Optical dating of dune sand from Blombos Cave, South Africa: I—multiple grain data

https://doi.org/10.1016/S0047-2484(03)00048-4Get rights and content

Abstract

An aeolian sand unit overlies the Middle Stone Age deposits at Blombos Cave on the southern Cape coast. These deposits contained culturally-important artefacts, including bone tools and pieces of engraved ochre, as well as a large number of worked lithics. The aeolian sand and two other remnants of the sand dune formed against the coastal cliff were dated using optical dating. To determine the dose received since deposition, measurements were made on 5 mg aliquots of purified quartz grains using the single-aliquot regenerative-dose (SAR) protocol. The results of several internal check procedures are reported and at least 15 replicate dose determinations are presented for each sample. Combining these dose values with measurements of the radioactive content of each sample resulted in an age of 69.2±3.9 ka for the unit within the cave, and a mean age of 70.1±1.9 ka for all three dune samples. This provides a minimum age for the Middle Stone Age material at Blombos Cave.

Introduction

Optical dating provides a direct means of dating sedimentary units, with the age of last exposure to sunlight being obtained from measurements of the optically stimulated luminescence (OSL) and the radioactive content (Aitken, 1998). It is especially important for sediments that are beyond the range of radiocarbon dating (∼40 ka). Thus, optical dating is appropriate for providing chronological information for sediments relating to the Middle Stone Age in southern Africa. Studies of the OSL behaviour of quartz from Australian sand dunes, recently summarized by Wintle and Murray (2000), have led to the development of an improved laboratory procedure for measuring the radiation dose to which grains have been exposed in their environment; this dose is called the equivalent dose (De). The decay of elements in the uranium and thorium decay chains, and the decay of 40K, with a minor contribution from cosmic rays, produce radiation in the environmentresulting in the delivery of an annual dose to the quartz grains.

Several different luminescence-based dating procedures have been applied to Middle Stone Age sediments from sites in southern Africa, e.g. thermoluminescence (TL) and OSL analysis of quartz at White Paintings Rockshelter, Botswana (Feathers, 1997) and at Die Kelders, South Africa (Feathers and Bush, 2000), and TL and OSL analysis of quartz and infrared stimulated luminescence (IRSL) analysis of K-feldspars and polymineral fine grains at Klasies River and Duinefontein, South Africa (Feathers, 2002).

In this paper we have chosen to apply the single aliquot regenerative dose (SAR) protocol of Murray and Wintle (2000). Our choice is based on the excellent agreement between OSL ages determined by SAR and independent ages for about 50 samples reported by Murray and Olley (2002). This data set included sand-sized quartz from Holocene (Murray and Clemmensen, 2001) and Late Glacial (Hilgers et al., 2001) aeolian sediments with radiocarbon control, and silt-sized quartz from a marine core from the Indian Ocean, with a variety of independent age controls back to 120 ka (Stokes et al., 2003). Our study reports results for three samples of dune sand from Blombos, South Africa, where cave sediments have been found to contain lithic tools, shaped bone artefacts and decorated ochre (Henshilwood et al., 2001a, Henshilwood et al., 2001b, Henshilwood et al., 2002).

Section snippets

Site and sample description

Blombos Cave is one of a number of sites along the southern Cape coast of South Africa, including Die Kelders and Klasies River (Fig. 1), which have been found to contain evidence of Middle Stone Age occupation. The Middle Stone Age deposits contain a large number of finely-worked bifacial points, a range of bone tools (Henshilwood and Sealy, 1997, Henshilwood et al., 2001b) and several pieces of engraved ochre (Henshilwoodet al., 2002).

Blombos Cave (34°25′S, 21°13′E) is located 300 km to the

Sample preparation

The surface layer of samples ZB13 and ZB20 that might have been exposed to sunlight was discarded. Carbonates were dissolved in 10% HCl and organic matter was oxidised in 30 vols H2O2. Mechanical dry sieving separated the 212–250 μm diameter grain size fraction for ZB15 and the 180–212 μm grain size diameter fraction for ZB13 and ZB20. Quartz was obtained from these fractions via density separation using a sodium polytungstate solution of specific gravity 2.62 to reduce the potassium feldspar

Dosimetry

Dose rates were calculated using the conversion factors of Adamiec and Aitken (1998) with corrections for grain size and moisture content (Aitken, 1985) and the data are given in Table 2. The beta dose rate for ZB15 in Table 2 is about 1% lower than that used previously to calculate the total dose rate (Henshilwood et al., 2002) because we had previously omitted to allow for the effect of the HF etch on the beta dose. The external alpha particle contribution is considered negligible since the

Equivalent dose and age calculations

Fig. 9shows a plot of the equivalent dose (De) versus the preheat temperature for sample ZB15. By using a broad range of preheating temperatures, the pattern of sensitivity changes is accentuated, as previously shown in Fig. 5. Thus, being able to obtain the same Devalue, even when using very different preheat treatments, implies that the correction within SAR for luminescence sensitivity changes is effective, and hence the confidence in the final Devalue is enhanced. For all three samples the

Summary

Three dates have been obtained for remnants of a coastal sand dune at Blombos on the southern Cape coast of South Africa. Optical dating of purified quartz grains was achieved using a SAR protocol. The reliability of the procedure was demonstrated by undertaking a recycling ratio test, by confirming that the Deestimates are independent of preheat temperature, by recovering a given laboratory dose, and by confirming the purity of the refined quartz via IRSL tests. Comparing the results of the

Acknowledgements

The authors wish to thank Professor C.S. Henshilwood for providing access to the Blombos Cave, Dr H.M. Roberts for assistance in the field, and Bert Roberts, Stephen Stokes and an anonymous referee for extensive comments. ZJ acknowledges a grant from the Sir Henry Strakosch Memorial Trust.

References (39)

Cited by (0)

View full text