YELLOWKNIFE - Caribou herds aren't shrinking; scientists are just looking in the wrong place.

That's the opinion of some aboriginal elders and at least one outfitter determined to expose what he calls a fraud.

"This is a deliberate hoax, not good science," writes John Andre, owner of Montana-based Shoshone Wilderness Adventures, which operates commercial hunts in the Lac de Gras area of the Northwest Territories, northeast of Yellowknife.

Andre has maintained for years that official estimates of caribou populations have been manipulated. Government biologists, he says, have deliberately created the impression of declines by reassigning caribou from established herds to newly created herds.

Andre points to the Bathurst herd, which he says was divided in two in 1996. Animals calving on the east side of Bathurst Inlet were allocated to the so-called Ahiak herd, for which no population count exists.

That created an apparent drop in the Bathurst herd, Andre suggests.

"If you had five herds in 1986 and nine herds in the same geographic area 20 years later, you can't compare one herd with another over that same time ... without re-combining the herds you have split," he writes a in paper entitled "The Great Crashing Caribou Hoax."

He maintains a large part of the Bathurst herd simply shifted its calving ground east.

"The Ahiak herd are Bathurst caribou and they must be counted as such."

Some aboriginal elders also wonder if caribou haven't simply evaded researchers, said Fred Sangris, who heads the caribou committee for the Dene Nation.

"They're still trying to determine if those numbers are true," he said.

"If there's 100,000 caribou missing in a herd, then where are they? There's no evidence of caribou carcasses or anything on the land."

Andre has been persistent enough to convince the N.W.T. government to ask the Alberta Research Council to review caribou research by the territory's biologists. The council didn't reject the findings, but it did suggest some conclusions were based on sketchy data.

Caribou herds are indeed defined by their calving grounds, but government biologists deny that the Ahiak herd is a recent invention.

"The evidence for an (Ahiak) calving ground goes back at least to the 1970s," said Anne Gunn, a retired N.W.T. biologist who continues to study the herds.

Gunn theorizes that Ahiak animals have taken over territory once occupied by the shrinking Bathurst herd.

Scientists also doubt that caribou are migrating from one herd to another. Data from radio collars suggest that fewer than five per cent of cows switch herds, and they are just as likely to switch in one direction as another.

Finally, while they acknowledge there's no real estimate of the Ahiak herd size, the data that does exist isn't encouraging. Yearly flights over the same 10-kilometre stretch of the herd's range found 40 per cent fewer animals in 2009 than in 2006.

Still, Andre calls the assertion that caribou herds are diminishing a deliberate attack on both his business -- caribou tags issued to outfitters have declined from a high of 1,700 to only 503 this fall -- and all industry in the North.

"It's a simple strategy," he writes.

"Divide the herds and compare old herd definitions with the new ones. Frighten the First Nations groups that the caribou are in danger and they will shut down new development every time."

Sangris said the Dene just want to be sure.

"It's a great mystery and everyone's trying to find out what happened to these herds."