Search

A critical need in the management of public and non-urban private lands in the United States and elsewhere is assessing the efficacy of different management alternatives under different environmental and land use scenarios. One type of tool that would be extremely valuable is large-scale ecological simulation models designed to project effects of alternative climatic, usage, and management scenarios on ecological resources. A modeling challenge in this type of application is to link mechanistic simulations of small-scale ‘ecosystem’ processes to large-scale ‘landscape’ processes to provide more realistic and exhaustive projections of effects and ramifications of management alternatives. The Ecological DYnamics Simulation model (EDYS) is a general ecosystem simulation model that mechanistically implements relevant processes in ecosystem dynamics, including: climatic inputs, soil water and nutrient dynamics, plant uptake and growth by species, herbivory, fire, contaminants, physical disturbance, and management activities. In the EDYS model, ecological processes simulated in plot-level ecosystem cells are scaled up to the landscape level using a grid-based representation of the spatial extent of that ecosystem across the landscape. A significant practical challenge in applying complex ecological models is compilation of appropriate input data from a wide-variety of print and on-line media. A semi-automated database is currently under development, which will compile, organize, and format data sets to facilitate future EDYS applications. Another challenge is linking different types of models, each of which is specialized to simulate particular aspects of ecosystem and landscape dynamics. As so many different types of organizations are presently involved in model development and resource management, ownership of models and datasets will increasingly become an issue in their distribution and use among different types of land managers.

Grid-based models such as cellular automata are useful tools for investigating spatial dynamics in ecological systems. Quantitative measures have been proposed for characterizing static spatial patterns, but techniques for examining the dynamics of spatial pattern are less developed. I present an adaptation of Hamiltonian free energy equations to predict numbers of state changes resulting from different ecological mechanisms in spatial automata models. The three terms in the Hamiltonian estimate the number of grid cells changing state due to spontaneous state changes, to interactions among cells, and to heterogeneity in the behavior of different cells in the grid. PComp is a simple competition model in which each cell in the grid can be occupied by a single individual plant. The simulation calculates the Hamiltonian at each time step to indicate how many of the observed changes in cell occupancy are due to basic colonization and death rates, crowding effects, dispersal from adjacent cell, and differences in substrate suitability.