LAUSR

Co-ordinating Editor: Thomas Giesecke Abstract Questions: What are the latitudinal diversity patterns in Scandinavia over the last 8 ka? Have they been consistent over time? What are the longterm diversity trends at different sites? Is there a consistent rateofchange between sites? Location: Thirty sites in Norway, two sites in northwest Sweden. Methods: Pollen data from these 32 lakes. The sites lie within six of the vegetation zones of Moen (1998). The data are used to estimate pollen diversity using Hill′s effective species numbers N0, N1 and N2. Diversity relationships in time and along latitude are assessed using ANCOVA. Linear regression is used to investigate siteindependent longterm trends. Patterns in the regression slopes are investigated between vegetation zones and along the latitudinal gradient. Results: The latitudinal pattern of decreasing diversity with increasing latitude existed over the last 8 ka, although for the number of effectively common (N1) and abundant (N2) pollen taxa the relationship was slightly weaker between 8–6 ka. Longterm trends show that diversity increases at most sites over time. However, the magnitude and rateofchange vary between sites. Highest rateofchange in N0 over time is observed in southern sites, although the spatial relationship of rateofchange is not statistically significant. Conclusions: The constant increase in diversity over time may reflect additional taxa arriving and changes in abundances. Most sites change in the same direction, but at different rates and magnitudes. There is no clear spatial pattern in rateofchange, and the general patterns of increasing diversity over time are consistent with the hypothesis of postglacial (Holocene) dispersal limitation. Drivers of these changes have not been rigorously investigated, but our results suggest that environmental and historical processes are not mutually exclusive. During the last 2–1 ka BP diversity patterns have probably been increasingly affected by human land use.