Developing robust lake sediment chronologies using 210Pb, Pu and radiocarbon dating of pollen concentrates and macrofossil: A case study from Lake Surprise, Victoria, Australia

The development of reliable sediment chronologies is crucial for accurate interpretations of decadal to century-scale palaeoenvironmental changes in the late Quaternary. Although radiocarbon dating of sedimentary sequences is commonly undertaken, not all the organic fractions are representative of atmospheric 14 C levels, resulting in inaccurate age models. Whilst terrestrial plant macrofossils are widely considered ideal dating material – assuming they are contemporaneous with the horizons being dated – they are often sparse or absent. In this context, radiocarbon dating of pollen extracts is increasingly being used as alternative dating material. Here, we used pollen radiocarbon dating, alongside a suite of macrofossil and bulk sediment dates, to develop a chronology for the Holocene sediments of Lake Surprise, in Victoria, Australia. 210 Pb activity and Plutonium (Pu) concentrations and isotope ratios were also analysed to constrain the age of the uppermost sediments, augmented with recent historical markers, including the first arrival of Pinus pollen and the date of an earlier coring expedition at the site in 2004. With respect to the radiocarbon dates, we found an age offset between the plant macrofossils and bulk sediment dates of 260 ± 86 14 C years and an offset of ∼340 14 C years between plant macrofossil and pollen extracts. In both cases, macrofossil dates appeared to be “younger” than the bulk sediment and pollen dates. The offset between pollen and plant macrofossil dates was found to vary with sediment depth and generally correlate with carbonate concentration in the sediment. Using Fourier transform infrared spectroscopy (FTIR), we determined that the pollen extracts were not contaminated by either carbonate or charcoal. However, contamination by algal spores could not be ruled out, and we hypothesise that those algal spores may have assimilated aged dissolved inorganic carbon during periods of higher groundwater influx, thus altering the measured radiocarbon age of the pollen extract. Macrofossil and corrected pollen radiocarbon dates were incorporated in a Bayesian age-depth model which integrated 210 Pb activities and Pu data and bomb pulse C-14 dates validated using recent historical age markers. Our results suggest that it is possible to generate a robust geochronological framework for Lake Surprise using radiocarbon dating back to at least ∼10,846 cal yr BP. • The Holocene chronology is not well constrained in the South Australian region. • Bayesian Plum modelling minimizes the Pb 210 dating uncertainty by integrating Pu and 14 C data. • Pollen 14 C dating establishes accurate sediment chronologies, when terrestrial macrofossils are sparse. • FTIR analysis of pollen concentrates suggests old 14 C ages,may stem from reworked algal spores as contaminants.