Southern Estonia is characterized by the complex topography and soil distribution. Although good spatial-correlations between vegetation and soil have been found, the extent of soil’s role on the long-term vegetation dynamics is little known. We will develop and use modeling approaches to estimate the Holocene vegetation change at a local scale (< 100 ha) and assess factors and mechanisms that have controlled it. Our central hypothesis is that the spatial patterns of vegetation around the Haanja National Park in southern Estonia do not change through time but rather are controlled by that of soil. To test this hypothesis we pursue three specific aims: (1) to quantify the spatial dynamics of the Holocene vegetation using fossil pollen, (2) to model the vegetation dynamics to assess mechanisms that affect its spatial patterns, and (3) to model the effects of climate, soil and topography on the spatial patterns of vegetation in the past and future. For aim (1), we will use a model-based reconstruction method developed by Sugita (2007a,b). For aims (2) and (3), we will develop and use a dynamic individual-based model based on Smith et al. (2001) and assess the roles of soil, topography and climate on the vegetation dynamics. Collectively, we expect to have calibrated and tested the vegetation model, to have determined the mechanisms that affect the spatial dynamics of vegetation, and to have estimated potential impacts of climate change on vegetation around the park. Well-calibrated and tested models for quantitative assessment of the future changes in landscape are currently not available in Estonia. This project will provide such models and is therefore expected to have a positive and significant impact on the basic research and development of Environmental Technology and natural resource management in Estonia. After completion of this project, we will actively disseminate the results through various channels (e.g. RMK–State Forest Management Centre Tallinn). At the same time, the fundamental knowledge on the relationship between long-term spatial dynamics of vegetation and soil, even though widely assumed but little known, is expected to advance plant ecology, paleoecology, and conservation biology.