New AMS Dates on Human Bone from Mesolithic Oronsay
Richards, M. P., Sheridan, J. A., Antiquity
Here we report on new AMS dates on four human bones from two Mesolithic shell midden sites from the island of Oronsay, Inner Hebrides, Scotland. These new dates were commissioned by the National Museums of Scotland, to provide information for the `Early People' displays in the new Museum of Scotland which opened in Edinburgh on 30 November 1998. They complement the existing dates on shell and wood from Oronsay, published by Switsur & Mellars (1987), and indicate that these human bones date to the last phase of the Mesolithic period.
These four samples had also been used for stable isotope analysis, and the results were reported in this journal by Richards & Mellars (1998). Because of the obvious marine-based diets of these individuals we had to `correct' the dates for the marine reservoir effect before we could calibrate them. This correction is rarely applied to human bone, and its application can be controversial, but we argue it is crucial for correctly dating samples of this type.
Calibrating the dates
Collagen was extracted from the four samples as reported in Richards & Mellars (1998). The collagen quality was excellent, despite the age of the samples. The uncalibrated radiocarbon ages and associated [[Delta].sup.13]C values are presented in TABLE 1.
TABLE 1. Uncalibrated and calibrated radiocarbon dates from four human bone samples from the island of Oronsay.
lab no. sample site context no. OxA-8019 17157 Cnoc Coig Square H13, Unit 4 OxA-8014 17203 Cnoc Coig Square I 13, Unit 4 OxA-8004 18284 Cnoc Coig Square I 5, unit 4 OxA-8005 1281 CNG II Trench P/N, Layer 1/2, Unit 4 lab no. [[Delta].sup.13]C(a) uncalibrated [sup.14]C age OxA-8019 -12.4 5615 [+ or -] 45 BP OxA-8014 -12.0 5495 [+ or -] 55 BP OxA-8004 -12.4 5740 [+ or -] 65 BP OxA-8005 -16.0 5480 [+ or -] 55 BP lab no. calibrated [sup.14]C age(b) OxA-8019 4200-4000 cal BC OxA-8014 4000-3800 cal BC OxA-8004 4300-4000 cal BC OxA-8005 4200-4000 cal BC
(a) These [[Delta].sup.13]C values are from a second measurement of the collagen before radiocarbon dating. They have an error of [+ or -] 0.3% associated with them, and are statistically identical to the values reported in Richards & Mellars (1998).
(b) At 2[Sigma], rounded to the nearest century. Marine calibration data is from Stuiver & Braziunas (1993) and the atmospheric calibration data is from Stuiver et al. (1998). Calibrations undertaken using OxCal 3.3 (Bronk Ramsey 1995). Please see text for details of how the calibration was obtained.
Marine organisms have uncalibrated radiocarbon dates that are approximately 400 radiocarbon years older than contemporary organisms that derive their carbon from the atmosphere. In order to calibrate these samples to calendar years, it is necessary to use the marine calibration curve outlined in Stuiver & Braziunas (1993). Additionally, in some areas of the world there are local variations in the [sup.14]C contents of marine organisms which have to be taken into account before calibration. Unfortunately, although most researchers are aware of the problem of the marine reservoir effect, a suitable correction is not often applied to human bone (but see Arneborg et al. 1999), even in cases where the collagen [[Delta].sup.13]C value clearly indicates the individual has derived the majority of their collagen carbon from marine sources. Below, we apply this correction to the four Oronsay human bone dates.
In the UK, we can expect a human that derived all of their protein from marine sources to have a [[Delta].sup.13]C value of approximately -12 [+ or -] 1% (Richards & Hedges 1999), and a human that had exclusively terrestrial protein to have a [[Delta]. …