ISLE ROYALE NATIONAL PARK, MICH. — Rolf Peterson held up his arm for silence and pointed through the thick brush.

A hundred yards off the trail, a female moose sporting a shiny new mahogany winter coat was knee-deep in muck, munching on plants. She raised her head nonchalantly, then flicked up her ears and froze as she spotted observers. After a long minute, she plodded up toward firmer ground. A calf popped out of the brush and trotted after her.

Research happens up close in the world's longest continuous study of predators and prey at Isle Royale National Park in Lake Superior. Peterson has been watching and counting moose and wolves in this wilderness off Minnesota's North Shore for nearly 40 of the study's 51 years, in summer by foot and in winter by air.

Now that continuity is at a breaking point. The island's moose population is nearing a 50-year low, and what's bad for the moose is worse for the wolves that depend on them. Peterson can see the day when the wolves die out on Isle Royale, and scientists must confront far-reaching questions: Should we intervene to help the wolves survive, or let them die out and start again? What role should humans play to preserve an ecosystem?

Disrupting the extraordinary research has ramifications far beyond the wolves on Isle Royale.

Through the years, the study has provided unprecedented information about how long wolves live in the wild, and how much prey they kill. For wildlife managers around the world who want to reintroduce wolves into an ecosystem, as they have in the Yellowstone National Park area, the answers emerging from the Isle Royale research have been crucial to their efforts.

Big decisions will have to be made. "The risks for wolves seem to be pretty large and growing," Peterson said.

Peterson and his wife, Candy, live and work out of a one-room log cabin built by Jack Bangsund, a Norwegian bachelor fisherman, in 1931, the year Isle Royale became a national park. A Norwegian flag flutters outside the cabin in his honor. Under the cabin's slanting floor, otters have hollowed out space to sleep and squabble. Loons call from the waters of Rock Harbor just beyond the rickety dock.

The lifestyle marries the pioneer with the 21st century. There's no running water, and the outhouse is in the back. A sleek, vertical Finnish wind generator and a pair of solar panels power a couple of light bulbs — and the computers hooked via satellite dish to the Internet.

Under a ceiling papered with moose posters, family photos and island maps, Peterson has based his summer field work since 1970 when he joined the wolf-moose study as a 21-year-old graduate student.

Now 60, Peterson has retired from a teaching career at Michigan Technological University, but has no plans to give up this study, the love of his life. Trim and fit, he backpacks and bushwhacks through alder and hazelnut with the energy of someone half his age. He climbs a spindly ladder to the roof of an old fire lookout tower, grasping a portable antenna to check signals from radio-collared wolves. It's an iconic scene of a field scientist in a goofy sunhat leaning into the wind, listening for beeps and watching a meter with experienced blue eyes.

Isle Royale is an ideal outdoor laboratory. About 22 miles off the shore of northeastern Minnesota, its remote location makes it difficult to reach even in summer, when most visitors arrive by boat. Its terrain is a rugged mix of parallel valleys and rocky ridges, as if a giant claw scraped across its surface, as a glacier did 8,000 years ago. With no roads, no hunting and almost no human presence, it's a magnet for scientists to study plants and wildlife.

The moose-wolf study began in 1958. Durward Allen of Purdue University spent a decade learning how the two species seemed to be in balance, with each growing or declining at approximately the same rates. But that neat balance soon got messy, and the populations spiked and plunged over the decades with changes in snow depth, vegetation, diseases, parasites, higher summer temperatures and shorter winters.

By studying how much wolves prey on moose, said John Vucetich, research co-leader with Peterson, scientists can help answer a longstanding question of interest to sportsmen. "When we live with wolves, wherever it may be, we have to be concerned about the fact that they eat animals that humans also like to hunt" such as elk and deer, Vucetich said.

"In terms of the value of that study to science, nothing else rivals it," said Mark Romanski, biologist and acting chief scientist at Isle Royale National Park.

The park service contributes about $36,000 annually to Peterson's research, the National Science Foundation funds $90,000, and individuals and a small endowment provide about $25,000.

Peterson's field station, while visually idyllic, doesn't always smell so good. Part of his research is to collect the carcasses of wolves and the skulls and other body parts of moose. A hand-made sign in front of bones collected in 2009 explains why studying the bones is important: "When they get old, moose exhibit arthritis, periodontal disease, osteoporosis. Sound familiar?"

Wolf skeletons became especially significant after a Swedish researcher discovered three years ago that Isle Royale wolves have a spinal deformity that's a sign of inbreeding, Peterson said. The park service is planning a scientific review.

One day in September, Peterson cooked the carcass of a female wolf that died trying to give birth to eight pups last spring. He wore protective gloves and glasses as he lifted the skull from a boiling cauldron. He plunked the grisly mess on a picnic table, wrinkled his face at the odor, and tried to pull muscles, tendons and tongue away from the skull.

"Not done enough yet," he said, returning it to the cooker.

The death of this one female meant the collapse of an entire pack — one of four on the island. The pack had lost several members in recent years, and by early 2009 contained only one male and one female. She and her unborn pups were the pack's last hope for regeneration.

Wolves have struggled to survive since the 1980s, when a domestic dog brought illegally to the island spread canine parvovirus. Their numbers dropped from a peak of 50 to 14 in two years, and have never fully recovered.

The current populations of about 24 wolves and 530 moose are close to what they were 50 winters ago, Peterson said, but each is at risk.

The wolves have lost more than half of their genetic variability compared with wolves on the mainland, he said. Moose are being stressed by higher summer temperatures that cause them to eat less, produce less fat for winter survival and die of starvation, Vucetich said.

Moose are also plagued by life-threatening ticks that have multiplied in the rising temperatures.

"Climate change is certainly an additional pressure, and it's the big one," Peterson said.

So far the Park Service tradition has been not to intervene. "Our policy is basically to let nature take its course," Romanski said.

But the prospect of the wolves' extinction is forcing a new look at that policy.

"It's not just a science-based decision," Romanski said. "It's a blend of science and stewardship and conservation that will have to be addressed."