Software

Radiocarbon dating is based on the radioactive decay of 14C within a sample. High energy cosmic rays impinging on the stratosphere form14C from14N in the following reaction:

14_N+n→14_{C+p (3.9)}

where n is a neutron and p a proton. 14C is rapidly oxidised to 14CO2. Global at- mospheric circulation leads to the mixing of atmospheric14C around the globe, where it may be photosynthesised by plants and incorporated into plant tissue (Libby, 1952). The concentration of 14C within plant tissue remains in equilibrium with the atmo- sphere whilst the plant is living. Upon the death of the plant, the exchange of14C with

the atmosphere ceases, and the14C content of the dead plant tissue declines over time as the14C decays to14N as follows:

14_C→14_{N+β (3.10)}

where β is a beta particle (an electron). The half life of 14C was originally as- sessed assessed as 5568±30 years (Libby, 1952), though subsequent measurements revealed it to be closer to 5730 years (Goodwin, 1962). The established convention for radiocarbon age calculations is to use the “Libby” half life of 5568 years, necessitat- ing a correction in conversion to calendar years which is typically carried out during a calibration process (Stuiver and Polach, 1977; Bradley, 1985). Measurement of the radiocarbon content can be made by measuring the radioactive decay of14C using liq- uid scintillation counting (Bradley, 1985). In this process, the carbon of the sample is converted to benzene, mixed with a scintillation liquid and placed in a scintillation counter where β particles (electrons) emitted during the decay process are counted. Their rate of emission is proportional to the radiocarbon content of the sample allow- ing an estimate of the radiocarbon content and hence radiocarbon age of the sample to be made.

A radiocarbon age does not, however, equal the age of the sample in sidereal years. This discrepancy arises partly because the Libby half life for 14C not equal the true half life of this isotope, but more importantly, from variations of atmospheric 14C concentrations through time. Such variation departs from the assumption of uniform 14_{C concentrations (or alternatively the} 14_C/12_{C ratio) through time made in calcu-} lating the conventional radiocarbon age. To correct for this, a calibration procedure must be undertaken to convert conventional radiocarbon ages to calender ages (Stuiver and Suess, 1966). Version 3.8 of the Oxcal radiocarbon calibration software (Bronk- Ramsey, 2001) which uses atmospheric 14C data contained in Stuiver et al. (1998) derived from dendro-chronologically dated wood extending back to 11850 years BP

has been used in this study to calibrate conventional radiocarbon ages.

3.9.1.1 Dating Fluvial Deposits with Radiocarbon From Detrital Charcoal

Radiocarbon dating is often applied to detrital charcoal fragments found within fluvial sedimentary deposits in order to estimate the time of deposition of the sediment body. The radiocarbon age of charcoal from a sedimentary deposit may not correspond to the age of the depositional event due to a number of factors. Radiocarbon dating essentially measures the time elapsed since photosynthesis ceased on the plant from which the charcoal was derived. The time elapsed between death of the plant tissue, combustion to charcoal, transport through a catchment in a river system and burial within a fluvial deposit will vary according to any number of environmental factors, potentially giving charcoal found within a modern fluvial deposit some non-zero age.

The critical issue becomes the magnitude of this “in-built” age relative to the true age of the sample. Gavin (2001) found ages for charcoal fragments within a con- temporary soil horizon ranged from 0 to 670 14C years for samples from Vancouver Island of western Canada. Additionally and specific to the fluvial geomorphology con- text, erosion and re-deposition of ancient fluvial sediments containing “aged” charcoal fragments by the contemporary river also has the potential to give non-modern ages for what are essentially modern fluvial deposits (Blong and Gillespie, 1978).

Aside from these “time to deposition” effects just described, the movement of car- bon within a sedimentary body must also be addressed as enrichment of14C to a gen- uinely ancient and hence 14C depleted sample will contribute to an underestimation of the true sample age. Carbonate compounds may be mobilised within groundwater bodies, and variations in aquifer levels due to hydrologic changes such as seasonal pre- cipitation cycles or secular changes can lead to multiple wetting and drying episodes of floodplain sediments with concomitant dissolution and deposition of groundwater- borne carbonate compounds upon charcoal fragments. Additionally, humic and fulvic

compounds sourced from decomposition of organic material growing upon overlying ground surfaces can be dissolved by rainwater and then leached downwards through the floodplain sediments, again with the possibility of absorption or precipitation upon buried charcoal within underlying sedimentary horizons (Goh and Molloy, 1979; Gille- spie et al., 1992; Gillespie, 1997).

Laboratory pre-treatment methods are employed to remove such contamination. An acid-base-acid pre-treatment has been used here for treating charcoal fragments and involves boiling the sample in dilute hydrochloric acid to remove carbonate com- pounds, rinsing with copious amounts of de-ionised water followed by a second boil- ing step in sodium hydroxide to remove humic and fulvic acid compounds followed again by copious rinsing in de-ionised water (Olsson, 1979). Exchange with atmo- spheric carbon dioxide gas can occur during the sodium hydroxide boiling phase and to remove this contamination, a final boiling step using dilute hydrochloric acid is un- dertaken. The sample containing the purified charcoal is then rinsed and oven dried before the initial combustion step of radiocarbon dating. Radiocarbon analyses in this study were performed by Abaz Alimanovic and Damien Kelleher of the Radiocarbon Dating Laboratory at the Australian National University.