Explaining forest composition and biomass across multiple biogeographical regions

Current scientific concerns regarding the impacts of global change include the responses of forest composition and biomass to rapid changes in climate, and forest gap models have often been used to address this issue. These models reflect the concept that forest composition and biomass in the absence of large‐scale disturbance are explained by competition among species for light and other resources in canopy gaps formed when dominant trees die. Since their intiation 25 yr ago, a wide variety of gap models have been developed that are applicable to different forest ecosystems all over the world. Few gap models, however, have proved to be equally valid over a wide range of environmental conditions, a problem on which our work is focused. We previously developed a gap model that is capable of simulating forest composition and biomass in temperate forests of Europe and eastern North America based on a single model structure. In the present study, we extend the model to simulate individual tree species response to strong moisture seasonality and low temperature seasonality, and we modify the widespread parabolic temperature response function to mimic nonlinear increases in growth with increased temperature up to species‐specific optimal values. The resulting gap model, ForClim V2.9, generates realistic projections of tree species composition and biomass across a complex gradient of temperature and moisture in the Pacific Northwest of the United States. The model is evaluated against measured basal area and stand structure data at three elevations of the H. J. Andrews LTER site, yielding satisfactory results. The very same model also provides improved estimates of species composition and stand biomass in eastern North America and central Europe, where it originated. This suggests that the model modifications we introduced are indeed generic. Temperate forests other than those we studied here are characterized by climates that are quite similar to the ones in the three study regions. Therefore we are confident that it is possible to explain forest composition and biomass of all major temperate forests by means of a single hypothesis as embodied in a forest gap model.