Radiocarbon Dating of Samples from the Azokh Cave Complex (Peter Ditchfield)

A total of 18 samples from the Azokh Cave Complex have been submitted for radiocarbon dating by AMS at Oxford Radiocarbon Accelerator Unit (ORAU), the details of which are listed in Table 16.1. These consisted of nine charcoal samples and nine bone samples from Azokh 2. Six of the twelve samples (four bones and two charcoal samples) failed during pre treatment and yielded no datable material. Ten of the samples yielded finite ages ranging from 2366 ± 35 to 122 ± 23 14C BP (Before Present – AD 1950). Two charcoal samples yielded dates greater than 62,000 14C BP.

Pretreatment and Measurement

Chemical pretreatment, target preparation and AMS measurement of the samples were carried out by standard

methods employed at ORAU. Details of the current pretreatment methods used at ORAU can be found in Brock et al. (2010) and are briefly summarized below.

Collagen was extracted from bone samples using a sequential acid-base-acid wash at room temperature consisting of 0.5 M hydrochloric acid (3 washes over approx. 18 h), 0.1 M sodium hydroxide (30 mins) and 0.5 M hydrochloric acid (15 mins), with thorough rinsing with ultrapure (Milli-Q) water after each step. The crude collagen was gelatinized at 75 °C and pH3 for 20 h. The gelatin solution was filtered using a 45–90 micron Eezi-filter™ and then ultrafiltered using a Vivaspin™ 15–30 kD MWCP ultrafilter.

Charcoal samples underwent a similar acid-base-acid pre-treatment of 1 M hydrochloric acid (*20 mins or until effervescence has finished), 0.2 M sodium hydroxide (20 mins) and 1 M hydrochloric acid (1 h) at 80 °C.

The samples were then freeze-dried before being combusted and the resulting carbon dioxide collected cryogenically and graphitized prior to AMS dating, as described by Brock et al. (2010). For details of the target preparation and AMS measurement see Bronk Ramsey et al. (2004a, b, c).

Calibration

Of the twelve samples that yielded radiocarbon dates, two were “greater than” ages, and five were relatively recent. Thus there are only five dates where calibration is either possible or appropriate. These are samples AZK14 (OxA 17589), No. 121 (OxA 23364), No. 153 (OxA 23543,

OxA23544) and Sample 1 (OxA 18875). The latest is a bone fragment from Azokh 2, the rest are from Azokh 5 (top of the series, i.e. Unit A). These were calibrated using the OxCal calibration program version 4.1.6 (Bronk Ramsey 2010) and the Intcal 09 calibration data set. The calibration plots of these samples are shown in Figs. 16.4 and 16.5.

Electron Spin Resonance (ESR) Dating

(Rainer Grün, Wendy Lees, Maxime

Aubert)

All samples were collected in situ and submitted by the excavators along with sediment directly attached to the samples. The dating procedures followed those routinely applied in the ANU ESR dating laboratory. From each tooth, an enamel fragment with attached dentine was removed and analysed for uranium and thorium using laser ablation ICP-MS (Eggins et al. 2003, 2005). The sediments were analysed for U, Th, and K by solution ICP-MS (Genalysis, Perth). For ESR dose analysis, the enamel was powdered and successively irradiated in 24 steps to 3188 Gy (samples 2380, 2382, 2383, 2384, in 2007) and 16 steps to 1839 Gy (2668, 2689, 2691 and 2692, in 2009). Radiation doses were monitored with alanine dosimeters and evaluated against a calibrated dosimeter set (A. Wieser, Messtechnik, München). Dose values were obtained fitting the natural spectrum back into the irradiated ones (Grün 2002).

For the calculation of the internal dose rate values, the beta attenuation values of Marsh (1999) and an alpha efficiency of 0.13 ± 0.02 (Grün and Katzenberger-Apel 1994) were used. No in-situ gamma spectrometric measurements were carried out. Considering that about 50% of the gamma dose rate is generated by the 5 cm surrounding of the sample (Aitken et al. 1985), the gamma dose rate was calculated to 50% from the sediment attached to the sample and to 50% from the average of all sediment samples from that particular bed. A time averaged water content of 15 ± 5 was assumed for the sediments. Age calculations were carried out with the ESR-DATA program (Grün 2009a). The ESR results are presented for early U-uptake (EU) as well as combined U-series/ESR. In most cases it was not possible to use the p-value system for combining the U-series and ESR data sets (Grün et al. 1988). Instead, ages were calculated according to the closed system U-series (CSUS) ESR system (Grün 2000), which is more robust and allows solutions when U-leaching occurs. Because most EU-ESR and closed system U-series ages were quite close, the choice of the U-uptake model was not critical (Grün 2000). At most archaeological sites, faunal samples experience delayed U-uptake (Grün 2009b). Here, the EU-ESR age presents the minimum age estimate, and the CSUS-ESR result the maximum possible. However, when U-leaching may have occurred (see below), the EU-ESR age is a maximum age estimate and the CSUS-ESR result gives an indication of the age overestimate.

Results and Discussion

All analytical data are listed in Table 16.2, sorted according to the depth in the profile. The only non-provenanced sample is 2383. The four digit sample numbers indicate an individual tooth, A and B repeat analyses on the same tooth, but on different enamel pieces.

Surprisingly, the EU ESR results of all provenance samples are in stratigraphical order because at most cave sites reworking of teeth leads to a large scatter in the ESR results. However, there are some problems. Some of the U-series results, particularly from Unit II are older than the EU ESR results. This is only possible, if either some U-leaching has occurred in the samples, or if they were reworked from other layers with a lower environmental dose rate. If leaching has occurred (perhaps influenced by guano), the CSUS calculation gives an indication of how strongly the age results were affected by this process. If the samples were reworked any assumption of when they were incorporated into Unit II, i.e. when Unit II was formed, is pure conjecture. The base of Unit II seems to have been deposited during MIS 6. If the teeth below Unit II were deposited in-situ, the contact between Units IV and V has an age of around 200 ka, and Unit Vm was deposited between *200 and *300 ka.

Amino Acid Racemization Dating

of Fossil Bears from Azokh 1 (Trinidad Torres, José Eugenio Ortiz, Arantxa Díaz Bautista)

Introduction

The amino acid racemization dating method is based on a chemical reaction (racemization) which allows the establishing of aminozones that reasonably have the same age (isochronous). If independent (radiometric) datings are available, it is possible to calibrate amino acid racemization analysis results and to use the method as a numerical dating system.

In general terms, cave bear localities are the best sites for sampling because they are clean, and the cavern’s thermal history did not show strong variations, with average temperatures moderately low (Torres et al. 2003). In a recent experiment the winter-summer temperature variation within the cave sediment at 30 cm deep was 0.5 °C. Outstanding amounts of amino acids were found in bear dentine samples with ages ranging from 10,000 to more than 300,000 years (Torres et al. 1999).

(continued)

Dating Appendix: 16

331