Low temperature (LT) combustion of sediments does not necessarily provide accurate radiocarbon ages for site chronology

Radiocarbon dating of soils and sediments is notoriously problematic for the purposes of dating a specific event due to their heterogeneous mix of multiple organic fractions, each of which may have a different radiocarbon age. Numerous studies have failed to agree on which sedimentary fraction or radiocarbon pre-treatment method, if any, provides the closest agreement between the age of a sedimentary fraction and that of associated plant macrofossils or charcoal. We tested the stepped-combustion method of McGeehin et al. (2001), as well as standard radiocarbon humin and humic extraction techniques, using samples from a chronologically well-constrained perennially-frozen site at Quartz Creek, Yukon Territory, Canada. The ages in closest agreement with associated radiocarbon-dated plant macrofossils and with the overlying Dawson tephra were given by the humic and humin fractions, but even these were still older than the macrofossil ages by up to 4195 +/- 260 radiocarbon years. The low temperature (LT) humin method recommended by McGeehin et al. (2001) yielded ages older than the macrofossils by up to nearly 4425 +/- 240 radiocarbon years. These fractions, while still providing information on the mobility and potential residence times of carbon in soils and sediments, should not be relied upon to provide consistently accurate site chronologies. (C) 2010 Elsevier B.V. All rights reserved.