availability of sea ice are already leading to great changes in the distribution of polar bear dens. Historically important denning areas, especially on the more peripheral islands, may conceivably become unsuitable in the future. Increased use of alternative areas will also lead to greater changes in habitat. This might impose a larger predation pressure on potential prey species in new areas.

About the monitoring

The polar bear is monitored because it is Red Listed and is vulnerable to pressures deriving from climate change and environmental pollutants. It is also a species that has great appeal so that knowledge about how it is being affected will more easily lead to enhanced focus on the effects of climate change, pollutants and other pressures linked with human activity which may harm the ecosystem in the Arctic.

The most important human-induced pressure factor for polar bears is the loss of sea ice due to changes in climate. To be able to detect the effects of this as quickly as possible, the population monitoring of the polar bear is focusing on reproduction and body condition.

An aerial survey to estimate the size of the shared Norwegian-Russian Barents Sea subpopulation was carried out in 2004, and the resulting estimate was 2650 animals +/- ca 30%. A new survey of the Norwegian part of the population was conducted in August 2015, and the results should be ready before the end of the year.

Places and areas

Relations to other monitoring

Monitoring programme
International environmental agreements
Voluntary international cooperation
Related monitoring