Potential Pitfalls of Pollen Dating

Abstract

Pollen extracted from ocean and wetland sediments cored from the eastern Canadian Margin, James Bay region, and Atlantic provinces of Canada have been radiocarbon dated and results are reported here. Pollen dates from ocean sediments were compared with marine carbonate (mollusk shells or foraminifera) dates from the same core levels, dates for which validity was assessed via correlations with other cores, and for which reworking has been excluded. Pollen samples from 3 tidal wetlands were taken from levels dated with 137Cs and 210Pb profiles. Pollen dates from 2 additional wetlands were compared with 14C dates of botanical macrofossils. Most pollen dates disagree with 14C dates based on macrofossils or carbonates, with age differences typically exceeding 250 yr and reaching 4000 yr in one instance. In some cores, pollen dates show age reversals. Significant proportions of reworked pollen grains in ocean and wetland samples are associated with pollen dates that are too old. Prolonged core storage could result in pollen 14C ages that are too young, possibly because of growth of fungi or other microbes, but more work is needed to verify this hypothesis. Despite the problems we encountered, some pollen dates are consistent with other 14C dates from the same core levels, suggesting this dating method can work, but at present, more work is needed to understand the conflicting results obtained.