A comparison of pollen extraction methods: confirmation of dense-media separation as a reliable method of pollen preparation

Palynology is a crucial proxy for understanding Quaternary environmental change. A range of laboratory preparation techniques has been developed to deal with the extraction of pollen in different sedimentary contexts. Here, we present a comparison of the conventional hydrofluoric acid method and the dense-media separation method using sodium polytungstate. We examine pollen, non-pollen palynomorphs (NPPs) and microcharcoal by undertaking parallel preparation and counting of 30 paired samples from three sites in the Middle Atlas Mountains, Morocco, with contrasting environmental and sedimentary conditions. Differences between microfossil counts are assessed visually and statistically, using ANOVA and linear regression. We observe that the typical offset between counts for the five most common pollen and NPPs produced by the two methods (µ = 1.6%, SD = 0.9%) is considerably smaller than the 95% uncertainty associated with the standard counting procedure (µ = 6.8%, SD = 2.7%). There is no statistically significant difference associated with the preparation methods, and no significant deviation from a linear relationship between the results obtained by the two methods. There are advantages and disadvantages to both methods in terms of preparation time and visual characteristics. This study confirms that results for pollen, NPPs and microcharcoal obtained by either preparation method can be directly compared.