Sie sind hier: Die Projekte / Zeitliche und räumliche Differenzierung der TBK / Projektbeschreibung
Zeitliche und räumliche Differenzierung der Trichterbecher Kultur


The constructions of chronologies

The only way to construct the precise chronologies needed to synchronise the evolution of the different micro-regions and connect these to the archaeo-botanical and limnological evidence is to use several 14C dates in models constructed using Bayesian (e.g. Bronk Ramsey 1995, or Buck et al. 1996) or classical statistics (Van den Bogaard et al. 2002) or wiggle matching, synchronising using the shape of the calibration curve, as in Sarnthein et al. 2007 or Staubwasser et al. 2002.

The obvious question, then, is how many samples are needed to build a chronology to the precision needed and at which point do extra measurements bring little to the overall model. The answer to this question is divided in two cases, sporadic deposition with some ordering such as archaeological sites, continuous deposition, although not necessarily at a constant rate, such as lake sediments, peat bogs etc.

The period under study corresponds to 5100 to 3100 14C years BP with an emphasis in the later part from ca 4600 14C years BP. Experience tells us that one can expect a 1σ measurement uncertainty varying from 25 to 35 14C years for samples containing from 8 to 1 mg C respectively. The models below therefore assume a 1σ measurement uncertainty of ± 28 14C years. They are based on the approach of Bayliss et al 2007 where a hypothetical site or core is imagined in calendar years, transposed via the calibration curve into 14C ages with a measurement uncertainty of 28 14C years. These hypothetical 14C results are then put into a bayesian model calculated using OxCal 4.0. The comparison between the result of the model and our assumptions can then give us an indication regarding the number of ideal samples needed during a certain period, given the 14C calibration curve, to reach a good precision.

Sample preparation of archaeological 14C samples

Sample preparation is a key issue in radiocarbon dating. The intention of the sample preparation is to remove non-contemporaneous materials (particulate materials such as rootlets and fibres; chemical compounds such as carbonates, fulvic- and humic acids, and conservational agents) by avoiding additional contamination from the treatment itself. Common sample materials for radiocarbon dating are organic macrofossils such as plant remains or charcoal, bone samples and pottery (temper or food crusts). While organic macrofossils such as plant remains usually allow an unambiguous dating, radiocarbon dating of bone and pottery is more challenging in terms of interpretation due to possible reservoir effects. A reservoir effect is when the carbon incorporated in an organism does not reflect the 14C content of the atmosphere but has an admixture of older, 14C depleted, carbon, making 14C age of the organism older than it should be from its real age. This is found in marine and lake organisms such as fish and is transmitted to humans when they eat fish for example. This is also very strong in snail shells as some snails use carbonate dust to digest their food. The same has been at different degrees observed in eggs.

Most 14C samples (e.g. seeds, wood, charcoal, plant remains) undergo a simple acid-alkali-acid treatment. The first acid treatment removes the carbonate. The acid is washed with aqua dest. and an alkali solution is applied (usually NaOH). This dissolves a mobile contamination, humic acids, which can be introduced in ground water. After washing with aqua dest, the sample is treated with acid again to remove atmospheric carbon dioxide that might have dissolved in the sample during the alkali treatment. The sample is washed again but kept slightly acidic to prevent dissolution of atmospheric CO2 and dried.

Wood sample found in a peatland
Wood sample found in a peatland

Charcoal and wood samples

We will determine the species of charcoal and wood samples to select short lived species or the part of a charcoal piece close to the bark (when the tree fell) when possible, to avoid the “old tree effect” where one dates rings from several years, sometimes decades, prior to when the tree was cut and used for e.g. construction in the hope of dating the construction itself.

Collagen extraction from bone samples

Bone sample with a yardstick
Chemicals and tools for the collagen extraction

The radiocarbon dating of bones is usually performed on the collagen fraction, which is purified by dissolution from the demineralised bone as gelatine in hot acidic water (hydrolysis, e.g. Longin 1970) after the demineralised bone has been treated with alkali to remove mobile fractions such as humic acids and degraded protein fragments (Van Klinken and Mook, 1990).

Two different ultrafilters for cleaning the collagen
Two different ultrafilters for cleaning the collagen

Ultrafiltration (UF) has been suggested as an effective method to separate low molecular weight contaminants and degraded proteins from the high molecular and unchanged proteins of the collagen (e.g. Brown et al 1988, Bronck Ramsey et al. 2004). This cleaning method for extracted fossil collagen has received considerable attention (e.g. Mellars 2006), as it was shown that redating previously dated bones using UF resulted, in many cases, in 14C ages that were significant older and more consistent (e.g. Bronck Ramsey 2004, Higham et al. 2006). However, comparative dating studies also found evidence for contamination introduced by the filters used (Bronk Ramsey et al., 2004, Brock et al., 2007, Hüls et al. 2007, Hüls et al, subm.).

Radiocarbon dating of bone apatite

In some circumstances, the preferred collagen fraction may not be available in a sufficient amount. Bones, which were burned at high temperatures (> 650 C, cremated bones), do not contain organic material nor collagen. Alteration of bone material due to oxidative processes or microbiological decomposition will may also degrade the original organic bone material (e.g. bones buried in hot dry or very humid environments). In those cases, the only carbon originating from the animal/human available is stored in the mineral fraction of the bone, the apatite.

So-called structural carbonate (~ CO32-), which is formed in the organism by metabolic processes, will be partially substituted for PO43- in the apatite during biomineralisation and survives, although at reduced amounts, the burning process and permits therefore the 14C dating of the apatite of burned bones. The formation of larger apatite crystalls and compaction during loss of CO2 and water during cremation is assumed to preserve an almost unchanged isotope signal of the structural carbonate (van Strydonck et al 2005).

Energy of Infra-Red (IR) light is in the range of the vibrational energy of the chemical bindings of molecules, e.g. adsorption of the electromagnetic energy leads to a stimulation of molecule vibration at specific amounts. The adsorbed energy is thus characteristic of specific binding types and materials making which makes IR an ideal tool for identifying different materials. The measurement of bone powder or demineralised bone material allows the immediate recognition of cremated bones, organic preservation (e.g. collagen content and quality), and amount of secondary calcites, which is needed to evaluate quality of radiocarbon dating of bone apatite. We will assess the degree of preservation of all bone samples with Fourier Transform Infra Red (FT-IR) spectrometry before preparation to select the samples that are likely to give good results.

IR spectra of a bone sample
IR spectra of a bone sample
Soxhlet extractor

Material treated with conservation agents

Finds that are kept in museums are sometimes treated with conservation agents to ensure their preservation. These agents acts as contamination for 14C dating and must be removed. These samples will be subjected to a soxhlet type serial extraction (So) to remove fatty and waxy organic contaminants. In sequence, the samples are extracted three times each with boiling tetrahydrofurane (THF), chloroform, petroleum-ether, acetone, and methanol, and then rinsed with demineralised water. This elutrope sequence was designed to remove most conservation agents commonly used. If the effectiveness of this cleaning procedure is in doubt, the material treated once or twice with solvents, measured, treated again one or two times and measured again, if the results do not agree, the process is repeated until the results are stable.