Estimating relevant source area of pollen for small Danish lakes around AD 1800

The aims of this study were to obtain estimates of the relevant source area of pollen (RSAP) that can be used to interpret fossil pollen diagrams from medium sized lakes, and to elucidate factors affecting RSAP for such lakes. A data set of sediment pollen assemblages dating from AD 1800 from 25 Danish lakes (3-27 ha) and land cover around the same lakes from historical maps was used. The plant abundance data were distance weighted using a species-specific model of pollen dispersal/deposition and other distance weighting functions (1, Ild and lId²). Extended R-value models were applied to evaluate the relationship between pollen and plant abundance and to estimate RSAP. The choice of distance weighting function influenced the RSAP estimates. When using the species-specific model, wind speed and species-specific properties of pollen dispersal had little effect on the RSAP estimates, which were approximately 1700 m in radius, when all lakes were analysed together. When the pollen types were classified into fewer groups, the RSAP estimate increased, probably because this procedure affected the spatial pattern of the plant abundance data by, in effect, increasing patch size. When the lakes were classified into two distinct groups (western Jutland and remaining sites) using a multivariate statistical technique (TWINSPAN), we obtained different RSAP estimates for each of the two groups. This result could be explained by a difference in average patch size of the land cover types between regions. A series of simulations supported our interpretation that the patch size was the main factor determining the size of the RSAP for the Danish lakes at AD 1800.