The circumboreal tundra-taiga interface: late Pleistocene and Holocene changes.

Creating a global perspective on past treeline changes is problematic due to the varying methods and definitions used. A general lack of a detailed description of the modern treeline position and vegetation complicates any comparative analysis of the magnitude of the most important changes. However, one seemingly common factor in most regions was an extremely rapid dispersal of trees when climate warmed drastically from full glacial conditions. Most Arctic treelines reached their northernmost positions in the early Holocene and receded to present positions starting at about 5.8 ka. The early occupation of the northernmost sites in ice-free and early deglaciated areas was possible because of the close proximity of invading trees in nearby glacial refugia, particularly in Fennoscandia and northern Russia. In Canada, the Northwest Territories and Quebec-Labrador were out of phase with this general trend due to their late deglaciation. However, even here colonization was rapid, indicating that the tree species were present adjacent to the glaciers. Following this trend and based on the present evidence, we propose a scenario of a continuous but modest occupation of eastern Beringia by spruce during the late-Pleistocene instead of an exceptionally rapid spread of conifers from the glacial refugium south of the Laurentide ice sheet (2000 to 3000 km in about 200 years), which typically has been assumed. Macrofossil evidence of scattered occurrences of "exotic species" (for instance Siberian larch in central Sweden) far from their natural range limits in the early Holocene highlight the disparity between pollen and macrofossil analyses. It questions the validity of assigned pollen percentages to indicate the presence of a species within a region as these species were not observed in the pollen record. Thus, it is likely that trees were present at any given site well before the rise in pollen abundance. There is still a large potential to improve our knowledge about the environmental history of the circumboreal treeline areas. In particular, future research should concentrate not only on patterns of species displacement, but on finding the factors, apart from climate, which cause treeline shifts.