Climate change and European forests: what do we know, what are the uncertainties, and what are the implications for forest management?

The knowledge about potential climate change impacts on forests is continuously expanding and some changes in growth, drought induced mortality and species distribution have been observed. However despite a significant body of research, a knowledge and communication gap exists between scientists and non-scientists as to how climate change impact scenarios can be interpreted and what they imply for European forests. It is still challenging to advise forest decision makers on how best to plan for climate change as many uncertainties and unknowns remain and it is difficult to communicate these to practitioners and other decision makers while retaining emphasis on the importance of planning for adaptation. In this paper, recent developments in climate change observations and projections, observed and projected impacts on European forests and the associated uncertainties are reviewed and synthesised with a view to understanding the implications for forest management. Current impact assessments with simulation models contain several simplifications, which explain the discrepancy between results of many simulation studies and the rapidly increasing body of evidence about already observed changes in forest productivity and species distribution. In simulation models uncertainties tend to cascade onto one another; from estimating what future societies will be like and general circulation models (GCMs) at the global level, down to forest models and forest management at the local level. Individual climate change impact studies should not be uncritically used for decision-making without reflection on possible shortcomings in system understanding, model accuracy and other assumptions made. It is important for decision makers in forest management to realise that they have to take long-lasting management decisions while uncertainty about climate change impacts are still large. We discuss how to communicate about uncertainty - which is imperative for decision making - without diluting the overall message. Considering the range of possible trends and uncertainties in adaptive forest management requires expert knowledge and enhanced efforts for providing science-based decision support.

Application of volume growth and survival graphs in the evaluation of four process-based forest growth models.

Volume growth and survival (VGS) graphs, which show volume growth rate and risk of mortality for individual trees (or tree size classes), have been proposed as a tool for assessing the validity of models that describe the development over time of tree size distributions within forest stands. We examined the utility of the VGS method in evaluating four process-based models. The performance of the models FORSKA, 4C, MORG, and PipeQual is analyzed against long-term data from a Scots pine stand in Evo, Finland, and the models FORSKA and 4C are also assessed with respect to data from a beech stand in Fabrikschleichach, Germany. Comparison of the measurement-based VGS graphs with those produced from the model-based data shows that although the models yield similar stand-level predictions, they can differ widely in their projections of individual tree growth and size distributions. Examination of the discrepancies between models and data in the context of the VGS graphs reveals several areas in which the models could be improved. We conclude that the method is useful in model evaluation, especially if used in combination with indicators of stand structure, such as the height/diameter ratio.

Volume growth and survival graphs: a method for evaluating process-based forest growth models.

We investigated the relationships within forest stands between tree size and (a) stem volume growth rate and (b) risk of mortality for individual trees. Values of both x and y variables were plotted relative to the largest value in the stand. We refer to the resultant presentations as relative volume growth and relative survival graphs (VGSs). A pair of VGSs can be produced readily from an individual-tree growth model. It can also be constructed from consecutive sets of field measurements. Comparison of VGSs derived from model and measurement data provides a test of the validity of the components of the growth model. We have analyzed VGSs based on measurement data for Scots pine (Pinus sylvestris L.) in central Finland and for beech (Fagus sylvatica L.) in southern Germany. The graphs based on measurement data varied as a consequence of differences in competition, stand management, and tree species. We analyzed the relationship between VGSs and stand dynamics using a simple growth model. We found that different features of the VGSs imply characteristic tree size distributions in subsequent years. Thus, we conclude that if the VGSs generated by a model do not correspond to those based on field measurements, the model cannot be relied on to reproduce the development of tree-size distribution correctly. Relative growth and survival graphs thus provide a tool for evaluating complicated growth models.