4. Results and discussion

Figure 2 details results of each Mesolithic individual. Results from Aveline's Hole cluster well in terms of width, but not length. The one individual from Totty Pot has a similar distribution of scratch dimensions to the Aveline's Hole individuals, possibly supporting its Mesolithic attribution, though this can only be confirmed with further radiocarbon dating. Individuals from Téviec do not cluster well in terms of width or length, although the median scratch widths from Téviec are continuously distributed between 0.8 and 1.7µm. The one individual from Hoëdic has smaller features than most of the individuals from Téviec. The two individuals from Oronsay have very large scratches compared to the other Mesolithic populations analyzed. In fact, the Oronsay individuals have larger scratches than any other population sample analyzed, which raises the possibility that some taphonomic process has altered the overall microwear signal via the removal of small features. This is unlikely however, as many more large (width > 2.5µm) scratches were found on the Oronsay teeth (Table 1).

Table 1: Number of wide scratches from every individual on each site.

SiteN scratchesN scratches width >2.5µm%
Oronsay 707 26337.2
Aveline's Hole 1258 806.4
Totty Pot 196 94.6
Téviec 1763 1599.0
Hoëdic 450 102.2
Hambledon Hill 3953 72118.2
Thornwell Farm 2638 58322.1
Port-Blanc 312 6019.2
Beg an Dorchenn 740 16822.7
Torlin 501 112.2
Clachaig 753 699.2
Audleystown 108 32.8
Millin Bay 471 5611.9
Ballynahatty 721926.4

For the purposes of comparison with Neolithic populations, results are discussed regionally. On sites from southern Britain (Figure 3), most individuals from Thornwell Farm and Hambledon Hill have very similar distributions of scratch sizes, which are significantly larger than those of Aveline's Hole and Totty Pot. In Brittany (Figure 4), more overlap is evident between the results from the Neolithic and Mesolithic populations, although as per southern Britain, Mesolithic individuals tend to have smaller microwear features than Neolithic individuals. In the northern Irish Sea area (Figure 5), this trend is reversed. The Neolithic Arran sites (Torlin and Clachaig) and the Co. Down Sites Audleystown and Millin Bay have small scratches, even compared to the other Neolithic sites, whereas the Oronsay individuals have large scratches. The one individual analyzed from Ballynahatty has similar scratch sizes to other Neolithic sites from southern Britain and Brittany. A summary of scratch widths at each site appears in Table 2.

Table 2: Median microwear scratch minor axis length (mnx) results for all sites mentioned in the text.

SiteScratch mnx (µm)95% C.InPeriod
Oronsay 2.00 0.14 2Mesolithic
Aveline's Hole 1.05 0.21 5Mesolithic
Totty Pot 0.94 - 1Mesolithic?
Téviec 1.18 0.25 7Mesolithic
Hoëdic 0.67 - 1Mesolithic
Hambledon Hill 1.54 0.11 14Neolithic
Thornwell Farm 1.57 0.14 12Neolithic
Port-Blanc 1.37 0.11 2Neolithic
Beg an Dorchenn 1.56 0.25 4Neolithic
Torlin 0.84 - 1Neolithic
Clachaig 1.06 0.31 2Neolithic
Audleystown 1.07 - 1Neolithic
Millin Bay 1.27 0.03 2Neolithic
Ballynahatty 1.64 - 1Neolithic

Dental microwear interpretation

Surprisingly little is known about the effects of specific foodstuffs on dental microwear. Noble and Teaford (1995) have shown that there is little or no food in modern western diet capable of causing the kind of microwear features observed among prehistoric human populations - this is apparent from the essentially juvenile aspect of our teeth today, with minimal wear even of high cusps well into adulthood and indeed old age. It is thought that most dental microwear is caused by contaminants in foods rather than intrinsic food particles. Possible exceptions are silica phytolith-rich plant foods, although the abrasive potential of these is under considerable doubt (cf. Danielson and Reinhard 1998; Sanson et al. 2007). Most dental microwear studies attribute microwear formation to a load of gritty particles (e.g. Gügel 2003; Mahoney 2005; Nystrom and Cox 2003; Teaford and Lytle 1996). What, therefore, does this tell us about human diet? The significance of dental microwear in prehistoric dietary reconstruction is revealed when considering the source of these gritty particles and the rate of dental microwear formation.

Grit can be introduced into food incidentally via natural processes as unwashed plant foods contain abrasive dust of atmospheric or hydrological origin (Ungar et al. 1995). Presumably meat left in the open, for the purposes of preservation for example, will also accumulate such abrasives. Processing of food also has much potential to introduce abrasives into food, especially when done with friable stone tools (Teaford and Lytle 1996). Dental microwear is thus direct evidence for the consumption of such broad classes of foods, even though individual dietary components cannot be identified using dental microwear at the current state of knowledge. Experimental work, using microwear simulations (Gügel et al. 2001) and living subjects (Teaford and Lytle 1996) has provided some insight into the relationship between microwear data and the abrasive particles that caused the microwear fabrics to form. However, these studies lack the detail necessary to make specific inferences about the foods that cause microwear features of certain sizes. Further work is planned that addresses this, using living subjects and microwear simulations under controlled conditions.

Microwear forms very quickly on dental surfaces and is hence a record of the last few meals an individual ate - the "last supper phenomenon" (Grine 1986). This is important as variability within a population sample (a number of individuals selected for analysis from one archaeological site) reflects dietary differences in the short-term. Nevertheless, results from sites studied in the current project show more between-population variability than within-population variability. Different scenarios can be evoked to explain this:

  1. Environmental parameters create similar dental microwear potential at each site, irrespective of the actual foods consumed.
  2. Diet among the individuals studied was somewhat monotonous over time at the site level and different foods were consumed at different sites.

Scenario 1 can be partly refuted by the contrast between Mesolithic and Neolithic sites within each region. However, some environmental parameters such as vegetation cover, prevailing local weather conditions and (in the case of coastal sites) sea-level can change over the millennial time scale that separates the Mesolithic and Neolithic sites in time. Besides, the regions defined in the current study are Cartesian and arbitrary, whereas factors such as distance to coast, climate and surface geology could all potentially influence dental microwear. Despite this, all Mesolithic sites studied have produced different microwear results to those of nearby Neolithic sites, hinting at possible culturally determined factors common to local communities (scenario 2 above). Of course, these scenarios are not mutually exclusive, and it is possible that both environmental and cultural parameters have influenced the observed dental microwear signals.

It is interesting to note that overall variability is higher within the Mesolithic Aveline's Hole and Téviec samples than within the Neolithic Hambledon Hill and Thornwell Farm samples. It is possible to interpret this result as evidence of a more varied or perhaps seasonal Mesolithic diet, compared to that of the Neolithic, although studies of other Mesolithic and Neolithic sites are needed to confirm or refute this trend.


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