"We explain to the kids why, from the beginning of time in our homeland, we had the mounds. You can feel it in the classroom. There's a sense of dignity and a sense of loss."
The rate of mound destruction in northeast Louisiana is discouraging. Add land leveling to the equation, and it becomes staggering. In one parish, it is estimated that approximately 20 percent of the agricultural land has been leveled: so if one assumes an equal distribution of sites throughout the parish, 20 percent of the sites (prehistoric and historic) has been destroyed.
In the good old days, plowing lowered mounds, scattered artifacts, and disturbed subsurface features. Minimally, one still could determine a site's location, age, and (in many instances) function. It was possible to reconstruct settlement patterns through time. But where land has been leveled, one can't even determine exactly where the site was. All one can conclude is that a particular site was occupied during certain periods of time.
A 1993-94 survey of 35 mound sites in one parish determined that at least one mound (if not all) had been leveled at 12 of the sites. Conversely, only 5 of the sites (14 percent) had at least 1 mound that had not been disturbed. A 1995 survey of 9 sites in a second parish identified 5 where at least 1 mound had been leveled. All of the mound sites had been disturbed, with only two retaining visible evidence of being an earthen structure.
These figures may be unfair because they represent the destruction of mounds recorded as far back as C.B. Moore's early 1900s excursions in Louisiana.1,2 Also, landowners today are preserving most of the large mound complexes. They are aware of their significance and take pride in protecting them. But that awareness changes with ownership, as recently demonstrated by the leveling of the south end of Ridge 6 at Louisiana's Poverty Point.3 Furthermore, smaller mound sites are often overlooked, continue to be plowed or leveled.
These data strongly suggest that prehistoric sites are rapidly disappearing. All sites, historic or prehistoric, mound or non-mound, are impacted by land leveling. However, earthen mounds are fewer in number and represent only 361 of 2,547 recorded sites in northeast Louisiana. Mound sites probably number less than 300, given the erroneous identification of natural features as mounds (11 in two parishes alone) and the misplotting of sites, with multiple site numbers for the same mound complex (one mound site was assigned three different site numbers because of plotting errors).
This destruction, plus the fact that earthen mounds provide a wider range of data than other surface sites, is why we have focused our attention on this area for the past three years. As an outgrowth of our field work, we have developed new techniques for estimating the age of mounds, defining their stratigraphy, and perhaps eventually determining how old they are without excavating them.
Estimating the Age of Mounds
Louisiana has an advantage over other states in having developed a qualitative method to estimate the age of earthen structures. At least eight mound sites have been dated to >5000 B.P. (before present)4-9. almost doubling the known antiquity of mound building in the southeast United States. With a greater span of time to examine, the process of soil development in mound fill (i.e., how the soil changes over time) offers a convenient means for estimating the age of mounds as pre-Woodland period (>3000 B.P.), Woodland period (<2000 B.P.), and post-Woodland period (<1000 B.P.).
The weathering of soil provides clues that help us estimate the age of mounds. The rate and degree of soil formation on mounds is determined by time, parent material, environment, slope, and organisms. The process begins with the parent material, the soil and sediment used to construct the mound. The unweathered parent material is defined as the C horizon. As time passes, organic matter is mixed with the surface of the C horizon, forming an organically enriched A horizon on the surface of the mound. With the passage of time, the movement of water through the soil strips clay and iron from the A horizon and begins forming a B horizon between the A and C horizons.
The B horizon develops in two stages. First, an increase in iron, and a slight increase in clay, changes the color of the B horizon (usually redder), and this is called a cambic B (Bw). Second, the clay particles continue to increase in the Bw horizon, and through time, transform the Bw horizon into an argillic (Bt) horizon. In some soils, an E horizon forms between the A and Bt horizon. The E horizon is an horizon whose sediments have been stripped of clay and iron and lacks organic matter. Particle size analysis of samples from each horizon help to measure the degree of weathering that has occurred in the mound fill.
One can conclude that a mound with only A and C horizons is younger than a mound with A, B, and C horizons. Furthermore, a mound with an argillic B (Bt) horizon is older than a mound with a cambic (Bw) horizon. The actual amount of time necessary for the development of these horizons is unknown. However, preliminary research suggests that in northeast Louisiana, Bt horizons can be seen in mounds that are 3,000 years or older. Bw horizons develop in mounds that are about 2,000 years old. Mounds that are less than 1,000 years old usually lack distinct Bw horizons.
Soil coring provides an efficient means of defining mound stratigraphy. With examination of a soil core, one can identify stages of mound construction and if there is evidence of human occupation beneath the mound.
The people who built the mounds did so by depositing basketful after basketful of earth in a heap. When examining a coring sample, it is sometimes difficult to distinguish between individual basket loads of earth and the surfaces of succeeding stages of mound construction. Taking multiple cores, though, clarifies the stratigraphy because earlier stages of mound construction will be found at roughly the same depth in different areas of the mound.
The Absolute Age of Mounds
There are two ways we know of to evaluate the absolute age of a mound, and they can be employed without excavating. We are currently evaluating both methods.
The first is by recovering small fragments of charcoal from buried A horizons. With accelerated mass spectometry it is possible to obtain radiometric dates. Once the stratigraphy of a mound has been defined by coring, charcoal taken from the surfaces of earlier stages of mound construction or from sub-mound surfaces should provide accurate dates for the sequence of building and occupation.
We assume that charcoal recovered from a buried A horizon is in situ, undisturbed, a product of activities on the mound or sub-mound surface. Charcoal from basket-loaded fill is not dated, because it was transported to the mound from somewhere else.
The first time we tried this method, we were not successful. The site in question was remote, and we were unable to get to it with the coring truck. Instead we used a bucket auger (a mechanical arm with a bucket-like device on the end) to retrieve charcoal from the mound's A horizons. On its way in and out of the excavation shaft however, the bucket would knock charcoal from the surface of the mound down to the bottom of the shaft. Mistaken for material from the mound's buried A horizons, the charcoal was duly collected and dated. A site that had repeatedly dated to circa 3500 B.C. was suddenly dating to 1600 A.D.
With that episode in mind, we recommend that only samples obtained by continuous coring be submitted for dating. A continuous core collected from Mound 2 at Marksville, Louisiana, recovered organic material at 155 centimeters below the surface. A second core from a different place on the same at 157 centimeters down, indicating that we had located a former organic surface.
Both cores contained Bw horizons (a B horizon's earlier stage), suggesting a date between 2000 and 1000 years before present. Radiometric analysis of the organic matter produced a corrected date of about 1,460 B.P.
The second method is luminescence-dating sediments from buried A horizons in mound fill. Luminescence is a measure of the stored energy in crystalline materials (such as quartz) that accumulates by the action of natural radioactivity. When the crystalline materials are exposed to light or heat, this stored energy is released, thereby erasing the luminescence signal. So the signal is proportional to the time that has passed since the sample's last exposure to heat or light.
The layer of organically rich soil on the ground's surface—what eventually becomes the A horizon—contains quartz grains that, exposed to sun and daylight, lose their luminescence. Once the moundbuilders began their work, however, dumping their basketloads of earth over this surface, the quartz grains would have been shielded. Luminescence loss would have stopped and accumulation of stored energy would begin again. Thus, luminescence-dating of the A horizon would determine when the various stages of mound construction began.
Dr. James Feathers of the University of Washington's luminescence lab collected three sediment samples from radiocarbon-dated buried surfaces at two sites. The samples were collected from walls of test excavation units. If the dates prove reliable (results should be available by this spring), further samples can be collected from continuous cores.
We are aware that coring of mounds cannot replace proper excavation. Cores provide an extremely limited view of mound stratigraphy and virtually no information on associated artifacts. However, the rate of site destruction is so rapid that it will never be possible to conduct even limited excavations of impacted mound sites. So alternatives must be sought.
Obviously, preservation is the preferred option. But as we previously stated, commitment to preservation changes with ownership. Our second option is to collect as much data from existing mounds while minimizing impact on the sites.
Coring of mounds is perhaps the most efficient method available. There are several sound reasons why it should be pursued. Minimally, it will allow us to distinguish natural hills from mound sites. Second, it will provide accurate information about mound stratigraphy. Third, allowing that further refinement is necessary, the use of soil development for dating a mound is promising.
Fourth, AMS and luminescence-dating are exciting developments. They may prove to be an accurate method of dating mounds. Although both methods are considerably more expensive than standard radiometric dating, eliminating the cost of site excavation quickly balances the scales. The truth is that we continue to lose valuable information on a daily basis. The sadder truth is that the process cannot be stopped.
Joe Saunders is Regional Archeologist at the Regional Archaeology Program, Department of Geosciences, Northeast Louisiana University, Monroe, LA 71209, (318) 342-1899, fax (318) 342-1755. Thurman Allen is Resource Soil Scientist with the Natural Resources Conservation Service, 1605 Arizona St., Monroe, LA. (318) 387-8683, fax (318) 388-4275.
1. Moore, C. B. Antiquities of Ouachita Valley l4:7-170, 1909.
2. Moore, C.B. Some Aboriginal Sites in Louisiana and Arkansas 16:7-99, 1913.
3. Gibson, J. L. and J. Saunders. "The Death of the South Sixth Ridge at Poverty Point: What Can We Do?" SAA Bulletin 11(5):7-9, 1993.
4. Gagliano, S.M. Occupation Sequence at Avery Island. Coastal Studies Series No. 22. Louisiana State University Press, Baton Rouge, 1967.
5. Manuel, J.O., Jr. The Hornsby Site—16SH21: An Archaic Occupation in St. Helena Parish, Louisiana. Manuscript on file, Regional Archaeology Program, Northeast Louisiana University, Monroe, 1983.
6. Neuman, R. W. Report on the Soil Core Borings Conducted at the Campus Mounds Site (16EBR6) East Baton Rouge Parish, Louisiana. Museum of Geoscience, Louisiana State University. Submitted to Technical Report, 1985.
7. Russo, M. "A Brief Introduction to the Study of Archaic Mounds in the Southwest." Southeastern Archaeology 13(2):89-92, 1994.
8. Saunders, J., T. Allen, and R. T. Saucier. "Four Archaic Mound Complexes in Northeast Louisiana?" Southeastern Archaeology (Winter), 1994.
9. Saunders, J. W. and T. Allen. "Hedgepeth Mounds: An Archaic Mound Complex in North-central Louisiana." American Antiquity 59(3):471-489, 1994.