Das Projekt "Walddynamik bei Klimaaenderungen - Daten und Modelle zur Quantifizierung mittel- bis langfristiger Aenderungen im Oekosystem" wird vom Umweltbundesamt gefördert und von Universität Bern, Systematisch-Geobotanisches Institut durchgeführt. Ecosystem processes in forests such as immigration, expansion and declines of tree species (and thus their competition) are of great importance but difficult to observe because of the time scale involved. We will combine refined paleoecology with forest-development modelling in order to deal with this problem. We will choose specific periods of immigration, expansion and decline in the record of lake sediments and refine the temporal sampling resolution of pollen analysis. Previous work has shown that (with some exceptions) pollen proportions provide for most tree species a reasonably good proxy for the population sizes. As external perturbing factors affecting migration, expansion and declines of trees three types of potential disturbance will be considered: rapid climatic change, fire, and human impact. We will study some lowland and montane sites in the Southern Alps, where the history of climatic changes is known at least in its outlines. We will concentrate on periods of immigration, expansion and decline of trees and provide from identical samples the analyses of pollen, plant macrofossils and charcoal. We will test the hypotheses of climatic change, forest fire, and human impact as controlling factors for vegetation change by randomisation tests such a Monte Carlo permutations. By applying a forest succession model of the FORET family we will mimic the pollen profiles and test whether the selected key elements of forest dynamics (species migration, expansion, decline and competition intrinsic to the forest site) are able to explain the successional patterns exhibited in the pollen profiles. The model will be applied to various grid points (not only pollen sites) in both an environmental but also geographical space. With this sampling design formation of altitudinal belts can be simulated under various climate change scenarios and compared with the pollen flux of the different time windows. In landscapes with high relief the four processes of migration, expansion, decline of tree taxa, and thus their competition control the formation of altitudinal belts of vegetation. Therefore in both areas vegetation change in four dimensions (including time) will be recorded and correlated to perturbations. In addition competition among species can be studied in the Southern Alps by modelling.