Carbonates, Speleothem Archaeological (U-Series)
KeywordsInductively Couple Plasma Mass Spectrometry Thermal Ionization Mass Spectrometry Drip Water Archaeological Find Thermal Ionization Mass Spectrometer
Archaeological carbonate speleothems: Deposits of secondary carbonate minerals (usually calcite or aragonite consisting of CaCO3) that formed in caves containing artefacts or other evidence of past human or hominid occupation
U-series dating such as U-Th (or 230Th/U) is a well-established, accurate, and precise method with a wide range of applications in earth sciences, palaeoclimate research, and archaeology (Bourdon et al. 2003). U-series dating can be used to determine the age of materials that formed with a well-constrained radioactive disequilibrium between U and its daughter isotopes, including precipitates of calcium carbonate (CaCO3) in cave environments, i.e., speleothems like stalagmites, stalactites, or flowstones. U-series dating is essential for palaeoclimate research using speleothems but is less commonly used for archaeological applications. However, cave environments are often associated with archaeology and archaeological finds. Excavations in caves often exhibit stratigraphic relationships with speleothem formation. Where a stratigraphy can be unambiguously established between archaeology and CaCO3 formation, U-series dating provides a powerful chronological tool and can be applied to constrain minimum and/or maximum ages for associated archaeological finds.
Basic Principles of U-Series Dating of Speleothems
U-Th dating can be applied to materials like secondary CaCO3 that incorporate disequilibrium between U (238U and 234U) and the daughter isotope 230Th at the time of formation. For example, speleothems such as stalagmites or flowstones precipitate from percolating waters entering a cave. The water also contains trace elements dissolved from host rock above the cave, including U, which are subsequently incorporated in the CaCO3. The key for U-Th dating is the difference in solubility between U and Th which leads to elemental fractionation in the percolating water. In contrast to U, Th is largely insoluble and thus not incorporated in secondary carbonates when they form from drip waters in cave environments. Thereafter, 230Th starts to build up until radioactive equilibrium is reached. The return of isotope activities to equilibrium allows quantification of time, i.e., the present 230Th/238U and 234U/238U activity ratios enable calculation of the time since formation.
The reliability of U-Th ages strongly depends on whether the dated material has remained as a so-called closed system, i.e., U and Th isotopes are neither lost nor gained after precipitation, and whether any initial 230Th was present. Presence of initial (or detrital) Th is tested by determining the abundance of common Th (232Th) in the sample. In case of significant amounts of 232Th, a correction for initial 230Th needs to be applied by mathematically subtracting an assumed or measured detrital component or by using “isochron” methods (Ludwig and Titterington 1994).
A single U-Th age does not provide a means of confirming that the dated material represents a closed system. However, there are ways to check reliability of U-Th ages of speleothems with respect to closed-system behaviour. The most rigorous test is to employ the alternative 235U-series chronometer and measure 231 Pa/235U in addition to 230Th/U on the same material (Cheng et al. 1998). However, U-Pa dating requires much larger sample sizes due to the significantly smaller concentrations of 235U relative to 238U (235U/238U = 0.0073) and thus cannot be employed in all cases, especially where sample size is limited. Alternatively, where open-system behaviour cannot be excluded, it is recommended to perform repeat U-Th analyses of coeval samples. If dating results on different subsamples are concordant, the system can be regarded as closed. Stalagmites, flowstones, or other carbonates that form with internal stratigraphic constraints provide an additional possibility to check reliability by comparing ages of subsamples precipitated successively along a growth axis, which must become progressively younger from bottom to top.
U-Series Isotope Measurements and Sample Sizes
The applicability of U-series methods depends on availability of sufficient suitable material for dating. In many cases only small samples of less than 100 mg CaCO3 are available, making the sample size critical. Recent advances in thermal ionization mass spectrometry (TIMS) and multi-collector inductively coupled plasma mass spectrometry (MC-ICPMS) instrumentation and protocols not only led to very high precision measurements of U-series ratios (Andersen et al. 2004; Potter et al. 2005) but also greatly reduced the sample size needed for a precise U-Th age. Currently, sample sizes between 10 and 100 mg are sufficient in most cases (Hoffmann 2008; Hoffmann et al. 2009). Overall, the ability to work on small samples in the range of 10 mg is now the key to many applications in archaeology where only very small sections or thin layers of CaCO3 are found to have direct stratigraphic relations to archaeological finds.
U-Series Age Constraints for Excavations in Cave Environments
U-series dating has been applied to speleothems from archaeological sites for over 40 years employing alpha spectrometry (e.g., Fornaca-Rinaldi 1968; Schwarcz and Blackwell 1983; Blackwell et al. 1983), TIMS (e.g., Shen et al. 2001; Zhao et al. 2001) or more recently MC-ICPMS techniques (e.g., Mercader et al. 2009; Clark-Balzan et al. 2012; Hoffmann et al. 2013). Generally, sediment fillings in caves containing archaeology can often be found on top of flowstone formations or are covered by a layer of flowstone or in some cases both. Flowstone thus formed before or after artefact-hosting sediments accumulated, and U-series dating of such speleothems provides maximum or/and minimum ages for bracketed sediments. Similarly, stalagmites that form locally on top of sediments at a drip site or that are covered by sediment fill also may provide age constraints for the sediments. The stratigraphy between carbonate deposits and archaeology bearing sediments is essential; only carbonates with unambiguous relationships provide diagnostic ages of the target sediments and should be analysed.
For capping flowstones, the lowest layer is closest in age to the underlying sediment, and for underlying flowstones, the uppermost layer. However, the layers in contact with sediments are also most prone to incorporation of silicate detritus (dirt) within the carbonate matrix and therefore do not necessarily yield the best dating results. Thus, in most cases, subsamples of detritus-poor carbonate should be taken from a layer within several mm of the flowstone–sediment boundary. It is generally recommended to target subsamples from the most pristine and cleanest parts of the speleothem. It is also essential to perform repeat analyses of a single growth layer and/or a suite of subsamples along the growth axis to verify the presence of closed-system behaviour. In samples that contain high detrital components, “isochron” methods may need to be applied (Ludwig and Titterington 1994).
The measured 230Th/232Th activity ratios shown in Fig. 1 indicate the degree of contamination with detrital material. Typical bulk earth detritus has a 230Th/232Th activity ratio of 0.8 ± 0.4. The value of 1.5 for the topmost subsample confirms that this part of the section has high dirt content. Consequently, the resulting U-Th isotope data are dominated by the detrital component, and a reliable U-series age cannot be calculated for this subsample. Subsample 2 also shows a high detrital component; however, an age can be calculated for this analysis using a bulk earth Th-correction, although the large correction leads to high dating uncertainty. The results of subsamples 3–6 have 230Th/232Th activity ratios >400 indicating that negligible detritus is included. Calculated ages require little correction and are in stratigraphic order indicating reliable results. In the case of this flowstone, U-Th dating constrains the underlying sediments to be older than 500 ka and the overlying to be younger than 150 ka.
Buried stalagmites, stalactites, or soda straws have also been suggested to provide age constraints for sediment fills in caves (Schwarcz and Blackwell 1992; St Pierre et al. 2009). Speleothem fragments such as broken soda straws can often be found mixed into cave sediments and must have formed prior to the sediment accumulation. U-series ages of these materials thus can provide a maximum age of the sediment in which they are found.
U-Series Age Constraints for Artefacts and Cave Art
In cave environments, artefacts (bones, tools) or archaeologically relevant sections such as painted cave walls may become exposed to percolating waters and subsequently covered by secondary carbonates. In most cases, only very thin layers of CaCO3 are found on artefacts, which poses a major restriction for application of U-series dating. Reliable U-series ages can only be obtained where sufficient amounts of CaCO3 can be removed from the artefacts and where the CaCO3 has unambiguous stratigraphic relations to the artefacts. Since the calcite formation postdates the archaeology, its age provides a minimum age constraint. For example, Frank et al. (2002) constrained the age of Trojan artificial water-supply tunnels by dating calcite that formed on the tunnel walls. Fu et al. (2008) used calcite coatings to constrain the age of human skeleton fragments from northeastern China.
Summary and Conclusions
In cave environments, archaeologically relevant sections or artefacts can be found with stratigraphic relations to secondary calcite deposits which can be dated by U-series methods. State-of-the-art U-series techniques minimize sample sizes and thus allow selection of the most suitable calcite formations to provide many possibilities for employing this technique in archaeological contexts. The stratigraphic relations between calcite formation and archaeology need to be unambiguous in order to use U-series ages of calcite to constrain the age of archaeological finds and provide minimum or maximum ages.
- Bischoff, J., Garcia Diez, M., Gonzalez Morales, M. R., and Sharp, W., 2003. Aplicacion del metodo de series de uranio al grafismo rupestre de estilo paleolitico: el caso de la cavidad de Covalanas. Veleia, 20, 143–150.Google Scholar
- Bourdon, B., Henderson, G. M., Lundstrom, C. C., and Turner, S. P., 2003. Uranium-series Geochemistry. Washington, DC: Mineralogical Society of America. Reviews in Mineralogy and Geochemistry, Vol. 52.Google Scholar
- Clark-Balzan, L. A., Candy, I., Schwenninger, J. L., Bouzouggar, A., Blockley, S., Nathan, R., and Barton, R. N. E., 2012. Coupled U-series and OSL dating of a Late Pleistocene cave sediment sequence, Morocco, North Africa: significance for constructing Palaeolithic chronologies. Quaternary Geochronology, 12, 53–64.CrossRefGoogle Scholar
- Schwarcz, H. P., and Blackwell, B., 1992. Archaeological applications. In Ivanovich, M., and Harmon, R. S. (eds.), Uranium-series Disequilibrium: Applications to Earth, Marine, and Environmental Sciences. Oxford: Oxford University Press, pp. 513–552.Google Scholar