Variation in 14C age of macrofossils and different fractions of minute peat samples dated by AMS

Accurate C-14 dating of peat samples is crucial if peat stratigraphics are to be used as historical archives. To analyse the variation in C-14 age within small peat samples, the C-14 age of different mire plant remains and peat fractions (n = 39) from 2 cm thick peat slices was determined using accelerator mass spectrometry. The C-14 ages of mosses were assumed to be the most accurate. To verify this we also tested the possibility that uncertainties could be introduced by mosses at the mire surface reassimilating CO2 produced during the decomposition of old peat. The average C-14 content of living mosses (110.5-110.9 absolute modem % (pM)) did not differ between three Sphagnum species, representing hummock-, carpet- and lawn-growing species, Their C-14 contents were also identical, within experimental limits, to the current C-14 content of the atmosphere. Depending on sampled depth and peat type, the differences in calibrated C-14 age between different fragments or fractions, within specific 2 cm thick peat samples, varied between 365 and similar to 1000 years, with one exceptional deviation of 2125 years in a sample from a mesotrophic sedge peat. Alkali-treated peat samples always gave greater ages than the corresponding untreated peat samples and small-size fractions (<0.045 mm), indicating that the fraction removed by standard alkali treatment is very similar to the small-size fraction. When compared to alkali-treated moss fragments, most other fractions and fragments deviated considerably. Even alkali-treated bulk samples deviated from, the alkali-treated moss fragments. We propose that Sphagnum plant fragments should be used in C-14 AMS dating of peat since they yield the most reliable C-14 dates.