Carey only had to drive a few minutes from Duke's West Campus to be surrounded by her experimental subjects-the trees of Duke Forest. The forest is a research treasure to the University of Minnesota ecologist because it's a site where tree age has been meticulously recorded for decades.

Such record-keeping is crucially important if she is to understand how water use and photosynthesis rates in pine trees change with age, Carey says. "I'm working over there with trees under fifteen years old, over there around thirty, down there they are seventy, and across the road where they are over ninety," she says, sweeping her hand around the expanse of forest. She also needs to study trees growing in the same soil types, because similar soil types have similar water-holding capacities. In addition to the age records, Duke Forest scientists keep information on its soils. Is it easy to find places with such meticulous data files? "No!" Carey replies emphatically. "Absolutely not."

Photo: Duke University Photography

For some seven decades, the university-owned Duke Forest has served the research needs of legions of scientists like Carey, who enter its expansive woodlands on projects both microscopic and monumental. The research reserve, which totals about 8,300 acres in six divisions, has been radar-mapped from orbit, overgassed with carbon dioxide, and analyzed for natural contributors to ozone pollution. The U.S. Forest Service even values death in this living laboratory, studying the roots of felled trees.

"There is absolutely no question of Duke Forest's value as a research enterprise," says John Harer, Duke's vice provost for academic affairs. "About 60 percent of it at any one time is being used for some research project or another, and it rotates around."

Many of these research projects take advantage of one of Duke Forest's greatest assets-its long recorded history. From the beginning, Duke Forest custodians have meticulously tracked what grows there. Other researchers are lured by the forest's isolation-as are local residents seeking an escape into nature. The forest offers trails for hiking and horseback riding, and scenic waterways beside which to picnic or meditate, away from the traffic and sprawl of the burgeoning city. The forest is also a scenic attraction for motorists. State Highway 751, the winding artery that leads motorists through a section of the forest to Duke from points west, has been designated a state scenic highway because Duke Forest provides unbroken vistas of trees.

But the tempting fact that all its acreage is undeveloped also places Duke Forest under increasing pressure as surrounding real-estate prices skyrocket. That's a reality that is not lost on university officials, who have responded with a land-use plan that preserves most acreage for teaching and research but also designates some acreage as an investment. That land-use planning process clearly recognized the research value and the extraordinary history of Duke Forest. In particular, says Norman

Christensen, the dean of the Nicholas School of the Environment, there were the extraordinary records of Clarence Korstian, the first director of Duke Forest and the former School of Forestry's first dean. Every five years for about forty years, Korstian would rigorously inventory the forest, "so we today would know what was there," says Christensen, first dean of the school that absorbed Duke's forestry program. "He set up experimental plots where he went in and located every tree."

These seventy years of data have "given us a glimpse at the ways forests grow and how they change that we could only infer very indirectly otherwise," he says. "This has really put the Duke Forest on the map and has stimulated interest across the country in long-term studies of areas we can be assured we can come back to twenty years or forty years from now."

When Christensen came to Duke in 1973 as a young ecologist whose previous experiences were confined to the western United States, he depended on Duke Forest to learn about eastern woodlands. After rediscovering some of Korstian's meticulous records, he joined forces with a colleague at the University of North Carolina at Chapel Hill to begin surveying other parts of the forest. "That really launched my serious career as an ecologist in the East, studying changes in forests," he says. "We have re-sampled these plots every five years since 1975. Many of them have been dramatically altered by hurricanes, disease, and other impacts. I think the significance of this kind of long-term research is much better appreciated today than it was even in Korstian's time."

The Nicholas School's Landscape Ecology Laboratory studies land-use patterns using computers, satellite imagery, and on-site observations. Each summer, the lab sends students out from Duke and UNC to re-measure some of the oldest Duke Forest plots, including Korstian's. "It's part of their initiation into the world of environmental sciences," says Patrick Halpin, an assistant professor for the practice of landscape ecology and the laboratory's assistant director.

Ironically, Halpin and his colleagues, including lab director Dean Urban, are also using Duke Forest as an "urban laboratory," tracking via satellite images the changing patterns of "urbanization." Of course, Duke Forest doesn't have parking lots or housing developments, but Halpin says it does have other analogous features like clear-cuts and changing tree densities. "We're also looking at the effects of open space on housing values and other features of land-use planning, and Duke Forest is one of the largest areas of open space in the Triangle. Eight thousand acres in forest is something that is quite remarkable, and is becoming even more remarkable as every other acre gets cut down in the Triangle."

Meanwhile, botany and earth-sciences professor James Clark is simulating hurricane-style damage in Duke Forest to test new ideas about how forests maintain species diversity. His previous experiments, conducted in the North Carolina mountains, have found that small gaps in the forest canopy from a few fallen trees are not enough to give light-dependent seedlings a chance to grow. Now his research team is creating gaps up to 120 feet in diameter in the mountain test plots, and will do so in Duke Forest, Clark explains, "looking at a totally different forestry type."

Clark can see other advantages to using Duke Forest. "We drive four and a half to five hours to get to our southern Appalachian site," he says. "We are much less efficient there than at Duke Forest, where we can just run out whenever we want. I don't know where else in the Triangle we could pull that off."

Decay was an attraction for Kim Ludovici, a U.S. Forest Service soil scientist working out of Research Triangle Park. "We wanted to find a way to monitor how fast roots decompose," she recalls. To do that, she decided to compare roots of trees cut at known times in the past with the roots of newly cut trees. "It turned out the records at Duke Forest were good enough. We went to ten different sites where stands had been cut at different ages." Last spring, using a backhoe and ditchwitch, she and forestry technicians removed some well-documented loblolly roots and surrounding soils and took them back to their lab for analysis. "Duke Forest made it easy, and made it possible."

Photo: Will Owens

How much tree emissions contribute to pollution is the intriguing question that brings the U.S. Environmental Protection Agency to Duke Forest. From a mobile research lab parked on a grassy field there, EPA instruments are sniffing out chemicals in the air.

Standing in the door of the converted school bus, EPA atmospheric chemist Bob Arnts says that emissions from both natural and human-made sources can combine in theatmosphere to produce the pollutant ozone. Natural ozone "precursor" chemicals are emitted from different kinds of oaks, pines, and grasses that grow in the forest, and Arnts' agency has developed an elaborate mathematical model that incorporates levels of all contributing chemicals and climatic conditions to gauge the ozone threat.

"When you get it all done, everybody asks, 'How realistic is this?' " Arnts says. So he and Nicholas School research associate Fred Mowry are setting up their complex air-monitoring instruments in various Duke Forest settings to see if the EPA models do reflect reality. "If there weren't Duke Forest, we wouldn't have a place where we could set up to have reasonably secure measurements," says Mowry, who has collaborated with Arnts since 1976.

"Let's say we rented a commercial forest. Next year it might not be there." Swathed in trees on the ground, Duke Forest's most dramatic current experiment may resemble an alien spaceport when seen from the air. The Forest-Atmosphere Carbon Transfer and Storage (FACTS-1) study site, funded by the U.S. Department of Energy, features seven rings of towers symmetrically arranged in the pine-dominated woodlands. The site's objective is to mimic the higher atmospheric carbon-dioxide levels expected over the next century due to fossil fuel burning and other human activities, and to test the effects of those levels on forestland.

Using an elaborate computer-control system, four of the tower rings emit enough carbon dioxide to create open-air "bubbles" registering one and a half times current atmospheric CO2 levels. The other three rings discharge no gas, thus acting as experimental controls. Experts expect such high carbon dioxide concentrations to be typical by about 2050. Since the full experiment began in 1997, researchers from Duke and other universities have documented higher loblolly pine growth rates under elevated carbon dioxide.

The high-CO2 pines also seem to be maturing earlier and producing more cones, according to recent data. "This is one of the landmark projects of carbon cycle and global change research in the nation, if not the world, right now," says William Schlesinger, James B. Duke Professor of Biogeochemistry and the principal investigator for FACTS-1. "And we have it here."

Lightning strikes, not trees, are what draws Steve Cummer to Duke Forest. Cummer, assistant professor of electrical and computer engineering at the Pratt School of Engineering, is studying "sprites," colorful but ephemeral flashes high above especially strong lightning discharges. The lightning bolts themselves are 1,500 miles away, striking in the Midwest, but Cummer is trying to avoid power-line "hum," electromagnetic emissions with frequencies that partly overlap those of his lightning signals. His Ph.D. adviser at Stanford actually set up an antenna in Antarctica to get away from the hum. Cummer's compromise was a dense thicket in Duke Forest, where he set up his small, magnetic-field-sensing antenna detector, strategically nestled beside a log to keep running deer from smashing his equipment.

Duke Forest has also been studied from space. In 1994, the forest became one of nineteen international "supersites" scanned by the orbiting space shuttle Endeavour in a test of advanced environmental radar mapping technology. As the shuttle aimed its radars at the forest's trees, clearings, and bogs, Nicholas School researchers and students from Durham's Riverside High School positioned themselves on the ground to observe and measure some of the same targeted plants and soils to compare with the radar data.

Duke Forest is a powerful tool for teaching as well as for research, a fact Christensen brings home each August when he takes incoming Nicholas School of the Environment graduate students there to learn about ecosystem management.

"The lesson I try to convey to the students, that I think is so amazing about Duke Forest, is the intertwining of history and change," says Christensen. "By 1800, most of the Duke Forest was in cropland up through the Civil War, and then muchof it was abandoned between Reconstruction and the Great Depression. The forests we see today are a product of human activity over hundreds, maybe thousands, of years.

"People do things to landscapes, and landscapes change, which provide new opportunities for people. Today's pine forests there are a product of old fields. And we are doing research on those pine stands that is going to affect the future. This cycle of human activity and ecological change, of realizing that there is not a single square foot of the forest that hasn't been heavily influenced by human activities, is really the focus of one day out in the forest."

That day puts the students in a classroom defined not by walls and blackboards, but by trees and fresh air. The setting serves to reinforce Christensen's emphasis on the ecological cycle. "It's really to get them to think about that," he says, "and then to think about what that means from the standpoint of how we manage those resources, how we study them, and how we understand them."

The wide variety of experimentation and teaching opportunities found in the forest have not only yielded invaluable scientific data and hands-on learning opportunities, but have also forestalled incursions by some land-hungry neighbors. In 1995, the university granted the Department of Energy an easement to use ninety-three acres of the forest's Blackwood Division as the site for FACTS-1. In 1996, NASA entered into an agreement with Duke to use 500 acres of the division for future "spaceborne imaging research and similar studies." Those two legal moves helped the university repel an attempt by the Landfill Owners Group, represented by Chapel Hill, Carrboro, Hillsborough, and Orange County, to take part of Blackwood for a new public trash landfill. That unsuccessful attempt "is a good example of why we need to have better communications," says Orange Commissioner Barry Jacobs '72. "Duke Forest should not be a target for land grabs for other public uses."

Orange County was by no means the first to raise the possibility of developing Duke Forest land. In 1985, with Triangle land values beginning to rise and Duke working on a long-range planning process, the university itself brought in the Urban Land Institute of Washington, D.C., to evaluate Duke Forest and other off-campus university land holdings. The institute's subsequent report raised eyebrows when it included development as an option for most of the forest divisions.

"Certainly the Urban Land Institute saw the forest as a bunch of trees on some land, and that maybe the university was missing some opportunities for the generating revenue," Judson Edeburn, Duke Forest's resource manager, recalls wryly. But there was a positive side that followed the immediate negative faculty reaction. The report, says Edeburn, "precipitated a much more in-depth look at the land, parcel by parcel, division by division, in terms of the history of teaching and research."

The final result was the 1989 report of Duke's Land Resources Committee, proposing a new classification system that would assign Duke Forest's current holdings to one of four different categories of potential use, including the two categories of continued teaching and research. The report, which was adopted by Duke's trustees, designated holdings in a third category for eventual campus expansion and other institutional uses. Acreages in the fourth category-"residual endowment land"-might indeed be leased, traded, or even sold, though the report emphasized that "this does not imply that such land will be commercially developed or otherwise used in a manner inconsistent with its current forested state." Most Duke Forest land, the document stressed, should be committed to "a continuation and enhancement of current academic uses."

Says Christensen, "I think the university, all the way up to the senior administration and board of trustees, has really bought into the importance of the forest." However, he notes that these land-use principals also recognize that some parts of the forest may someday become too surrounded or fragmented by development to remain useful for research. Already, places in the Durham Division-the most affected by urbanization -"are looking more and more like Central Park from the air," he adds. "There are parts of the forest that 99.9 percent of the university community have never visited. They're not scenic. They're not pretty. From a research standpoint, they're not all that valuable.