Program for Ecosystem Research
Research Project Multi-agent simulation of hierarchical organization in ecological systems: a computational approach to a theory of scaling and system response to environmental perturbation
Principal investigator: Anthony W. King
To investigate, using multi-agent system simulation (MAS), whether (1) hierarchical organization of ecological systems is an emergent property, and higher-level organization and function emerge from interactions among lower level agents; and (2) hierarchical organization buffers the higher-level system from lower-level perturbations, and only those lower-level perturbations that alter system organization are likely to cause observable change at the whole-system (or ecosystem) level.
Ecosystem being studied
Theoretical, generic ecological systems are being simulated. Understanding of system organization common to all ecosystems is important to understanding specific ecosystems of economic or other value.
Simulations with an MAS-system are being completed.
Why this is important
Reliable models of ecosystem response to anticipated environmental change from increasing atmospheric greenhouse gas concentrations are essential tools for determining safe levels of those gases. A rigorous theory of ecological organization is an important contribution to the development of those models.
Multi-agent system simulation will be used to model ecological organization and the consequences of that organization for the flow of information through those systems. MAS is a formalized, object-oriented approach to modeling complex systems in which a collection of autonomous but interacting entities (agents) communicate with each other and interact with an environment. Higher-level, system-scale organization and function emerge as a consequence of local agent-scale behavior and interaction. In the proposed application of MAS, no a priori assumptions are made about the organization of the system into levels. Rather, the MAS is used in a computational test of fundamental hypotheses from hierarchy theory about the organization of ecological systems. The results of these tests will help determine the value of hierarchy theory as a theoretical basis for understanding the scaling and translation of information in ecological systems.
Anthony W. King, Oak Ridge National Laboratory
Funding period: February 2005 to present