Spatial modelling of vegetation change in dynamic landscapes: a review of methods and applications

Because of the spatiotemporal scales involved and the logistical constraints in collecting landscape-level data, spatially explicit simulation models have become important tools in ecological and biogeographical studies conducted over broad extents. Here we review the methods used and some of the applications of landscape-level models of succession and disturbance dynamics. Mechanistic and stochastic models are compared and contrasted and the development, over the last 15 years, of spatial landscape models of ecological change is discussed. Coarse-grained spatial landscape models are compared with finer-grained individual-based approaches (eg, forest gap models). Management and monitoring applications of landscape models are considered alongside a discussion of the appropriate use of models in this context. A key area where spatial landscape models of the type described here need to develop is improved integration with the social sciences - both in terms of the parameters and the processes that the models incorporate. Finally issues related to scale and scaling are outlined and, in particular, the utility of methods for linking ecological models operating at disparate scales ( eg, forest gap models versus landscape models) is examined.