Substrate dependency of Lateglacial forests in north‐east Germany: untangling vegetation patterns, ecological amplitudes and pollen dispersal in the past by downscaling regional pollen

Aim  Palynology has revealed that during the Lateglacial Allerød period, Pinus and Betula dominated the forests of north‐east Germany. Because of implicit restrictions, however, palynology fails to reconstruct in detail the distributional patterns and whether monospecific or mixed forests prevailed. Here we test the hypothesis that the distributional patterns of Pinus and Betula were largely determined by substrate. Location  Fifteen sites in north‐east Germany. Methods  As pollen data reflect the abundance of taxa around a sample site, our hypothesis implies that the abundance of e.g. P inus pollen should reflect the abundance of substrate types that are favoured by Pinus . (In order to differentiate clearly between taxa and pollen types, the latter are displayed in small capitals. ) We analysed the percentage of P inus and B etula pollen from an interval in the Allerød, and their relationship to the present‐day proportion of the two dominant substrate types (sand and boulder clay). The substrate proportion was determined in rings from 1 to 50 km around all sites, using four distance‐weighting functions. Results  Pollen percentages of P inus are linearly related to the proportion of sand, and B etula percentages to the proportion of boulder clay. The highest coefficients of determination ( r 2  = 0.89 and 0.91, respectively) were observed for radii of c. 30 km, with distance weighting by 1 and 1/ d . Main conclusions  The present‐day distributional pattern of substrate types strongly indicates the distributional pattern of Pinus and Betula in north‐east Germany in the Allerød. Assuming that the pattern of substrates remained broadly constant, Pinus dominated on sand and Betula on boulder clay. Our new method thus enables us to refine significantly the reconstruction of habitat characteristics and distributional patterns of taxa in the Lateglacial and Holocene, independent of their present‐day ecological amplitudes. The good performance of the distance‐weighting functions 1 and 1/ d indicates that the pollen source area of a site has rather discrete boundaries determined by convective air movements during daytime. Within these boundaries, pollen is evenly distributed over various distances. This implies that light pollen types are transported further than predicted by Sutton’s equation on particle dispersal.