Until a few months ago, the border between Ecuador and Peru was missing a piece. For years the two countries skirmished over a 50-mile strip in the Condor Mountains. Ecuadorians wanted to be able to send goods down the wider, navigable rivers to the Amazon and on to the Atlantic Ocean. Peruvians said the land belonged to Peru. Official U.S. maps didn't recognize either border.

Add to the mix oil fields in Ecuador, near one of the most biodiverse forests on Earth, and now everyone is consulting a map—American and European oil companies, NASA scientists, even a team of geographers from Chapel Hill.

The geographers aren't looking for oil. They're looking at what the discovery of oil has done to a virgin section of rainforest, now crisscrossed with roads and dotted with homesteads and cattle. In short, they're mapping a human footprint. But their maps don't come in the two-dimensional variety with straight political lines. They're sophisticated, computer-derived maps that reveal the big picture—using Geographical Information Systems (GIS).

"We're trying to understand land-use change—site conditions, road accessibility—in a way that arcs across social, biophysical and cultural domains," Stephen Walsh says. The professor of geography led a group of graduate students and Carolina Population Center staff to northeast Ecuador in February. Power boats stowed with video cameras, yards of film, and surveying equipment carried one more item down the Napo River. It looks like a cell phone, and without it even NASA's best satellite images of the river valley don't look like much more than grainy, although pretty, pictures.

The yellow hand-held device connects researchers to the global positioning system, or GPS. Geographers locate their position on the ground by picking up high- frequency radio waves off of orbiting satellites. "The digital imaging satellite only knows roughly where its sensor was pointing when it collected the imagery," says geography graduate student Joe Messina. "By looking at the imagery without any sort of ground truth whatsoever, we really don't know field conditions or locations." Once researchers collect coordinates, then the satellite images line up in geographic space.

Two days before leaving, Messina and fellow graduate students are busy packing. Messina, Greg Taff, and Gabriela Valdivia make sure their GPS units are set to go.

"Gabriela, did you take your malaria pill today?" Messina asks. She makes a note to herself. Valdivia, who is Peruvian, says she's also steeling herself for a meeting with Ecuadorian border guards. The first time she visited the site, they stopped her for questioning. The group plans to fly into Ecuador's capital, Quito, then fly or hop a bus for the 10,000-foot drop to Lago Agrio in the Amazon.

"It takes eight hours just to go fifty kilometers." Messina says. "I should just bring my mountain bike."

For his dissertation research, Messina is building a computer model smart enough to predict landscape changes over time. He says so far no one has built a model like his, especially for deforestation in the Amazon. New settlers are changing what was virgin landscape into a rural landscape. One way to measure change is to find out, for instance, how much of each Ecuadorian frontier farm is cleared.

"What's interesting from an economic and human standpoint—and from an environmental one as well—is really how much of the farm do they chop up? How much do they let reforest? Or how much do they reforest with another type of plant, like coffee, the primary cash crop?"

With rates of change in hand, Messina will instruct a computer to go back and mimic the development of the Napo and Sucumbios provinces from the first oil drilling. If done correctly, the model should match the actual growth on the ground. Let the model keep running and it will predict the future. He says soon cellular automaton-based models like this will be created by geographers elsewhere. One reason is that the computer systems theory required to analyze many layers of data has become readily available.

"Geographic databases tend to be large," says Rhonda Ryznar, assistant professor of planning. "We can do the kinds of spatial analysis that before we had computers was a very tedious process."

In the last five years, geographers—even real estate agents—have added GIS to their PCs. After the Pentagon, McDonald's may have the most-advanced GIS service in the world, used to triangulate the best location for a Big Mac and a playground within complex human-traffic patterns.

Ryznar says that at its heart GIS is really just an elaborate database manager. "That's what map data are, locations with data attached to them. If we choose a location, we have access to all of that other data associated with that location. It can be anything. You can map voter behavior, pollution information. You name it."

Graduate students see the possibilities. Ryznar has taught GIS to public health specialists, political scientists, archae-ologists, environmental engineers, as well as planners. Ryznar says researchers can use GIS to look at large amounts of infor-mation about a single place over time. "GIS is useful because locations have an impact on events," she says. "Of course, we've known that for a long time."

GIS may be one reason geography de-partments are humming again with spatial analysis research. "Geography lost favor in this country," Ryznar says. "In fact, the University of Michigan's department was closed down around 1980. Now there's been a renewed interest in geography nearly twenty years later—that's almost a whole generation of students—largely through the computer tools."

One researcher who appreciates the new software, but confesses her limited exper-ience with GIS, is planning graduate stu-dent JoAnn Carmin. "Often GIS becomes the project," she says. "In my case, I needed to use it as a tool." Carmin traveled to the Czech Republic to conduct her dissertation research. Since the 1989 revolution, Czechs have stretched their newfound freedom to participate in public decisions. She wanted to see how communities respond to new development proposals—landfills, inciner-ators, and highways.

"I found that Czech people have strong ties to their communities," Carmin says. "Unlike development in the United States, where sprawl stretches everywhere, you have lots of villages in the Czech Republic, over six thousand of them. People are well-connected to their cultural traditions."

To evaluate her statistical models, Car-min turned to Ryznar for help. She needed to find out if one town's response to a development proposal influenced the response of neighboring towns. "If town A protested and town B protested, was it a spatial effect?" she asked. It turns out that in a number of cases it was.

Carmin used intensive GIS analysis for just a few paragraphs in her dissertation. "Without GIS, I'd be missing a way to understand the quality of my data and to fully understand factors influencing behavior in Czech communities."

Sam Brody's store-bought tourist map of Maine doesn't inspire statistical confidence. In GIS parlance, it lacks projection. An over-sized Maine bursts from his office wall in full color. The rest of New England, and the world for that matter, fill in the edges. New York City sits squished underneath. Hollywood stands vaguely off to the left.

Ask Brody, a planning graduate student, about his own Gulf of Maine map and he'll escort you to a computer screen. From Cape Cod to the tide-tortured Bay of Fundy, the gulf spans some of the finest lobster, cod, and haddock fisheries in the world. Or it used to. That was before trawlers lugging huge drift nets brought the equivalent of forestry's clear-cutting to the ocean floor. Fish populations collapsed.

"I don't have a lot of hope," Brody says. "Mostly because of the way people view the marine environment." He explains how fishermen who know better are forced to overfish simply because their neighbors will beat them to it if they slow down.

Brody's map explains fishing policies. A patchwork of red and blue represent miles of protected areas and fishing closures. One of the biggest spans 900 square miles, about the size of North Carolina's Wake County. But click one area, and the details belie the protection. Some areas are only closed to one fish species—if they're observed at all—or closed during part of the year.

"With a visual map you can get a sense of scale and you can see where different re-sources lie," Brody says. "The ecosystem is interconnected." For instance, lobster brood stock grow up off the coast of Maine, but they're seeded off the coast of New Brunswick. Because the current runs counter-clockwise in the gulf, lobster is affected by far flung international decisions.

"When you're making decisions that effect natural resources and people's lives, it's not so much a map as a management tool," he says.

At first the federal government didn't want Brody to show all the protected areas. "They were afraid when people saw the map they'd question it, stir up some controversy," he says. "My objective was to give people something to fight over and then help them collaborate on a solution."

Brody already plans further research. He'd like to create a predictive model, like Messina's model in Ecuador, to show what level of protection would keep fisheries, rather than rainforests, from collapsing.

"I'm mapping the conflict between the need to use resources and the need to conserve them," Brody says. "It's so useful, I haven't realized all of its uses yet."

Like the Gulf of Maine, Great Smoky Mountains National Park is being mapped, not like Brody's map, but similarly. The National Park Service, with some help, aspires to identify every species in the park. All 100,000 of them. Biologists, taxonomists, data specialists, and ecologists will scour one of the most diverse collections of plants and animals in the temperate world.

"I want GIS to be a fundamental part of the project," says Peter White, professor of Biology and vice-chairman of the All Taxa Biodiversity Inventory. "Each microbe and plant has to be identified and tagged with a structure and place in the park."

But the park spreads half-a-million acres large, and includes some of the highest elevations in the eastern U.S. How will a biologist know where to start looking?

"I'm hoping to come up with the money to hire John Boetsch," White says.

Boetsch, a graduate student in Biology, has come up with a way to find plant species without actually going there first. As a park service biologist, he built a GIS model of four plant species that grow at high altitudes. By punching in the known locations of mountain bittercress, Cardamine clematitis, Boetsch figured out where it likes to grow: what elevation, slope, solar radiation, and water levels. From 18 known populations he modeled 187 coordinate positions, and then started hiking.

Boetsch spent two years trudging up and down some of the steepest parts of the Smokies, eventually finding 25 new populations. "I found Cardamine only ten percent of the time," he says. Even so, he considers the model a success since it gave him an idea of the plant's overall range.

Now he wants to use Cardamine like a canary in a coal mine. Cardamine grows at high elevations where ground-level ozone pollution gathers. Ozone damages plant tissue much as it can damage human lungs, but so far ecosystem effects are unknown. The park has some of the worst air pollution in the U.S. Boetsch says mountain bittercress and other high- elevation plants would make good indicator species.

"GIS allows you to step back and look at the larger picture," he says. "If we simply wanted to identify the species, that's not dependent on GIS. But if we want to see large scale patterns or do long-term species conservation, then we're definitely assisted by using GIS."

Collaborative research in Ecuador by the Carolina Population Center and the Department of Geography is supported by NASA.