Pollen productivity estimates for a pine woodland in eastern Scotland: The influence of sampling design and vegetation patterning

Abstract

Quantitative vegetation reconstruction models have been developed to allow more faithful accounts of changes in vegetation structure to be attained. Widespread application of these models is encouraged, however for this to be realised necessary parameter values, such as pollen productivity estimates require further investigation. This study shows results of tests undertaken to detect the influence of a homogenous vegetation structure and sampling strategy upon the calculation of PPEs from contemporary data. Using a collection of contemporary moss samples, sub-groups were created using defined sampling strategies (random or systematic) and nearest neighbour distance thresholds to investigate the influence of these conditions upon PPE generation. Results suggest that a random sampling strategy achieves the most reliable values, but using a systematic strategy is possible. Furthermore, in areas where vegetation patterning is non-stationary, a large degree of variation is incorporated into the values obtained, and the use of quantitative models in such situations could be problematic.