SIMULATION OF FOREST STAND DYNAMICS, USING REAL-STRUCTURE PROCESS MODELS

The dynamic richness of forest ecosystem response to climate, management and pollution stress cannot be adequately described by traditional descriptive growth-and-yield simulators, but requires sufficiently detailed models of the underlying eco-physiological processes and their structural interactions (real-structure models). Two such models are described, and representative results presented. The generic real-structure model TREEDYN describes stand growth, and its carbon and nitrogen dynamics in interaction with corresponding soil processes, using ten nonlinear differential equations. The state variables are: leaf, fine-root, wood, fruit, and assimilate biomass; carbon and nitrogen in litter and soil organic matter; and plant-available nitrogen. The model has been applied to both a broadleaf tree (Acacia auriculiformis) under tropical conditions, and to a coniferous tree (Picea abies) under European conditions. The real-structure model SPRUCOM simulates Norway-spruce stand growth, including the light-competition-induced self-thinning process, using 30 non-linear differential equations for each of four tree competition classes. The state variables in this case are: needles (7 classes); corewood; sapwood (18 rings); branches; coarse roots; fine roots; and assimilate. The model has been used, in particular, for the study of stand production as a function of air pollution affecting leaf efficiency and/or fine-root turnover. Both models employ a full description of light attenuation and photosynthetic production in the canopy. The models are parameterized by species- and region-specific real-system parameters; no time-series data are required for this purpose. For more efficient computation in management tools, a condensation of complex models to simpler models may be required, while maintaining structural, behavioral, empirical, and application validity. Work on this aspect is briefly reported.