You've got to hand it to coyotes in the North. They're a tough lot; so tough that some Alaskan coyotes will even eat porcupine -- and lots of it -- if they have to.

Given the choice, northern coyotes prefer to eat snowshoe hares, and when bunnies are abundant, they are definitely the coyotes' meal of choice.

But northern hare populations crash dramatically about every ten years, and when this happens, coyotes in the Central Alaska Range significantly increase the amount of porcupine in their diets.

Laura Prugh made this discovery while doing research in a rugged area south of Fairbanks. While hare numbers were declining the UBC doctoral student tracked coyotes, poked through their kills and looked at their caches over the span of more than three years.

She also collected a lot of coyote scat -- about 1500 over the course of the study. After washing and analyzing them, she found a surprising 25-fold increase in the amount of porcupine in the scats.

This was not what Prugh had expected to find when she started her work. Coyotes also like to eat Dall sheep lambs, and biologists suspected that coyotes might be eating more lambs when hare numbers are down, possibly helping to drive down the total number of sheep.

Sheep populations in the Alaska Range and Kluane's Ruby Range have declined by about 60 percent since the 1970s, and no one is actually sure why. Alaskan biologist Steve Arthur has been monitoring the sheep population in the Central Alaska Range, putting radio collars on 20 to 25 lambs every year.

He found that lamb survival actually increased as much as three fold when the hares declined. Most likely, this is due to the fact that both coyotes and golden eagles -- the main predators on Dall sheep lambs -- have fewer or no young when the hare cycle is near its bottom, so there are not as many predators eating lambs.

Prugh's supervisor is Dr. Charles Krebs, who headed up the Kluane Boreal Forest Ecosystem Project. His research team found that when bunnies are less abundant, coyotes in Kluane eat a lot more voles.

They even resort to eating vegetation such as juniper buds, grass and needles, but porcupines do not seem to be on their menu. "The coyote scats in Kluane had no porcupine in them," says Prugh.

Prugh and her assistants spent a lot of time figuring out the densities of small mammals such as ground squirrels, snowshoe hares and voles in the Alaskan study area. They counted 64,838 bunny pellets, and marked and recaptured scores of ground squirrels and voles.

But she can only guess how many porcupines there might be in the area. "We only saw four of them in three years. I have no idea of the numbers."

She did find evidence that porcupines are indeed a risky prey for coyotes. "In 2001 we collected scat from coyote dens and one of them was in an area where there was no access to sheep. The coyotes' diet was 47 percent porcupine and the mom died from injuries from the quills. She had a stomach full of quills."

Strictly in terms of biomass, porcupines are not a bad option. "One porcupine equals 350 voles or seven hares. So the porcupine, if you can get it, is a really nice package."

The Alaskan coyotes also ate more voles when the hare numbers were down, but Prugh found that carrion was consistently the Alaskan coyotes' preferred alternative prey.

Unlike Kluane, the Central Alaska Range has abundant numbers of both moose and wolves, so the coyotes there often benefit from dead animals killed by wolves.

Prugh says it is too early to tell why Dall sheep numbers have declined in the North. While coyotes and golden eagles prey on lambs, wolves kill adult sheep. Other factors could also be at play such as changes in climate and hunting pressure.

"The records for Dall sheep are not that long, and I do not think there is reason to think that their populations are still on a downward trajectory. It looks like they may be wavering now at a lower equilibrium point."

For the next phase of her research, Prugh is gleaning even more information from her collection of coyote scats. By scraping them, she can collect intestinal cells and genetically identify individual coyotes, and figure out what they are eating.

"From my point of view, I'm just finding out more about how the system works, which can help with management in the area."

For more information, Laura Prugh can be contacted at prugh@zoology.ubc.ca or visit her website at www.zoology.ubc.ca/~prugh/.