Improving the simulation of stand structure in a forest gap model

There is currently great interest in improving the applicability of forest gap models to changing environmental conditions, in order to facilitate the assessment of possible impacts of climatic change on forest ecosystems. Moreover, for the development of mitigation strategies, it is necessary to include forest management options in the models. Both the simulation of transient effects of climatic change and of forest management regimes require a realistic representation of stand structure in gap models, since tree species respond to variations in stand density in characteristic ways, depending on their ecological strategies. In this study, we compared the effect of five different height growth functions that are sensitive to stand density on simulated stand structure of the FORSKA forest gap model. We used long term observation data from a beech thinning trial at Fabrikschleichach, Bavaria, to test the alternative functions. First, we compared simulation results of the original FORSKA model with measured stand development from 1870 to 1990. Whereas simulated stand level variables (e.g. biomass, mean diameter and height) showed good correspondence with observations, individual tree dimensions and simulated stand structure were quite unrealistic. After calibrating parameters of the height growth functions with data from a lightly thinned plot at Fabrikschleichach, we ran the model with data from a heavily thinned plot for validation. All five functions considerably improved the simulation of height/diameter relationships and stand structure. However, there were distinct differences between functions. The best correspondence with measurements was shown by a function which uses the relative radiation intensity in the centre of a tree crown as an indicator of the competition status of the tree. This function is rather simple and needs only two growth parameters, which can be derived for different functional types of species, according to their shade tolerance. With the new, flexible height growth function it should be possible to extend the applicability of gap models to more realistic simulation experiments including forest management and natural disturbances. To our knowledge, this was the first attempt to employ long term forest observation data for the calibration and validation of a forest gap model. The results suggest that such data could be very useful in model testing and improvement. (C) 1997 Elsevier Science B.V.