A Colorado State University statistician ended a years-long competition in London this month when his statistical model of whaling populations tied in an international contest.
Geof Givens, the only U.S. scientist to enter the competition sponsored by the International Whaling Commission, will now have the algorithm further scrutinized to ensure that it designates safe quotas of bowhead whales that may be hunted by native Alaskans without endangering the whale population. It has already bested all but one of the 13 procedures submitted by scientists around the world.
"If this model is adopted, every few years Eskimos will ask for a quota (for bowhead whales) and this computer algorithm will allow the IWC to set safe quotas," Givens said.
With some 40 nations as members, the IWC sets policy for cetacean (whale and dolphin) management. The commission has banned commercial whaling since 1986, although controversy exists as Japan and Norway take some whales through loopholes in the agreement.
Givens, also a member of the U.S. delegation to the IWC Scientific Committee, must balance the complex intermixture of ecological, economic, cultural and political concerns over whaling by the aboriginal cultures living adjacent to far northern waters. Some natives have been hunting whales for thousands years. In a harsh environment, these people rely on whales as a source of food, and they see whale hunting as part of their culture.
Givens’ model focuses only on the bowhead whales taken by native Alaskans. These communities have financed extensive research and data collection on the local bowhead stock, including dangerous whale censuses conducted from the edge of constantly shifting pack ice.
"The goal is to produce a statistical model that limits the effects of traditional whale hunting carried on by aboriginal peoples along the Alaskan seacoast by allowing some hunting, while ensuring that the whale population continues to grow" he said. The challenges have been, first, how to model that population, and second, how to test the algorithm’s effectiveness itself.
The difficulty for Givens is to combine estimates of the number of bowhead whales in the Alaskan area (which the IWC puts at about 8,500), then to factor in such variables as the whales’ productivity and age at sexual maturity, their survival rates, their age distribution, the number captured and widely varying scenarios about what may happen in the future.
In an attempt to make it the most powerful and accurate, Givens’ algorithm strategically tries to combine the best features of his competitors’ algorithms rather than modeling whale biology from scratch.
Now that they’ve been chosen (the tying model was submitted by a team of Icelandic scientists), the models will be put through more rigorous testing to determine which one is safest and best meets the IWC’s management goals in a vast array of future scenarios, Givens said.
A final decision will be made at an IWC meeting in Japan next year.
Givens said his interest in environmental issues blossomed during a summer spent in Yellowstone National Park, and his subsequent doctoral dissertation, completed at the University of Washington, involved statistical work that formed a background for the present model. Givens began his preparations for the whale management contest in 1994 with the assistance of several graduate students.
While the modeling is referred to as a competition, he said, there is camaraderie between the competitors worldwide.
"I’m not going to make any money on this," he said. "But I’m a very competitive person, and I hope to be able to give the other contestants a friendly ribbing next time I see them."