Models of Forest Dynamics

It is difficult to study the tremendous role that spatial relationships have on community structure. The tallest canopy individuals have their own leaves in the full sun and yet rob smaller individuals of light. As a result, recruiting seedlings often experience low and unpredictable light. Large gaps in the canopy caused by a disturbance result in increased light on the forest floor. At high latitudes, this region of increased light can be significantly offset from the location of the canopy opening (Canham 1988; Canham et al. 1990; Baldocchi and Collineau 1994). Thus, it seems obvious that the local spatial structure of trees will be necessary in order to model the dynamics of the forest.

Botkin et al. 1972	Shugart 1984, Urban et al. 1991	Pacala et al. 1993, Canham et al. 1994
JABOWA uses discrete cells that do not communicate at all. Dispersal is from an external list of potential colonists.	FORET uses discrete cells with some communication between neighbors. Dispersal is internal, but often spatially unconstrained.	SORTIE is both individually based and spatially explicit. Both competition and dispersal are estimated continuous functions.

Early simulation models (notably JABOWA) acknowledged the potential role that horizontal heterogeneity has on community dynamics. The description of a forest as a mosaic of closed canopies and gaps captures the idea that variance in light is critical to the success of some species (Botkin et al. 1972; Shugart 1984; Horn et al. 1989; Urban et al. 1991). Gap-phase species need holes in the canopy, but it is assumed that one can ignore the explicit spatial structure of these holes. The JABOWA model assumes this "pure variance" ideal of forest dynamics in its complete dismissal of horizontal spatial structure (Botkin et al. 1972). In addition, recruits are drawn from a predetermined, external list of species. Together, these assumptions preclude competitive or dispersal linkage among cells.

The FORET class of models are logical descendants of JABOWA. These models still assume that the forest can be divided into small (typically the size of a large canopy tree) cells. The most recent versions of the FORET class of models [for example, ZELIG (Urban et al. 1991)] can track the spatial structure among cells and thus can include the effects of local competition. Many of the FORET models assume that new seedlings are produced by the adult trees on the landscape, which is an important improvement over JABOWA (Shugart 1984; Shugart and Prentice 1992; Shugart and Smith 1992a). However, these seedlings are often dispersed globally, precluding the possibility that localized dispersal influences forest development (Urban et al. 1991; Pacala and Hurtt 1993).

Spatially explicit forest simulations can incorporate functions for local competition and dispersal, obviating the need to constrain competitive interactions to cells. SORTIE is an empirically based model of forest dynamics derived from data collected in the Great Mountain Forest in northwest Connecticut (Pacala et al. 1993; Canham et al. 1994). In SORTIE, forest dynamics emerge as the result of local competition for light among the constituent trees. The responses of the trees to their local light and the local dispersal of seedlings are estimated from field data (Pacala et al. 1993; Pacala et al. 1996).

Forest models have become increasingly detailed in their description of local interactions among trees. The SORTIE model tracks the exact position of each tree and determines tree performance based on the individual's local neighborhood. The natural emergence of community dynamics resulting from the empirically estimated, biologically realistic interactions among local trees makes SORTIE an ideal model system to investigate which interactions control forest dynamics.