Relative pollen productivity estimates of major anemophilous taxa and relevant source area of pollen in a cultural landscape of the hemi-boreal forest zone (Estonia)

Abstract

Estimates of relevant source area of pollen (RSAP) and relative pollen productivity (PPE) are critical parameters for quantitative reconstructions of past vegetation and land cover. This study provides estimates for PPE relative to Poaceae for ten taxa, characterizing the cultural landscape of south Estonia and the RSAP for 40 lakes with an average radius of approximately 100 m (22–274 m, average 101 m) in the region. We evaluate the effects on those estimates of various combinations of factors, such as the analytical methods (i.e. three Extended R value (ERV) sub-models), the distance-weighting methods used to determine plant abundance, and alternative classification schemes of vegetation and land cover data around study sites. Different combinations of ERV sub-models and distance-weighting methods lead to estimates of RSAP varying between 1500 m and 2000 m. The differences in the estimated RSAP are strongly related to the patch size of the vegetation data. According to ERV sub-model 1, which shows the highest log-likelihood among the three sub-models, most of the taxa have PPE that are higher ( Picea, Pinus, Quercus), slightly higher ( Salix, Artemisia, Filipendula) or similar ( Betula, Cerealia, Cypreraceae) to that of Poaceae. The three ERV sub-models produce only slightly different PPE. However, the selection of distance-weighting method for vegetation has considerable influence on the PPE values. The inverse distance-weighting methods, which do not consider inter-taxonomic differences in pollen dispersal, tend to underestimate PPE for taxa with heavy pollen grains and overestimate PPE for taxa with light pollen grains, compared to the results obtained using other methods that consider taxon specific pollen-dispersal. General land-cover data, based on the classification scheme similar to the CORINE (COoRdination of Information on the Environment) database, could be used for estimating the RSAP and PPE, although some potential problems remain when the classification is too simplified and consolidated.