3. Methodology

A number of different tooth surfaces have been used by different researchers for dental microwear analysis, but the molar occlusal surface is the most commonly employed. Only adult teeth are were used for the current study as few deciduous teeth were available for analysis and structural differences exist between the enamel of adult and juvenile teeth (Scott and Turner 1997). Occlusal force can vary according to position along the dental arcane (Spencer 1998). Therefore, in order to eliminate possible variability due to tooth position, the second molar was preferentially selected here for microwear analysis. However, in order to increase sample size, other molar teeth were used when necessary. Recent work by Mahoney (2006) suggests that between-tooth differences in microwear are rarely significant, barring changes in masticatory behaviour caused, for example, by dental pathologies. There should be no difference in dental microwear between equivalent wear-facets of maxillary or mandibular teeth, although significant differences in microwear patterns are observed on different facts of the same tooth (Teaford and Walker 1984).

The teeth from Oronsay were imaged directly using a CamScan MX2500 Scanning Electron Microscope (SEM) in low-vacuum mode at the electron microscope facility of the National Museums of Scotland. Fields of microwear were recorded using digital imaging, at x500 magnification, which has become the standard working magnification of most SEM-based dental microwear studies (Mahoney 2005; Nystrom and Cox 2003; Schmidt 2001).

Teeth from Aveline's Hole, Téviec, Hoëdic and the comparative Neolithic sites were replicated prior to SEM analysis using Coltène President Microsystem lightbody, a commercially available polyvynalsiloxane compound designed for use in dentistry. These moulds were then cast into positive replicates using a high resolution epoxy, Araldite MY753 with XD716 hardener. The replicates were then sputter-coated with gold and imaged in a JEOL JSM-6500 field emission SEM at the Electron Microscope Unit at Queen's University Belfast at x500. The resulting photomicrographs were recorded digitally.

Microwear features were marked-out on a computer using the SEM digital photomicrographs and Microware (Ungar 1997), a piece of software which has become the de facto standard among microwear researchers. The resulting lists of co-ordinates of pits and scratches were analyzed further using custom utilities written specially for this project in GNU R, creating summary statistics and for each photomicrograph, individual or site, depending on the unit of analysis required.

A number of variables can be extracted from dental microwear data. Many studies have concentrated upon the relative number of pits and scratches on the tooth surfaces or the preferred orientation of the scratches (e.g. Molleson et al. 1993). Additionally, the size of the pits and scratches (i.e. their major and minor axis lengths) has also been used to characterize populations using dental microwear data (e.g. Mahoney 2005; Nystrom and Cox 2003). Dental microwear data, as well as responding to dietary factors, is also sensitive to the tooth contact area, or facet, used for analysis (Teaford and Walker 1984). Mahoney (2005; 2006) has shown that dental microwear data display significant within-facet differences. Ideally therefore, exactly the same part of every tooth in the study should be examined. However, this is not possible with the material considered for the present study for two reasons. Firstly, with advancing age, the enamel on the teeth of an individual wears away, causing an obliteration of the distinctive pattern of cusps and facets. Secondly, the small number of Mesolithic individuals available in museum collections does not allow for the age of each individual to be controlled. Therefore, it is necessary to analyze dental microwear data in such a way that between- and within-facet differences are kept to a minimum. Pits form under certain mechanical conditions on so-called phase II facets, thus are not particularly useful unless the age-at-death of each individual can be controlled. Scratches are useful because they form on all facets, although their length varies according to facet position and tooth location (Mahoney 2006). Scratch width tends to be constant over different facets (Teaford and Walker 1984), and arguably relates more directly than scratch length to the size and nature of the abrasive particles. Therefore, results presented here concentrate on the scratch size variables, and scratch width must be seen as the most discriminating variable. Up to 600 dental microwear features were measured from each individual. Dental microwear data are not normally distributed, so microwear scratch variable distributions are summarized here using their median and the estimated 95% confidence interval.


© Internet Archaeology URL:
Last updated: Tues Oct 2 2007