Detecting human impact in the pollen record using data-model comparison

Palaeoecological data are compared with output from climate-driven forest simulation models to separate human influence as a driver of vegetation dynamics from other drivers such as climatic change. The transition from Tilia cordata to Fagus sylvatica dominance in a small forest hollow in Denmark was not predicted by a climate-driven forest simulation model and could be ascribed to anthropogenic impact. This transition can be upscaled to a large region of north-west Europe and contributes to a data-model mismatch for the European distribution of Fagus 6,000 years ago. A data-model mismatch for Picea abies during the last few centuries in southern Scandinavia can also be attributed to anthropogenic impact. Combining pollen data and vegetation models can help with the important task of upscaling from the scale of the forest stand, where anthropogenic impact is readily detectable, to regions and continents, where it is more challenging to distinguish anthropogenic impact from the impacts of climatic change.