A Trace Element Analysis of the Gordon Town Site (40DV6): An Investigation of Intra-Individual Variation and Dietary Inference

The results of trace element analyses of archaeological human bone have been used to investigate questions pertaining to the dietary practices and subsistence behavior of various prehistoric populations. As a technique still in the experimental stage, however, a number of methodological difficulties have been identified. This thesis examined two aspects of trace element analyses of human bone. First examined was a methodological problem that has received increased attention since the initiation of this study: the documentation, measurement and control of intraindividual variability of trace element concentrations. Secondly, the validity of dietary inference from trace element concentrations were addressed using the results from the analysis of the Gordon Town (40DV6) skeletal sample, a Late Mississippian group from Nashville, Tennessee.

Five elements were measured in from two to four bones in 12 adult individuals. The elements chosen for measurement were manganese, magnesium, strontium, zinc and copper. High levels of the first three elements are indicative of a diet containing a large amount of plant foods. High levels of the last two elements indicate a diet composed of a high proportion of meat resources. The bones selected for analysis were the femur, humerus, rib and vertebra. The 12 individuals that comprised the sample met a number of stringent criteria designed to control for extrinsic (environmental) and intrinsic (physiological, biological) factors that may alter the trace element concentrations of an individual from those determined by diet.

Descriptive statistics, Student's T-tests and Analysis of Variance (ANOVA) techniques assessed mean bone element concentration differences between males and females, individuals of different age groups, individuals buried in different areas of the site and different bones of the skeleton. Results indicate that no differences in element concentrations existed between males and females, age groups and different areas of the site. Significant element concentration differences were found to exist in the four different bones analyzed. The differences seem to be related to the type of bone (compact, trabecular), bone turnover rates and stability of the trace element. It was also determined that intraindividual variation of element concentrations is in some instances greater than interindividual variation. These results indicate that the element concentrations of one bone cannot legitimately be compared to the element levels of a different bone.

Dietary inferences from the trace element concentrations of the bone analyzed were limited. Absolute or even relative proportions of plant versus animal foodstuffs in the diet were impossible due to the unusual elemental composition of the carnivore sample and the inaccurate strontium concentrations. It was concluded from the relative element levels, however, that the Gordon Town inhabitants consumed a diet composed of a variety of both plant and animal foodstuffs. The consumption of a varied diet by the Gordon Town people is in contrast to the predominant subsistence strategy of Late Mississippian peoples.