Oakley lives with her husband and two daughters in Haines Juction (pop. 800) which is nested in the Shakwak Valley, shown above.

I'm working with a small herd of isolated and genetically unique woodland caribou that range on the southern Yukon/Alaskan border, about 200 miles north of Haines Junction (where I live). It's called the Chisana herd, and its numbers have dropped from 1,800 in the late 1980s to about 300 now. Most of the remaining animals are very old, at the end of their lifespan, and the population faces extinction.

People out on the land first noticed problems with the Chisana herd. The local big game outfitter reported the decline to Yukon's Fish and Wildlife Branch, and we began to look at the herd more closely. Native people from the White River First Nation and Kluane First Nation also knew something was wrong with the herd. The reason for the decline seems to be almost no calves have survived, apparently because predators are eating 90 to 95 percent of them in the neonatal period, or first week or two of life. At this rate, this herd is calculated to disappear within three to four years.

Michelle Plantinga Oakley '92 grew up in Munster, Indiana. She first went to the Canadian Yukon as a field assistant on an arctic ground squirrel ecology study when she was an undergraduate zoology major at Michigan. There, she met her husband to be, Shane Oakley, a local forest fire fighter. She received her doctorate in veterinary medicine in 2000 from Atlantic Veterinary College (University of Prince Edward Island), then trained in zoo/wildlife medicine at the Calgary Zoo.

I was charged with the task of coming up with a plan for captive breeding at a game farm in the Yukon; that is, to take 20 animals and breed enough for reintroduction five to seven years later. Although captive breeding has saved a lot of endangered species, it comes with some problems and won't work for every situation. Some of the problems I was considering were:

· It takes a minimum of five to seven years to breed enough caribou to release back into the wild, but in this case there would be no wild herd left by then!

· We risk reintroducing diseases/parasites with these animals—whatever they may have picked up on the game farm or captive breeding facility. With testing, treatments and quarantines, this is a somewhat manageable risk, but it is expensive and you never reduce those risks to zero.

· When you take 10 to 20 animals for captive breeding, that's the extent of the genetic diversity you're ever going to have. All the rest of the herd's genetics are lost, so you put that herd into what is called a "genetic bottleneck." That means they can be more susceptible to disease and less able to adapt to environmental challenges.

· Since there won't be animals left when we are ready to reintroduce captive-bred animals, the reintroduced animals, after being raised on a farm, would have no wild caribou to show them the ropes. They would not know the migration patterns, the best places to find food, the best ways to avoid predators and so on. These herds have a huge communal knowledge that is lost when the wild herd is lost, and a naive animal can learn it only if enough of the wild herd is left to teach it.

Cabin home of the rescue team. Oakley's duties involve 'saving caribou and neutering wolves,' because wolves and bears, though not specially numerous, for some undetermined reason have been killing Chisana calves as fast as the caribou can birth them.'

Because of these problems with captive breeding, I came up with an idea to temporarily pen some of the caribou on their range just before and during the time of calving, from late March to late June. Then we don't interfere with breeding (and genetics), we don't risk introducing disease and we don't have to worry about reintroducing the animals years later. After ten weeks of holding them and protecting their calves from predators, we let them go. The calves are about three to four weeks old then, and at that age their ability to survive and avoid predators is excellent.

I approached our caribou biologist Rick Farnell with the idea last summer, prefacing it with, "This is going to sound a bit crazy, but . . . " and he agreed it sounded pretty "out there" but also like it could work. The idea is not entirely new, as other endangered species recovery programs have used a similar approach—building what is called a "predator exclosure" with sea turtles and various other threatened bird species when predation was seriously affecting their offspring/nests. To our knowledge, no one had tried it before in the wild with large mammals like caribou.

A cow seems almost transparent in the Yukon mist. Colleagues credit Oakley with the idea of the 'captive-rearing' technique. So far, it's working so well that it's greenlighted for next year and fundraising is well under way.

I recently talked to a reindeer herder in Finland who said some of them deal with predation problems on their free-ranging reindeer herds this way. They bring the herd into fenced areas by their homes, where they can guard them and scare off predators. Our situation is a bit more difficult, however, because we are in a very isolated area, making the logistics of getting out there and living for a few months pretty tough. And our animals are wild.

Once under way with funding, equipment and crew, we built the pen near the outfitter's camp, which he lent to the project. It has a few log cabins for us to live in—no electricity or running water, but we are on a lake, have woodstoves to keep warm and satellite phones for communication.

You can reliably access this camp only by air. We snowmobiled in and out through the mountains (takes about 1.5 hours each way) a few times in early March, when three other people and I built the 20-acre caribou pen. But you have to travel on rivers, and as it warms up here the thawing ice gets too dangerous for snowmobiles.

'Two big tundra swans lift off from Tchawsahmon (Tu-wash-amon) Lake. Bird species abound in the region, and mobs of ravens sometimes distract cows while others blind and maim their calves.

Construction was an incredible challenge. The weather was very cold for March. Most mornings were -22 F, and one morning was -58. It would warm up to about +5 degrees during the day. We made the caribou pen out of Typar geocloth, a black tarp material that is used to line roads for construction. It is 7 feet tall and hung on 1/8-inch-steel cable wrapped around tops and bottoms of trees for fence posts.

Caribou don't challenge fences like deer and moose will, and the opaque blackness seems to keep them from trying to go over the barrier. Wolves also seem to stay away from the fence, but we electrified it because in addition to wolves, there are grizzly and black bears, wolverine, coyote and lynx in the area.

We also walked the perimeter a few times a day, hoping that our presence, along with the electricity, would keep predators from trying to break in. A big grizzly came running down the lake ice one evening, nose in the air following the caribou smell. He charged as soon as he saw us but ran away when we shot cracker shells to scare him off. We saw him pass by again a week later, and we also found large bear tracks a few feet away from one of the cabins one morning, but he did not cause any problems. I suspect he encountered the electric fence.

Rescuers blindfolded 20 pregnant cows, tied them in deer bags and helicoptered into the 125-acre enclosure. Each cow received a blood test, an ultrasound and a radio collar. Handling time per animal was limited to 30 minutes to minimize stress.

My original idea was to herd the caribou using caribou fences similar to what native people began using here thousands of years ago. These were fences made of brush that gently pushed caribou a certain way, usually into a narrowing V passage. Near the narrow point of the V, there would usually be snares and hunters ready for the concentrated animals. But we decided that, in this pilot project year, we would just shoot a net over the animals from a helicopter (called net-gunning) and fly them to the pen.

We captured 20 pregnant cows over three days in late March. Everything went smoothly. I used a portable ultrasound to check each female for pregnancy and started them out on a diet of reindeer pellets. The caribou adapted to our presence amazingly quickly. Within two days they stopped looking for a way out. At first they would trot away when we entered their pen to fill up the feed troughs, but within a week they just looked up when we came in and then went back to whatever they were doing. Within 10 days all ate regularly from the troughs.

I thought it would take longer for them to adjust to a new diet because there is a fair bit of natural feed in the enclosure, but they really watch and learn from each other. As soon as a few of the dominant ones began eating at the troughs, the others came over and gave it a try. We also picked a lot of lichens (their favorite food—caribou candy!) with the help of Yukon school kids. We mixed lichens in with the pellets to tempt them to the troughs and also to help the bacteria and protozoa in their rumen (the first division of their stomachs) slowly adjust to the new diet. If we changed the diet too quickly, the caribou could get bloated or diarrhea, and even die from serious cases. Fortunately, we had no such cases, probably because we provided lichens and because they had so much natural forage in the pen to snack on.

The Yukon Fish and Wildlife department, where I've been a wildlife biologist and veterinarian for three years, is leading the caribou project. But we've had help from people from the Canadian Wildlife Service and the White River First Nation. We also get funding from the Yukon Fish and Wildlife Management Board, and just recently the World Wildlife Fund. Since the herd ranges in Alaska, too, we've been trying to pull Alaskan agencies into our recovery effort. They are slowly coming around, thanks to the efforts of a few dedicated biologists there.

Yukon schoolchildren picked hundreds of bags of licen for the caribou. Lichen ('caribou candy,' Oakley calls it) is mixed with commercial food pellets to tempt the caribou to the troughs and to ward off digestive problems their new diet could set off.

The Chisana is the only woodland caribou herd in Alaska. Alaska's other herds are barren ground caribou. But the area where the herd ranges is in Alaska's Wrangell St Elias National Park, and thus far, the park has taken the stance that the decline is "natural" and therefore should not be interfered with. We do not know the ultimate cause of the decline, however, so we are very hesitant to declare it "natural" at this point.

Our latest count showed 14 of the 17 calves were still out there doing well. That's an 82 percent survival rate, excellent compared with the 6 to 10 percent survival we have in calves born out of the pen. Although I loved observing the caribou close-up, I was relieved and hopeful watching them run from the pen that day, back to the wild where they belong, taking much needed youth and vigor back to the herd.

We're going ahead with the project again next spring, but with 40 to 50 caribou this time. So we'll be out enlarging our fence in early fall and start our captures next March and early April.

Dr. Michelle Oakley received her BSc in zoology from U-M in 1992. She is the first veterinarian to be a regional biologist in Canada's Yukon Territory. She works for the Fish and Wildlife Branch of Canada's Department of Environment and welcomes inquiries from students interested in classroom projects. Her e-mail address is Michelle.Oakley@gov.yk.ca.