Some Ecological Consequences of a Computer Model of Forest Growth

The complexity of a forest ecosystem makes difficult any attempt to synthesize knowledge about forest dynamics or to perceive the implications of information and assumptions regarding forest growth. Although digital computer simulation seems to offer a potential for creating a complete model of forest growth, little progress has been reported. Computer simulation has been carried out for the growth of trees in even-aged stands of a single species (Mitchell 1969), and for meteorological energy exchange in a forest canopy (Waggoner but apparently no one has successfully reproduced the major characteristics of a mixed-species, mixed-aged forest from a conceptual basis. A computer simulation of forest growth is now developed that successfully reproduces the population dynamics of the trees in a mixed-species forest of north-east North America. The simulator is designed to be used in the Hubbard Brook Ecosystem Study and to provide output in the same form as the original vegetation survey of that study (Bormann et al. 1970). However, the underlying concepts of the simulation are general. The properties of each species are derived from its entire geographic range and in theory any non-hydrophytic species whose relevant characteristics are known can be entered into the simulation. In the present version of the program, the description of the environment is restricted to those features that have been recorded for the Hubbard Brook Forest, but the relative importance attached to each environmental factor has been influenced by the environmental characteristics of the north-eastern United States. It is hoped that a wide dissemination of this simulator will encourage others to test this version with their data and hence lead to later versions of wider usefulness and applicability. The basic goal was to produce a dynamic model of forest growth, a model in which changes in the state of the forest are a function of the present state and random components. This approach has two advantages over the curve-fitting approach to forest growth: first, the simulator can be regarded as a repository for an integrated knowledge of the ecosystem; second, additional hypotheses can be formulated and tested using Monte Carlo samples of simulator runs and comparing the results with observed data. For