Forest models defined by field measurements: I. The design of a northeastern forest simulator

We introduce a new spatially explicit model of forest dynamics. The model is constructed from submodels that predict an individual tree's growth, survival, dispersal, and recruitment, and submodels that predict the local availability of resources. Competition is entirely mechanistic; plants interfere with one another only by depleting resources. We also describe maximum likelihood methods for estimating each of the submodels from data collected in the field. Over the past two years, we collected the necessary data for the dominant tree species in the Great Mountain Forest (Norfolk, Conn.). We report estimates of submodels for each species, and show that the calibrated population dynamic model predicts the structure and dynamics of natural forests. Finally, we contrast our model with the JABOWA–FORET family of forest models.