Military training is often assumed to negatively impact archaeological resources, but methodologies to estimate or infer damage to these resources are expensive and time consuming. Cultural resources managers require a technique allowing them to estimate past and current impacts of training on archaeological site condition and determine if the site is capable of supporting future training activities.

The technical objectives of this research were (1) to demonstrate that naturally occurring and culturally induced stratigraphic differences in a suite of soil parameters, such as soil organic carbon and chemical elements, at the interface of near-surface soil horizons can be disrupted to varying degrees by military training and other (e.g., agricultural) human actions, and (2) to use this predictable stratification as the basis for an innovative, low-cost, widely applicable, reliable method to identify the onset and quantify the extent of adverse impacts to archaeological deposits that are associated with military training.

Using changes in the distribution of selected soil chemical variables, this research applied simple soil coring and laboratory analyses to estimate severity and depth of disturbance. At Fort Benning, Georgia, and Fort Riley, Kansas, soil cores were collected from sites representing disturbed and undisturbed archaeological and non-archaeological treatment groups. Each core was subdivided into 5 cm increments, dried at 55°C, processed to pass through a 0.15 mm mesh screen, and analyzed for total nitrogen, organic carbon, pH, lime buffer capacity, cation exchange capacity, extractable aluminum, boron, calcium, cadmium, chromium, copper, iron, potassium, magnesium, manganese, molybdenum, sodium, nickel, phosphorus, lead, and zinc and total phosphorus, potassium, calcium, magnesium, sulfur, manganese, iron, aluminum, boron, copper, zinc, sodium, nickel, lead, and chromium. Data from each archaeological and non-archaeological site within each installation were combined for statistical analyses.

Significant treatment group by depth interactions for carbon, pH, total phosphorus, and extractable calcium, magnesium, manganese, and nickel indicated divergence in distribution of soil variables with depth that could be attributed to physical disturbance (mixing, inversion, or burial). Soil carbon, pH, calcium, magnesium, manganese, nickel, and phosphorus values were different for disturbed treatment groups to depths ranging from 10 to 30 cm when compared with undisturbed treatment groups, suggesting these variables may have utility in estimating severity and depth of disturbance. This was especially evident when data were used in ratios or combinations with other soil variables that would serve to normalize changes in absolute soil concentrations, thereby providing a more standardized measure of soil mixing across multiple soil types.

The recognition that ratios or combinations of soil variables might be more stable indicators of disturbance across wide geographic and edaphic ranges than single variables led to the development of several small-scale validation tests involving the concept of top-to-total ratios, total-to-soluble ratios, and leachable-to-nonleachable elemental ratios. Significant top-to-total, total-to-soluble, and leachable-to-nonleachable ratios were observed for total phosphorus and total sulfur, calcium and potassium, and potassium:iron/sulfur:iron, respectively, indicating that soil mixing to depths between 10 to 20 cm had occurred and that these ratios may hold promise as measures of severity and depth of soil disturbance. Top-to-total, total-to-soluble, and leachable-to-nonleachable ratios were only developed on small subsets of data from each installation as proof of concept exercises, but preliminary analyses indicate these types of derived ratios may be promising areas for continued research.

The use of soil pH, soil organic carbon, extractable nickel, calcium, magnesium, and manganese concentrations as indicators of both severity and depth of disturbance may be enhanced when used in combination or ratio with any number of other soil variables, including organic carbon, zinc, chromium, lead, etc., and when used in combination with other archaeological surveying tools to locate archaeological sites. Additional site analyses are needed to better verify the use of these and other soil indicators for these surveys.