At their 2005 meeting, the last with published proceedings, the experts had an interesting discussion about what kinds of estimates should be including in a "status" table describing the polar bear population.

"There was considerable debate and disagreement about the use of point estimates for subpopulations that we really do not know much about," said the proceedings.

"There was further discussion and some agreement that there should be separate columns for estimates based on science and estimates based on something else," it said.

"Something else"? What "something" other than science might a polar bear scientist be tempted to use in estimating the polar bear population?

Well, "TEK" for one. "There was some confusion as to how the estimates based on TEK were derived," said the proceedings.

Now, "TEK" may look like an acronym for something technical -- but it isn't. It is "traditional ecological knowledge" -- or what local people say about polar bears. In some cases, researchers run "simulations based on the minimum size necessary to support local knowledge of subpopulation trends," said the proceedings, noting that estimates like these "have unknown and in most case inestimable errors."

In the end, the scientists wisely punted on presenting a conclusion about the polar-bear population trend for six of the 19 subpopulations they now believe exist, saying the data was "deficient" in these areas. "No subpopulation inventories have been conducted in East Greenland and therefore the size of the population is not known," said their population "status table," for example.

However, the report did say: "The total number of polar bears worldwide is estimated to be 20,000-25,000."

Almost 50 years ago, polar bear scientists first estimated there might be as many as 19,000 polar bears. Three years ago, they estimated there might be as few as 20,000.

Who knows? By the time the government decrees polar bears not merely threatened, but endangered, there may be as few as 30,000 roaming the Arctic wastes.