Assessing the potential future of current forest stands is a key to design conservation strategies and understanding potential future impacts to ecosystem service supplies. This is particularly true in the Mediterranean basin, where important future climatic changes are expected. Here, we assess and compare two commonly used modeling approaches (niche-and process-based models) to project the future of current stands of three forest species with contrasting distributions, using regionalized climate for continental Spain. Results highlight variability in model ability to estimate current distributions, and the inherent large uncertainty involved in making projections into the future. CO 2 fertilization through projected increased atmospheric CO 2 concentrations is shown to increase forest productivity in the mechanistic process-based model (despite increased drought stress) by up to three times that of the non-CO 2 fertilization scenario by the period 2050-2080, which is in stark contrast to projections of reduced habitat suitability from the niche-based models by the same period. This highlights the importance of introducing aspects of plant biogeochemistry into current niche-based models for a realistic projection of future species distributions. We conclude that the future of current Mediterranean forest stands is highly uncertain and suggest that a new synergy between niche-and process-based models is urgently needed in order to improve our predictive ability.

No abstract

Environmental classification addresses issues involving the representation and analysis of continuous and variable ecological data. This study creates a methodology to define topo-climatic landscapes (TCL) in the north-west of Catalonia, which is situated in the north-east of the Iberian Peninsula. TCL provide data regarding the ecological behavior of a landscape in terms of its topography, physiognomy, and climate, which are the main drivers of an ecosystem. The variables selected are derived from a variety of different sources, such as remote sensing and climatic atlases. The methodology employed combines unsupervised iterative cluster classification with supervised fuzzy classification. Twenty eight TCL, which can be differentiated in terms of their vegetation physiognomy and vegetation altitudinal range type, were selected for the study area. Furthermore, a hierarchy among the TCL is established which permits the merging of clusters and allows for changes in thematic resolution. By using the topo-climatic landscape map, managers can identify patches with similar environmental conditions and at the same time assess the uncertainty involved in classification.