Aim To assess the extent to which climate change might cause changes in potential natural vegetation (PNV) across Europe.
Method We parameterized a generalized dynamic vegetation model (LPJ-GUESS) for the most common European tree species, and, for the first time, modelled large-scale vegetation dynamics using a process-based model explicitly representing tree species, age cohorts, gap dynamics and biogeochemical cycles in a single framework. For projections, the model was driven with climate scenario data from two atmosphere–ocean general circulation models (AOGCMs), downscaled to 10 × 10′ spatial resolution (c. 18.5 × 12 km at 50° N).
Results At a general level, modelled present-day PNV corresponded better with an expert reconstruction of the PNV than most earlier plant functional type (PFT)-based simulations, but at a finer scale the model and the expert map showed substantial discrepancies in some areas. Simulations until 2085 showed considerable successional shifts in vegetation types in most areas: 31–42% of the total area of Europe was projected to be covered by a different vegetation type by the year 2085. In the long term, equilibrium changes are substantially larger: simulations with one climate scenario suggest that 76–80% of the European land surface could exist within another PNV if climate was stabilized by the end of the century and vegetation had unlimited time to achieve equilibrium with the new climate. ‘Hotspots’ of change include arctic and alpine ecosystems, where trees replace tundra in the model, and the transition zone between temperate broad-leaved and boreal conifer forest. In southern Europe, the model projected widespread shifts from forest to shrublands as a result of drought.
Main conclusions The model presents a considerable advance in modelling dynamic changes in natural vegetation across Europe. Climate change might cause substantial changes in PNV across Europe, which should be considered in the management of reserves and forestry.