Lance Craighead and Doug Ouren are standing on a rugged knob in southwestern Montana looking for grizzly 416. They aren't following the bear's dagger-clawed trail. They can't see its silver-tipped mane. They're just listening to a steady beeping on their tracking receiver.
"That's a strong signal," Ouren says. He checks the frequency and then his notes: Grizzly 416 is a young adult female, fitted with a radio collar. Craighead scours the receding ridgelines with field glasses. To the east a wooded valley rises to the gilded peaks of the Gallatin Range, with Yellowstone National Park just beyond. To the west a hard, vertical hike leads to the Madison River Valley, the home of Ted Turner's Flying D Ranch. Although not strictly part of the park, this is still the Greater Yellowstone Ecosystem, winter home for thousands of elk and at least three collared bears. When Craighead was a teenager, he used to hike through this country with his father and his uncle, two of Yellowstone's renowned grizzly researchers. Now he is a bear biologist too. He runs his father's institute in Bozeman, Montana, and collaborates with Ouren, who is a wildlife researcher for the U.S. Geological Survey (USGS). Compact and curly-haired, Craighead looks like an older, bigger brother of Frodo Baggins and has some of the same earnest good nature. He never flaunts his legacy, but when pressed he admits that this work has become personal: "I sort of feel responsible for the bears in Yellowstone." At first glance, the bears hardly seem to need his help. Some 500 live in this ecosystem, while another 800 get by in the northern Rockies—a remarkable comeback from the 1970s, when the Yellowstone grizzly was declared a threatened species (see "A Symbol Besieged," below). But, today's populations can't compare with the tens of thousands that roamed the Lower 48 when Lewis and Clark passed through. And they are still under siege. To the north, Gallatin County is filling up with new residents and ranchettes. Gateway communities such as West Yellowstone and Gardiner are booming. In 2001, 20 bears were killed in and around Yellowstone by hunters, automobiles, and wildlife managers. Two years ago, grizzly reintroductions in the Salmon-Selway-Bitterroot region were cancelled after Idaho governor Dirk Kempthorne took the plan to court. "I oppose bringing these massive, flesh-eating carnivores into Idaho," the governor declared. "Whenever there's an encounter between a human and a grizzly bear, the human does not fare well." The Idaho decision was a setback for the grizzly. Big animals need big country, and Yellowstone National Park just doesn't measure up. After a century of isolation and a major population crash, the park's grizzlies are genetically vulnerable. They need to mix with other bears across the country to preserve their genetic diversity. With each new road and subdivision along its borders, America's first national park is turning into a genetic prison.
A grizzly's habitat, represented by a geographic information system, or GIS, is composed of multiple layers of information—from food to ground cover to road locations—all superimposed on a digital map.
To keep grizzly 416 out of harm's way, Craighead and Ouren need more than just a vague idea of her home range. They need a precise record of her movements. They need to figure out which parts of the landscape matter the most to her—her hangouts, her aversions, her sense of place. Radio tracking is one solution. But Craighead also uses a far more sophisticated tool, one that can take the data transmitted by 416's collar and merge them with countless other bits of information. That technology is known as a geographic information system, or GIS.
The idea behind GIS is simple: It combines a database with a mapping program to build up digital maps layer by layer. A map of Yellowstone, for instance, might start with state lines, park borders, and other municipal boundaries. Snap on an elevation layer—contour lines—and the landscape jumps into relief. Add trails, roads, and forests and a more complex map begins to emerge. The ability to create such maps on demand is already an enormous boon to biologists, but it's just a start. To estimate how many bears an area can support, for instance, biologists can use GIS to map all the animals' potential food sources—from pine nuts to cutthroat trout—and their availability at different times of year. They can layer on the angle of the sun in autumn, where the bears hole up in the winter, where the sedges grow, and how deep the snow cover lies. What was a two-dimensional map now becomes a three-dimensional world, or even a four-dimensional one, in maps that show how the landscape changes over time. GIS was first conceived in the 1960s, at Harvard University and Canada's Land Inventory program. Because it juggles tremendous amounts of data, researchers needed almost a decade to create a working system, using million-dollar mainframe computers. Processors have since grown cheaper and exponentially more powerful, but GIS still struggles to keep pace with our evolving planetary surveillance. The first Landsat went into orbit in 1972, sparking a proliferation of space-based imaging. The 24-satellite Global Positioning System went online in 1995, allowing precision mapping on the ground. Between remote sensing and GPS-based monitoring, GIS now has the power to organize, analyze, and display every piece of data on the planet. The effect on ecology has been nothing short of revolutionary. Instead of spending thousands of hours tracking animals on foot or in small planes, field researchers now use radio-tracking collars that digitally store GPS readings. After a few months, a timer triggers the collar to fall off, the researcher collects the collar, and the data are downloaded directly into a GIS program. The result is a nearly seamless record of an animal's movements, tracked with satellite precision. Chuck Schwartz, head of the Interagency Grizzly Bear Study Team for the USGS, can remember when he kept track of animal populations with color-coded pushpins. Now he can easily print out digital maps that are more detailed by a few orders of magnitude. "We started seeing movements of bears—day, night, with people, without people, that were very, very different," he says. "It started explaining a lot of the basic science." Given enough data, some researchers say, GIS could describe everything that goes on in an ecosystem. "I don't know if that's ever going to happen," says Charles Convis, who runs the conservation program of GIS software giant ESRI. "But what's intriguing is the idea that GIS will solve ecology the same way that genetics solved organism biology and that calculus solved classical mechanics."
Lance Craighead, a third-generation wildlife biologist, has seen his field grow ever more tech savvy. "We have to go with the best science we have," he says.
Wildlife researcher Doug Ouren has radio tracked more than 30 bears, each one for close to a year. "One of the frontiers of spatial analysis is looking at things as they move across the landscape," he says. Photographs by Brent Humphreys
Sitting in his Bozeman office one morning, Craighead calls up a three-dimensional map on his computer monitor, then rotates it, contorting its features. This is the Madison River Valley, seen through the eyes of a bear. The contours and colors signal degrees of difficulty for a grizzly on the move. Purple and flat mean safe, easy going; steep orange and red signal danger. The result is a landscape turned upside down: The peaks are mostly flattened and purple, while the towns and highways in the valleys occupy the high ground, tinted orange to red. To build this map, Craighead began by cutting the landscape into a grid with cells 100 meters square. The first layer was ground cover. Agriculture and urban areas were shaded as dangerous, pine forests and subalpine meadows as habitable. Because bears like edges between habitats, the GIS took each square, found the nearest edge, and ranked it by distance and desirability. Next came layers denoting building density, roads, and trails. The map shows plenty of bear-friendly country, but the purple is chopped up by orange hazard lines: roads, towns, subdivisions. Craighead knows that Yellowstone's grizzlies can survive only if the park is somehow connected to the wildlands that surround it. The Salmon-Selway-Bitterroot region is only 185 miles to the west in Montana and Idaho, the northern Continental Divide only 155 miles to the north, and both areas are good bear habitat. By placing a conservation easement here, closing a Forest Service road there, Craighead can create a network of protected land that allows bears to mix genes across the landscape. "We need just a couple of small pieces," he says. "The habitat is already there. What we're trying to do is keep it from getting blocked off." He estimates this can be done by protecting as little as 2 percent of the privately owned land. But finding that 2 percent requires GIS. Grizzlies are at their most vulnerable in the winter, when they go into hibernation. Their dens are typically in lonely country—in high-altitude forests or far up remote canyons—but as hikers and snow machines move deeper into the wilderness and grizzlies recolonize land around the park, isolated spots are harder to find. Three years ago, Chuck Schwartz and his team loaded a quarter-century's worth of den-site data into a GIS program and analyzed the relationships between elevation, slope, sunlight, and forest cover. They then projected that information across the entire system to find favorable denning areas. Several national forests in the area then used this model to determine where snowmobiles should not be allowed. Snowmobiles and inbreeding aren't the only things threatening the grizzly. In much of the West, the whitebark pines whose nuts the bears depend on are battling a deadly Eurasian blister rust. Forty-two percent of the whitebark pines in western Montana have died in the past two decades, and 89 percent of the surviving trees are infected. Yellowstone's cutthroat trout, meanwhile, are being displaced by lake trout, which don't spawn in streams and therefore can't be caught by bears. Wildlife managers can use GIS to map these changing food resources and thereby predict the bears' movements: If it's a good year for pine nuts, the bears stay higher in the mountains, out of people's way; if it's a bad year, they have to venture farther for food. Everyone in the greater Yellowstone area—from the Forest Service to the Department of Transportation to Gallatin County planners—is now thoroughly dependent on GIS, Craighead says. It allows them to speak the same spatial language, and with it, to shape their world.
Last November, Craighead and his wife, April, who is also a biologist, had their first child. Given baby Willow's family history, she could well end up studying the Yellowstone grizzlies. If so, her field trips 20 years hence might go something like this: Stepping out of her truck near the Taylor Fork, Willow puts on special sunglasses that project GIS data onto her field of vision. When she looks through the lenses, the world takes on a virtual tint. Habitat layers peel away and geologic data hover forward; digital animal tracks wind through the trees, and underground springs send cobalt tracery along the ground. Willow checks for moisture stress in the trees and follows the boundaries of a proposed clear-cut. As she moves into elk territory, the rangeland flushes from green to pink to bright magenta, depending on how many elk the land can support under different management scenarios. And all of this happens in real time, in the field. When Willow wants to know where a certain grizzly is hanging out, she dials in her Interagency clearance code, and a flashing marker lights up ahead of her. She can even compare this bear's movements with those of its grandmother, grizzly 416. For now, though, 416 is still young and active. As the sun fades over the Taylor Fork, Craighead and Ouren swap hunting tales and stories about Bozeman and bears while their tracking receiver beeps away in the background. When the evening chill sets in, they descend to the valley floor and cross a stream, where the beeping suddenly stops. A quick line-of-site triangulation reveals that the signal has probably been blocked by a nearby ridge and that the bear is resting no more than a quarter of a mile away. Directly below her is a hunters' camp. Ouren investigates and finds a grill that reeks of meat. Tomorrow he'll ask the local manager to look in on it. But he and Craighead won't bother the bear. Sometime next summer, her collar will fall off, and Ouren will go find it. Then he'll know where 416 has been, and the real work will begin. He hopes the data will show her meandering through the outer reaches of the Taylor Fork. Maybe she'll have cubs and lead a more cautious life, as grizzly mothers do. Maybe she'll sniff out another dirty grill and wind up dead. Can people coexist peacefully with "massive, flesh-eating carnivores"? GIS alone can't answer that question. It combines detailed data on bear behavior with a comprehensive inventory of the landscape. But even the best virtual bear is a far cry from 600 pounds of true ursine fury. Computer models are driven by data choices and software limitations. Grizzlies are profoundly reliant on smell, for example, for which biologists have no data. "GIS is just a simplification of biological complexity," Craighead says. It may have the capacity to solve ecological problems the way the discovery of DNA solved the riddle of inheritance. But solving isn't the same as saving. As Craighead's uncle, John, put it 25 years ago: "We need the wisdom to apply space science, courage to act in the interest of endangered species, and foresight to properly manage a rapidly changing national landscape." Otherwise, a virtual bear in a simulated wilderness may be all we'll have left.
A Symbol Besieged
Yellowstone National Park is a landscape of legend, and the grizzly bear its royalty. More popular books and articles have been written about local grizzlies than any other animal, and they are the most intensely studied bear population in the world. The Craighead family laid that scientific foundation. Frank was a Forest Service entomologist who pioneered Everglades conservation. His sons, identical twins John and Frank Jr., pursued wildlife careers in tandem. In 1961, when miniature radio was in its infancy, they put the first VHF radio-tracking collar on a grizzly, working with electrical engineers to design their own equipment and lugging it into the wild. The ability to track an animal across rugged country opened a startling view into the bears' natural history and transformed wildlife biology. The National Geographic Society publicized their research in print and on film and turned the Craigheads into scientific celebrities. Then in 1967 came a disastrous coincidence: On the night of August 13, in Glacier National Park, two women camped miles from each other were mauled to death by two different grizzlies. Both bears, dependent on garbage from dumps for part of their diet, had grown familiar with human food before the maulings. So Yellowstone officials decided to close the park's dumps. The Craigheads argued that the closures should occur gradually, and that the bears should be fed elk and other animal carcasses until they adjusted. The park forced the bears to go cold turkey instead. The Craigheads had spent a decade collecting the most detailed grizzly data available, but the park authorities ignored them. When the brothers spoke out, officials gradually pushed them out of Yellowstone. The incident is finally fading from institutional memory, but one thing is undisputed: The bears lost. The Craigheads' census data suggests that there were 230 grizzlies in Yellowstone before the dumps were closed. In the decade that followed, bear after bear sought food from humans and was either killed in the process or hunted down. More than 200 grizzlies were removed from the Yellowstone ecosystem, and a small but healthy bear population slipped into crisis. In 1975 the Yellowstone grizzly was listed as a threatened species.