How soil can tell stories of the past
At the base of Monticello Mountain, just below Thomas Jefferson's historic estate in Charlottesville, Va., sits a 90-meter-long greenstone wall. The Rivanna River runs on one side. On the other, earth has piled up to the wall's top. Built up from sediments washing down the mountain for centuries, this soil holds clues to history. But rather than bits of tools or pottery, the clues are chemical elements in the soil.
A complex ecosystem, soil is home to matter animal, vegetable, and mineral. Numerous chemical, biological, and geological processes take place continuously among these players. Yet within this dynamic world, an imprint of the past appears in the variety and abundances of soil's chemical elements.
People directly influence these imprints. "As people live on a landscape, they leave all kinds of chemical residues around," says geoarchaeologist Vance T. Holliday of the University of Arizona in Tucson. Some elements build up over time, while others diminish. Studying these changes "can provide us with a record of how a landscape evolves," he says.
While archaeologists have long scrutinized soils to find traces of the past, advances in analytical technologies in the past 10 to 15 years have made the detection of dozens of chemical elements rapid and cost-effective for the first time. With that additional information, archaeologists are tying chemical signatures to agricultural practices and other human activities. But while soil chemistry offers new insights, it's most instructive in combination with artifacts and other historical evidence.
"Archaeology brings together many different lines of evidence to understand the narrative of the past," says Lisa Frink of the University of Nevada at Las Vegas. Soil chemistry is "another piece of evidence that adds to the understanding."
Both dirt's parent material—the bedrock that slowly fragments into soil particles—and its living constituents affect soil's natural chemical composition. Magnesium, calcium, phosphorus, and numerous other elements can show up within chemical compounds among soil particles or as ions bound to those particles.
When people occupy an area, they influence the amounts of the elements in the soil. In agriculture, fertilizer adds some elements to the soil and crops deplete it. Furthermore, domestic activities, such as preparing food, maintaining fires, and disposing of waste, concentrate certain elements in the soil, says archaeologist T. Douglas Price of the University of Wisconsin–Madison.
Archaeologists have the most experience with phosphorus, an element long recognized to indicate human activity. Burial sites, waste from people and animals, and meat, fish, and other food remains all add phosphorus to the soil, where it combines with ions to form stable phosphate minerals, says Holliday. Phosphorus accumulates while people occupy an area, and concentrations can rise orders of magnitude higher than those in soil relatively free of human contact, he notes.
Although many scientists have suspected for decades that "there ought to be other things besides phosphorus lying around," says Price, finding additional tell tale elements has only recently become practical. For example, instruments that perform an analytical technique called inductively coupled plasma spectroscopy are now available at many universities, notes Price. These machines identify elements extracted from a soil sample by weighing ions or by detecting characteristic wavelengths.
Having identified the elements in a collection of samples, archaeologists can map a plot of ground according to the elements' concentrations. By combining that information with details of artifacts and other evidence of people's presence, scientists are deciphering chemical signatures of various human activities. For instance, potassium and magnesium tend to be higher where wood ash once entered the soil from a hearth.
The evidence so far "makes us think that we are seeing some real patterns in the [chemistry] data," says Price. However, he notes that there are differences in conditions among sites, so archaeologists still "have to learn each site in a different way."
Soil-chemistry data can be especially useful in detecting human activities at sites of occasional gatherings. In such places, people carried away their food vessels or tools once they ended a ceremony or broke camp, so few artifacts typically remain for archaeologists to find.
An archaeological site in the northwestern part of Honduras, for example, has two main areas: a ceremonial plaza surrounded by pyramids and a residential patio encircled by dwellings. E. Christian Wells of the University of South Florida at Tampa works with his research group at this site, called Palmarejo, which dates to between A.D. 400 and 1,000.
As they expected, the researchers found almost no artifacts within the boundaries of either the plaza or patio. But near the plaza, they found large serving platters, grinding stones, and incense burners. Areas just off the patio held smaller versions of these items as well as bowls and cups.
This evidence suggests that the community cooked, ate, and held religious ceremonies in both places, but in larger groups in the plaza than in the patio.
To get a sense of where in these open spaces the activities occurred, Wells' group turned to soil chemistry. The researchers took 324 samples—one every 2 m in the plaza and every 5 m in the patio. They sampled from about 15 centimeters below the areas' current surface, which is the level of the original patio and plaza. By identifying elements such as barium, magnesium, and phosphorus, Wells and his team discerned several patterns, which they reported at the 2006 American Chemical Society meeting in Atlanta in March.
Phosphorus concentrations in the plaza displayed greater variation than those in the patio did, which suggests that cooking and eating occurred throughout the patio but only in certain spots on the plaza, says Wells. The team also observed variations in barium and magnesium concentrations in the southern part of the plaza but doesn't know what activities those patterns might represent.
To learn more about the ancient site, Wells' team is observing the activities and studying the soils of indigenous people currently living in the area. Such information is critical for "making inferences between soil chemistry and ancient activities that took place," says Wells.
Similarly, Frink and Kelly J. Knudson of Arizona State University in Tempe are examining modern-day indigenous communities in Western Alaska. The people there collect enough food in the short summers to last through harsh winters. Frink, Knudson, and their colleagues work in the Yukon-Kuskokwim Delta, about halfway up the state's coast along the Bering Sea, where the Yup'ik catch and process salmon and other fish at camps that exist only during the summer season.
Before the researchers begin excavating archaeological sites, they're observing activities and collecting samples from present-day Yup'ik fish camps. Families at the camps live in tents and set up fish-processing areas that include a fish-cutting station and covered drying racks. "If the site is ephemeral—you don't have people building large houses—then soil chemistry can be very helpful" in revealing a signature that might also locate an ancient camp, says Knudson.
In 2004, the team reported on three fish camps, one a site revisited by the same family for 30 years and two others that had each been inhabited for just one summer. At the older camp, the group found that manganese, phosphorus, and strontium concentrations in soil were an order of magnitude higher under the drying racks than in soils outside the camp. Drippings from the fish probably caused the difference.
The newer camps displayed a similar, but weaker, chemical signature under the drying racks. The poorly draining Alaskan soils appear ideal for retaining elements for analysis, the researchers say.
The group plans to look for chemical signatures at archaeological sites in the area. Ultimately, says Knudson, the goal is to "use the data we have from the fish camps to look at the past."
The soil also holds clues to past farming practices. Fraser D. Neiman, director of archaeology at Monticello, and his colleagues have excavated the layers of soil behind the meter-high greenstone wall at Monticello Mountain to study how changes in agricultural strategies affected the landscape as well as the lives of the plantation's slave community.
Jefferson's father, Peter, began growing tobacco and corn on a small field at Monticello in the mid–18th century. His slaves used hand hoes and slashed and burned trees to clear fields, leaving behind the tree stumps. This type of agriculture typically leads to some erosion, but stumps keep much of the sediment in a field, says Neiman.
Thomas Jefferson took over the farm around 1764. He continued his father's practices while establishing additional fields. In the early 1790s, Jefferson switched to growing wheat, a crop that required a completely different strategy. Because wheat fields need to be plowed, the younger Jefferson had his slaves remove the tree stumps. He also rotated the wheat crop with clover and other plants and, to keep fields fertile, added manure and gypsum, which is calcium sulfate.
"The plowing totally stripped the fields and set up a process in the mountain environment for lots more erosion," says Neiman. After Jefferson's death, farming on the mountain slopes ceased.
At the 2006 Society for Historical Archaeology meeting in Sacramento, Calif., in January, Neiman's colleagues presented their analysis of a 1-by-2-meter chunk of soil roughly 2 meters deep that they had removed from the area behind the greenstone wall. Its color and texture indicated rich, dark topsoil in both the top, modern layer and the fourth layer down, which was probably from the time when Peter Jefferson began the farm. The researchers took 50 sediment samples at regular intervals from the top to the bottom of the excavated material.
The chemistry data ties the soil's layers to past events at the farm. For example, the researchers observed a spike in the sulfur concentrations between the fourth and third layers of soil, which indicated the point at which Thomas Jefferson switched to wheat.
"Thomas Jefferson was a huge fan of gypsum as a fertilizer," says Neiman. "The sulfur spike tells us when we are beginning to see the use of fertilizer on permanent fields."
The researchers plan to test additional sites around Monticello.
In all of its work, the team incorporates soil-chemistry data, pollen analysis, and studies of artifacts and historical documents to learn more about how changing patterns of land use affected slaves. The work "gives us physical traces of changes in slave-labor routines [such as] going from tobacco to wheat, hoeing to plowing," says Neiman. "We can begin to think about the social implications for the kind of work slaves had to do and how they did it."
Soil-chemistry data also reveal how farming practices affect the soil itself. Jonathan A. Sandor of Iowa State University in Ames and his colleagues study sites in New Mexico that the Zuni people have farmed for at least 1,000 years. These are some of the oldest agricultural fields identified in the United States. Zunis and other Native Americans still farm in these areas today.
The Zuni people have farmed in the dry Southwest without using irrigation or fertilizers, says Sandor. Instead, their success rested in the placement of their fields at the base of hills. During summer in the Southwest, intense, brief rains brought water and organic debris tumbling down the slopes onto the fields. The runoff not only waters the fields but also replenishes the soil's fertility, notes Sandor.
To study how this roughly 1,000-year-long farming practice affected the soil's fertility, the researchers compared soil samples from hillside sites that are still being farmed with samples from sites that have no evidence of ever having been farmed and from sites that were historically farmed but are now abandoned. The team measured the soil concentrations of phosphorus and nitrogen, two nutrients that plants need to grow.
The researchers report in an upcoming Geoarchaeology that they found no significant differences in the elements' concentrations at the various sites, which suggests that the Zunis' agricultural practices maintained soil fertility. "What they did was apparently fairly sustainable over long periods of time," says Sandor.
This is in contrast to some modern cultivation practices, which can degrade the soil, he notes.
Sandor is also analyzing the elements in crops now grown on these fields and in the water and sediments that run off the mountains.
"Soil is an important natural resource," he says, "and understanding the long-term effects of human beings on soil is important to coming up with management practices that help us conserve it."
While soil chemistry is gaining ground in archaeology, scientists remain cautious about how much the data can say. "When people do chemistry work in soils, they can't do it just in that isolated context," says Sandor. He notes that soil is also a product of biological and physical processes.
Neiman observes that "most of the work that's been done more recently has tended to be empirical rather than driven by a well-grounded theoretical understanding of [soil] processes." To fully evaluate patterns of chemicals in the soil, scientists need a much better understanding of the mechanisms involved, he says.
Chemicals' movement within soil can confound interpretation of the data as well, says Holliday. While phosphorus tends to stay fixed, other elements, such as calcium, can be mobile. "If you are trying to make archaeological interpretations, you have to know that what you are measuring hasn't moved around," Holliday says.
Despite such difficulties, many researchers still consider soil chemistry an illuminating addition to other archaeological evidence. Says Neiman, "The payoff really comes in putting all this stuff together."
Department of Anthropology & Ethnic Studies
University of Nevada
Las Vegas, NV 89154-5003
Vance T. Holliday
Department of Anthropology
University of Arizona
Tucson, AZ 85721-0030
Kelly J. Knudson
School of Human Evolution and Social Change
Arizona State University
Tempe, AZ 85287-2402
Fraser D. Neiman
Department of Archaeology
Post Office Box 316
Charlottesville, VA 22902
T. Douglas Price
Laboratory for Archaeological Chemistry
5470 Social Sciences Building
University of Wisconsin, Madison
Madison, WI 53706
Jonathan A. Sandor
Department of Agronomy
Iowa State University
Ames, IA 50011-1010
E. Christian Wells
Department of Anthropology
University of South Florida
Tampa, FL 33620-8100