In 1996, the U.S. Geological Survey, in cooperation with the U.S. Air Force, installed a study to test whether stands of eastern cottonwood trees could reduce the amount of dissolved TCE in the ground water and the aquifer. The hypothesis was that cottonwood trees could accomplish this by pulling up contaminated water and releasing TCE into the atmosphere through transpiration or by accumulating TCE in wood-a process called phytoremediation.

Cottonwood was a natural choice for the project. It grows quickly, has high transpiration rates, and puts out deep roots that tap into ground water; previous experiments had shown that the tree can metabolize TCE. For the study, researchers installed hundreds of cottonwood cuttings (whips) and 1-year-old seedlings on a 9.5-acre site on the golf course of the Naval Air Station adjacent to the old Air Force plant. Whips and seedlings were planted in separate plantations, their long sides perpendicular to the contaminated ground-water plume coming from the Air Force plant. Each plantation was about 235 feet long and 50 feet wide.

An Individualistic Approach

To be an effective tool for cleaning up TCE, the cottonwood trees needed to transpire large amounts of water from areas in the soil where TCE concentrations were the highest. To determine the phytoremediation potential of the trees, researchers installed an extensive ground water and transpiration monitoring system. They constructed a ground-water flow model based on monitoring data and coupled it to the transpiration model developed by Coweeta researchers Jim Vose and Wayne Swank, who were using recently developed techniques to directly measure transpiration on individual trees to come up with more accurate estimates and models. Because of their expertise in transpiration measurement and modeling, Coweeta researchers were asked to participate in the project, with funding provided by the Department of Defense.

"It was assumed that cottonwoods might work well in the subhumid climate of northeastern Texas, but the agencies involved wanted more specific information about how well these stands would remediate TCE," says Vose. "Though it was expected that soil microbes would biodegrade some of the substance, transpiration was the primary mode for removing TCE."

At the time, not many studies had been done on tree or stand-level transpiration, and most of those were based on data from gauged watersheds. "You have to thoroughly assess water-use patterns- transpiration rates, depth of soil water uptake, interactions with climate-to provide accurate information for phytoremediation purposes," says Vose.

SRS researchers monitored sap flow of individual trees in each plantation to find out how much water was transpired, using a statistical sampling of trees and model-based scaling to estimate total transpiration from each plantation. "There isn't any way to directly measure stand-level transpiration," says Vose. "But this tree-based approach gets us closer than previous methods."

When they analyzed data after the third growing season, researchers were not surprised that TCE was still entering the nearby aquifer; it can take at least a decade for trees to reach peak efficiency. They were, however, encouraged to find that the volume of water moving past the trees was reduced by 12 percent, and that the mass of TCE moving out of the site was down by 11 percent. This showed that after only 3 years, the trees were removing TCE from the aquifer. After 7 years, project researchers concluded that the cottonwood stands could substantially reduce the mass of TCE in ground water, primarily through transpiration.

SRS researchers also compared the effectiveness of planting 1-yearold seedlings at a cost of $8.00 each versus whips at $0.20 each. At the beginning of the study, seedlings had much higher transpiration rates, but in just a few years, their findings showed sap flow rates in whips greater to or equal to those in seedlings. In the long run, planting the whips might be more cost effective for most phytoremediation situations.

Core Issues

Coweeta researchers Barry Clinton and Vose went a step further, using stable isotopes of hydrogen and oxygen and tree coring to find out which source-surface water or ground water-the cottonwoods were drawing most of their water from. The study was the first attempt to look at the water habits of a specific tree species on a site where the primary purpose for growing the trees was to reduce the contaminant load in ground water.

"Using plants to metabolize or store pollutants is very attractive, but you need a full understanding of the water-use patterns of the plant species to be successful," says Vose. "Water use by cottonwoods varies with how much water is available, which allows the tree to survive in dry climates. When it rains, they switch to surface sources, which means they won't be as efficient in taking up TCE from ground water."

The researchers irrigated a study area to the rate of a typical rain event. Using the isotopes to mark the two different sources of water, they looked at changes in soil water content, sap flow rates, and TCE concentrations in tree tissues. They found that water from irrigation on the surface was taken up first, slowing down the uptake of polluted ground water. If trees switch to surface water sources when there is adequate supply, the overall effectiveness of the treatment will decrease-especially in high-rainfall regions-because less pollutant is extracted from ground water, a finding that emphasizes the need for good local climate data.

More precise data allows site managers to plan phytoremediation projects for maximum effectiveness. "It isn't really practical to use individual tree studies to monitor and predict transpiration rates for every phytoremediation," says Vose. "The results from these studies can help develop predictive tools to provide the species and site-based estimates of transpiration needed to design effective phytoremediation projects."

For more information: Jim Vose at 828–524–2128, x114 or jvose@fs.fed.us.

For more information about transpiration and other water research at the SRS Coweeta Hydrologic Laboratory, read Widening the Lens in Issue 5 of Compass, available online at www.srs.fs.usda.gov/compass/ issue5/02wideninglens.htm.