THE APPLICATION OF POLLEN RADIOCARBON DATING AND BAYESIAN AGE-DEPTH MODELING FOR DEVELOPING ROBUST GEOCHRONOLOGICAL FRAMEWORKS OF WETLAND ARCHIVES

ABSTRACTWetland sediments are valuable archives of environmental change but can be challenging to date. Terrestrial macrofossils are often sparse, resulting in radiocarbon (14C) dating of less desirable organic fractions. An alternative approach for capturing changes in atmospheric 14C is the use of terrestrial microfossils. We 14C date pollen microfossils from two Australian wetland sediment sequences and compare these to ages from other sediment fractions (n = 56). For the Holocene Lake Werri Berri record, pollen 14C ages are consistent with 14C ages on bulk sediment and humic acids (n = 14), whilst Stable Polycyclic Aromatic Carbon (SPAC) 14C ages (n = 4) are significantly younger. For Welsby Lagoon, pollen concentrate 14C ages (n = 21) provide a stratigraphically coherent sequence back to 50 ka BP. 14C ages from humic acid and >100 µm fractions (n = 13) are inconsistent, and often substantially younger than pollen ages. Our comparison of Bayesian age-depth models, developed in Oxcal, Bacon and Undatable, highlight the strengths and weaknesses of the different programs for straightforward and more complex chrono-stratigraphic records. All models display broad similarities but differences in modeled age-uncertainty, particularly when age constraints are sparse. Intensive dating of wetland sequences improves the identification of outliers and generation of robust age models, regardless of program used.