Sediment density and the limits to the repeatability of absolute pollen frequency determinations

Abstract

The most commonly encountered measure of absolute pollen frequency (APF), grains per cubic centimeter of sediment, is highly dependent both on sediment density and lithology, as well as upon the sampling technique. This paper presents the results of a study designed to assess the relative contributions to the errors inherent in APF of sampling and variations in sediment density and lithology. A wide range of sediments was made up in the laboratory to simulate the spectrum of natural pollen‐bearing sediments with regard to sediment mineralogy, pH, hydration, and water content. They were sampled with several tools to determine (1) the precision with which different tools can be used to sample pollen‐bearing sediments, and (2) the repeatability of sampling in different lithologies. This study suggests (1) that some sampling techniques are clearly superior to others, (2) that operator errors and pH differences may bias all APF determinations, and (3) that the water content of the sediment is of paramount importance with regard to the precision with which a sediment can be sampled and to the validity of differing types of standardization of APF; this owes mainly to the hydration of organics and clay minerals. When APF can be expressed in “grains per unit area per unit time,” these factors may contribute considerable error to an APF measurement. When time‐stratigraphic control is not sufficient to allow APF to be expressed in these terms, expression of APF in “grains per unit volume” becomes almost meaningless. “Grains per unit dry weight” is far superior, both in theory and in practice.