Integrating Spatial Data into an Agent-Based Modeling System: Ideas and Lessons from the Development of the Across-Trophic-Level System Simulation

Abstract

Agent-based or individual-based models allow for variation in the state and behavior of the basic objects that interact within the model. Modeling each individual as a separate entity allows for spatially explicit components to be included so that the individuals can interact with a heterogeneous landscape and with each other. Realism is added to the models by incorporating spatially explicit data for the area of interest. Integrating spatial data into an agent-based system requires that a significant level of geographic information systems (GIS) functionality from traditional GIS be incorporated into the modeling system. This approach may seem redundant and costly, but current GIS systems do not offer a framework for building dynamic agent-based models. The across-trophic-level system simulation (ATLSS) is characterized by the integration of several distinct agent-based models and spatially explicit data into a single modeling system. One of the goals of the ATLSS (pronounced like “atlas”) project is to investigate the relative response of various interconnected trophic levels of the South Florida (SF) Everglades to different hydrologic scenarios over a thirty-year planning horizon. The ATLSS approach consists of several distinct component models, each of which represents different biotic components of the Everglades system, linked together as a multimodel. Currently, ATLSS includes component models for the Florida Panther, the Cape Sable Sea Side Sparrow, white tail deer, fresh water fish, wading birds, the Snail kite, and vegetation biomass. Additional models for alligators and various reptiles and amphibians are in development and will be added as they are completed. The list above by no means enumerates all of the species in SF. It reflects instead the initial attempt to include key components that biologists with many years of field experience in SF believed were critical to include. It also reflects the time and funding limitations to carry out empirical studies and develop both the theory and software needed to model each system component. The Florida Panther, the Cape Sable Seaside Sparrow, the Snail Kite and several wading bird species of SF are all listed as endangered species. The fish, deer, and vegetation on the other hand are included since they are critical resources for the endangered species.