By elevating the landscape to an equal player in agent-based simulations, a new class of scenarios become open for exploration via computer modeling. It has long been known that populations affect their environments, and that changes in the environment affect how the population acts. For reasons mentioned in section 1, it has not been practical until recently to explore these interactions via computer simulation.
Reusable software packages such as Swarm have dramatically increased the availability of complex system simulation exploration to researchers who are not experienced computer programmers. The ideas of dynamic populations and dynamic environments is certainly not new, and ingredients to create such simulations have been around for a while. Computer technology has been cheap and available for more than a generation of researchers. Newer theories about complexity and complex systems such as chaos theory, fractal geometry, and self-organized criticality are not quite new any more. Agent-based simulations and cellular automata have been well-used tools for simulating dynamic populations and landscapes. Their lack of application to agent/environment interaction has certainly not been for lack of tools or lack of talent on the researchers.
Though all of these ingredients have been around for some time, combining them into an integrated simulation has been difficult. The availability of software libraries such as Swarm may well be the catalyst that allows for creative recombination of research ideas into new packages, where the strengths of each of the components creates a picture that is greater than the sum of its constituent parts, and distinct from any of it's parent disciplines. The emergence of a new class of problems may well mirror the emergence of new disciplines similar to what occurred with the coupling of relational databases and computer cartography to create geographic information systems, or the combining of statistics and economics to create econometrics.
It is anticipated that this modeling tool will allow access to a number of questions that wildland ecologists will be able to explore via simulations. While it's not possible to name or even to anticipate the number of applications, some areas that might be addressed are:
While all these questions have been partially answered by existing simulation technologies, there are other questions that are difficult to address without a suitably dynamic representation of agent/environmnet interaction.