Isla de Mona, a four-by-six-mile island of carbonate rock, "sits like a limestone tile in the Caribbean," says Myrna Martinez, a Penn State doctoral student in geosciences.

Halfway between Puerto Rico and Hispaniola, Mona juts up, 50-meter cliffs supporting the high plateau that takes up all but a small strip of its land mass, a small coastal plain along the southwest border.

Below the island's flat surface, suffusing its porous foundation, is a lens-shaped aquifer of fresh water. "It's like a fishtank filled with sand," Martinez explains. "The water fills the spaces in the rock." The fresh water actually floats on top of salt water, she adds, due to the salt water's greater density. Fed by rainfall, which percolates down from the surface, the aquifer discharges its overflow via offshore springs.

Because Isla de Mona is uninhabited, this system has never been seriously disturbed. "There are only a few hand-dug wells," says Martinez, "and they are very shallow." Mona's aquifer remains close to its natural state, an ideal model for understanding how a small island retains and circulates groundwater.

Beginning in the summer of 1993, Martinez has worked as part of a team headed by the United States Geological Survey's Joseph Troester to complete the first geophysical reconnaissance of Isla de Mona.

To look for the aquifer's freshwater-salt water interface, the lower border of the lens, the researchers used two complementary techniques, both depending on the differences in electrical conductivity between fresh water, salt water, and dry rock.

In the first technique, they took two plastic-encased wire coils, like hula hoops, and laid them on the ground at fixed distances from one another. Transmitting an electromagnetic current through the ground from the transmitter hoop to the receiver, she measured the apparent conductivity between the two.

For the second approach, the team laid out a loop of wire on the ground and sent a current through it, setting up an electromagnetic field that would penetrate the surface to a depth of about 75 meters. Once the field was established, they shut off the current, then measured the time it took for the field to die out.

Martinez then placed the results into a computer, using a geo-electric model to determine changes in conductivity as a function of depth. The values she obtained located the island's fresh water-salt water interface. ("Salt water is a much better conductor of electricity.") Knowing the interface, she could then calculate the thickness of the lens.

Martinez estimates an aquifer under the plateau that is 14 meters thick at its center -- not nearly as much water as had been predicted by researchers extrapolating data from studies on nearby Puerto Rico.

A bigger surprise is that the island's small coastal plain turns out to possess a lens-shaped aquifer of its own, one nearly as thick as the main lens. The relative thinness of this secondary aquifer is puzzling, Martinez says, because the surface area of the coastal plain is so much smaller and its exposed rock so much younger than that of the rest of the island. "Because of all the alterations and weathering," she explains, "the older rock is, the more permeable it is." With younger, less porous rock, water can't flow through the system as quickly, and an aquifer tends to build up.

Next, the USGS team hopes to drill a series of wells in the coastal plain in order to measure water levels, salinity, and flow at the fresh water-salt water interface. Fortuitously, Martinez reports, project divers reconnoitering the island's underwater caves have found one that opens on the interface.

"At different depths," she adds, "we're installing conductivity sensors to see if the interface moves, and if so whether its movements can be correlated to tides or rainfall."

Once the team has gathered enough data, Martinez will calculate a water budget for the aquifer: how much flows in, how much drains out. Then, she says, she will simulate the system on computer, aiming eventually to be able to predict its behavior. "The data is somewhat limited by environmental constraints," she says, "but we should be able to give different scenarios of what could be happening."

Along with satisfying a scientific interest in carbonate-island hydrology, she concludes, understanding Mona may provide valuable insights for the management of less pristine groundwater resources.

Myrna Iris Martinez is a doctoral student in geosciences, 313 Deike Building, University Park, PA 16802; 814-863-3965. Her adviser is William B. White, Ph.D., professor of geochemistry. For the Isla de Mona project, Martinez is part of a research team headed by Joseph W. Troester, Ph.D., research hydrologist, U.S. Geological Survey, Caribbean District, Guaynabo, Puerto Rico. This research is funded by USGS. Martinez is a recipient of the Patricia Roberts Harris fellowship for minority students.