Scientist Responds To Misleading Polar Bear Coverage

september sea ice NSIDCCiting a recent study by the government of Nunavut in Canada, conservative media are claiming that the number of polar bears is “increasing.” The takeaway, according to these media outlets, is that concerns about the fate of polar bears in a warming world are overblown. But polar bear scientist Steven Amstrup says these commentators are mistaken.

The polar bears located west of the Hudson Bay are one of 19 polar bear subpopulations, and one of 8 subpopulations that are thought to be shrinking, according to a comprehensive review conducted in 2009. (One population was found to be increasing, three are stable, and there isn't enough data to assess the other seven). Amstrup and others previously analyzed bears captured from 1984-2004 and found that the West Hudson Bay population declined from 1,194 in 1987 to 935 in 2004.

But a new survey by the government of Nunavut, a largely Inuit territory in Northern Canada, puts the population size as of last August at 1,013, according to a widely circulated article in Canada's Globe and Mail. This new estimate is derived from a plausible range of 717 to 1,430 bears and, importantly, comes from an aerial survey, unlike the previous studies which involved capturing and recapturing bears.

Amstrup said media outlets claiming the aerial survey shows an increasing population are mistaking a single point estimate for a trend. “The population size is just a number. It is a valuable number to have, but from the standpoint of population welfare, it is the trend in numbers that is critical,” he wrote in an email. Because previous estimates used a different methodology, and covered a different geographic area, they cannot be easily compared to the latest figures, contrary to the media narrative. When the aerial survey is repeated in later years, it will then be able to tell us more about how the population size is changing. In the meantime, the Canadian government is expected to release its latest capture-recapture data next month.

Population estimates are used to determine how many polar bears can be killed each year. Hunting polar bears is a significant source of income among the Inuit, who have been skeptical of dire predictions of popopulation decline.

Amstrup emphasizes that “in the bigger picture, whether any one population is currently declining, stable or increasing is beside the point,” adding, “it is criticial to remember that our concern about polar bears is focused on the future.” The scientists who spend their lives studying polar bears have been unable to envision how the population numbers can withstand the long-term decline of the sea ice.

More detailed responses from Amstrup below:

Does the survey show the population is stable?

No it doesn't. This survey produced a single population estimate relevant to the time the survey was conducted. A single point estimate of population size says nothing about whether the trend is up down or stable. Trend can be addressed by multiple “point” estimates collected over time. Therefore, if we had two aerial survey population estimates conducted in different years, they might suggest the population trend (increasing, decreasing, or stable) between those years. Because of sampling error etc. and because the confidence interval (the statistical range of possible population size estimates generate by this survey) on this population estimate is wide (717-1430) having at least 3 such estimates conducted over multi-year intervals would be the best way to assure a meaningful trend is estimated. This one estimate, however, can say nothing about the population trend.

Also, this new aerial survey estimate cannot be compared to previous estimates, in order to estimate a trend, because it has used a different methodology and covered different geographic areas. Unfortunately, the press coverage of this survey has made the serious mistake of trying to make such comparisons. The previously published population estimates were derived by capture-recapture analysis. Even if the new aerial survey focused on exactly the same geographic area, it would not be surprising to derive a slightly different population estimate. It is important to remember that both methods provide estimates and not exact counts, and there is always some error and uncertainty in estimates (one expression of the uncertainty is the confidence interval described above). Each kind of estimate has its own potential biases and error rates, and to calibrate the relationship between them would require multiple simultaneous comparisons. You would expect the outcomes of the different methods to be close, but different point estimates would not be unexpected.

Also, and perhaps more importantly, the recent aerial survey covered geographic regions both north and south of the area of focus of the previously published capture-recapture estimate. It would be reasonable to expect that there would be a different number of bears occurring in a larger geographic area than in a smaller one. So, for both methodological and geographic reasons, this new estimate cannot be compared to previous estimates to assess a trend in numbers.

What does this survey say about the future of polar bears?

Because this survey produced a single population estimate for a particular point in time, it says nothing about the future of this population or of polar bears in general. Polar bears depend for their survival on adequate access to sea ice. It is only from the surface of the sea ice that they consistently can catch the seals on which they depend for nutrition. Planetary physics require that the world will continue to warm as long as greenhouse gas concentrations in the atmosphere continue to rise. There is absolutely no uncertainty that the world will warm without mitigating the rise in GHG concentrations! And, a warmer world will mean less sea ice. So, as I (and many of my colleagues) have predicted, all polar bear populations ultimately will decline and disappear if we do not reduce our emissions.

Now, the decline in sea ice and in polar bear numbers and distribution will not be a smooth curve. The natural chaos in the climate system means that there will be hump and bumps in these declines. And, there will be surprises in the greenhouse! Some populations may respond positively at first to a warming world. All polar bear populations, however, (even those that see a transient benefit from warming), ultimately will suffer from climate warming. There are no data anywhere in the range of the polar bear to suggest they can survive in anything like current distribution or numbers without adequate access to sea ice.

In Hudson Bay, the long-term trends in the important population parameters of body condition, reproduction and survival are all down. It is biologically impossible for a population to be experiencing declining physical condition and survival and yet to remain stable in numbers. Even if the adult animals can hang on and survive through tough times, the failure of new animals to come into the breeding population ultimately will spell decline.

The previous capture-recapture methods allowed scientists to not only estimate numbers, but also to keep track of changes in condition and survival. Aerial survey can provide an estimate of numbers, but does not provide these important other pieces of information. As I pointed out previously, therefore, one aerial survey does not allow an estimate of trend.

This new aerial survey does, however, include a piece of information relevant to trend. Of the 701 polar bears actually counted during the survey, only 22 (or about 3%) were yearlings. This is a very low percentage of yearlings (in Alaska during the good ice years of the 1980s, about 15% of the animals observed were yearlings). If that 3% figure is even close to the number of surviving yearlings that are out there now, it is not at all clear to me how the Hudson Bay population could be sustaining itself. This observation is very much in line with the previously published indications that survival (especially of young) is declining.