charcoal fragment sampled from a small 1 × 2 m test trench made from Unit A up to Unit E in the cave wall sedimentary section. AZK14 charcoal was found at the same height as and next to modern human teeth found in Unit A Azokh 5 (King et al. 2016). Charcoal AZK14 (OxA 17589) yielded an age between 722 and 384 years calBC. The bone was not affected by guano. These excavations, and especially the section located near the cave wall, are protected from bat guano deposition. Results obtained from bone (OxA 23543 and OxA 23544) recovered from the excavation on top of the sequence have a likely age between 126 calAD and 178 calAD (see Fig. 16.2). Charcoal (OxA 23364) gave an age between 715 and 888 years calAD, which may suggest possible contamination.

Uranium Series

Uranium series dating has also been applied to fossils from Azokh. The effective dating range of this method is between 1,000 to 350,000 years, but it can be extended (if the range of error is acceptable) to *400–500 ka. This method is

based on radioactive decay of uranium series isotopes (230Th/234U, 234U/238U). Uranium, relatively soluble, is

originally incorporated into the sample when the material (bone or stone) is formed, and thorium is incorporated into the sample with time. The ratio of uranium/thorium is then a direct measurement of the time elapsed since the sample formed. The most reliable material to date is cave speleothem, but so far the only one found developed in Azokh 1 is in a small chamber located at the front of the Azokh 1 cavity which was discovered in 2009 (see Sect. 16.5 below). Fossils (bones or teeth) have traditionally been considered to be unreliable due to their facility to uptake exogenous uranium after burial (Pike et al. 2002). In addition, uranium can be leached out of a bone, but not thorium, leaving a thorium excess leading to overestimated U-series dates. This, however, can be corrected using a diffusion-adsorption model (Millard and Hedges 1996) based on the geochemical context of bone-uranium-burial environment interactions (Pike et al. 2002). Initial dating by this method on fossil bones on the surface of Unit Vm in Azokh 1 gave ages between 191 +68/–36 ka and 186 +91/–48 ka. Simultaneously, results obtained by electron-spin resonance (ESR) as well as by racemization methods indicated an age ca. 300 ka for contemporaneous fossils. Although it is our intention to continue dating by this method, we need to come to a better understanding of the diagenetic and microgeomorphological processes operating in the cave to better understand the burial environment of the site. This is especially relevant as results from ESR methods indicate anomalies that could be explained by some U-leaching influenced by guano or fossil reworking (see Sects. 16.3 and 16.4 below). Reworking,

however, has been shown by taphonomic analyses not to have altered these fossil bones (Marin-Monfort et al. 2016).

As well as U-Th dating, Uranium-Lead (U-Pb) dating was used to attempt to determine the age of the speleothem deposits, in this case, stalagmites, in the cave. As outlined above, U-Th dating is useful for ages up to about 400 ka, beyond which the Th isotopes themselves have decayed away. U-Pb dating uses stable Pb istopes at the end of the U-series decay chain and has recently been successfully applied to speleothem (cave) carbonates from a few hundred thousand years (Richards et al. 1998) to material of several million years (Woodhead et al. 2006). Speleothem samples for U-Pb dating were collected from a large stalagmite boss near the entrance of the cave and from a number of small stalagmites situated at the very back of the cave. Initial attempts at dating Azokh cave speleothem were unsuccessful given the high Pb content of the material. A second attempt using cleaner, clearer calcite provided an age of 1.19 ± 0.08 Ma (see Sect. 16.5 below). This is currently the oldest age for any material from the Azokh Cave Complex and gives a minimum age for the formation of the cave itself. This opens up the possibility for the presence of older occupation layers.

ESR

Electron-spin resonance (ESR) dating was applied to several samples from Azokh 1 (see Sect. 16.3 below). This method is based on determining the natural radiation dose to which a sample has been exposed during its burial period. The sources of radiation are mainly from uranium and thorium in the sample itself, and from the radioactive isotopes of uranium, thorium and potassium in the surrounding sediment (Grün 2006). The most reliable material is tooth enamel because hydroxyapatite crystals are larger and more stable and closely packed than in bone. Modern enamel does not contain uranium, which is incorporated in the enamel crystals after burial, and uptake depends on the manner by which uranium enters into the enamel. Natural radiation generates new free radicals, trapped electrons and holes. The signal of the sample is called the natural intensity, which is dependent on the number of traps, the strength of the radioactivity (dose rate, D) and time (Grün 2006). To obtain a date, the fossil tooth is processed together with sediment underneath it. Dating of tooth enamel has been recognized as a useful tool for chronometric dating in the time range beyond the limit of radiocarbon and up to at least 2 Ma (Schwarcz et al. 1994).

Pleistocene fossils from Azokh 1 were dated by ESR (see Sect. 16.4 below). Six of these samples failed because the enamel was not thick enough. This happened for all bear canines and two bear premolars. Sample 2691 has been recovered from the base of Unit IV, close to the contact with

Unit V (equivalent to the top of Vu). Unit III has not yielded teeth with sufficient enamel thickness to be dated. A mandible of sheep from the section of Azokh 5, Unit B, was also processed, but dating failed. Results obtained by ESR have provided dates that are congruent with depth (Fig. 16.2). There is, however, an exception for sample 2384 (Table 16.1), which has given a younger age than the preceding and subsequent stratigraphically ordered samples. Once the depths of partial excavation Z coordinates have been referred to the datum, consecutive age results obtained by ESR according to depth support the lack of reworking processes involved in the site formation of Azokh 1.

Amino-acid Racemization

Racemization dating (see Sect. 16.4 below) measures the decay rate of protein amino acids in past living organisms. These amino acids can have two different chiral forms

(mirror images of each other), of which left-handed (levo, or L) is the condition when the organism was living. Once the organism dies the amino acids slowly turn into right-handed (dextro or D) amino acids until equilibrium is reached. This process is called amino acid racemization. The D/L ratio can be used for dating up to the time of equilibrium (D/L *1) (Fernández et al. 2009). Racemization is a chemical process that is highly temperature dependent and occurs faster under warmer conditions. These effects restrict the application of racemization and usually requires comparison with other dating methods. Diagenetic studies by Smith et al. (2016) indicate that collagen is generally absent from the Azokh fossil bones. However, racemization dating has provided ages that overlap the ESR dates (Fig. 16.3). This may result from sampling for racemization which uses dentine covered by enamel, and this may protect collagen at these particular areas from the destructive diagenetic effects observed in Azokh 1 due to bat guano (Marin-Monfort 2016). This explanation, however, needs further investigation.