How can individual-based simulation modeling contribute to development of ecological theory? Our collaborator Volker Grimm, in his classic review ("Ten years of individual-based modeling in ecology: what have we learned and what could we learn in the future?", Ecological Modelling 115: 129-148, 1999), pointed out the need to link individual-based modeling to ecological theory. We have identified a variety of ways that individual-based simulation can contribute to theory.
- Developing and testing theory for individual-based ecology. In his 1999 book Fragile Dominion: Complexity and the Commons, Simon Levin states that "The most important challenge for ecologists remains to understand the linkages between what is going on at the level of the physiology and behavior of individual organisms and emergent properties such as the productivity and resiliency of ecosystems." Using IBMs, we can develop and test theories of individual adaptive traits that explain system-level phenomena such as population dynamics. What models of individual behavior cause realistic patterns of emergent system properties in IBMs? We developed a "pattern-oriented" process for testing theories of individual behavior by examining their ability to reproduce a range of observed behavior patterns. This process can allow alternative theories to be compared and conclusions to drawn about how individuals traits determine population-level processes. The pattern-oriented model analysis procedure is described by Railsback (2001, Natural Resource Modeling) and illustrated by Railsback and Harvey (2002, Ecology)—our study of habitat selection theory for trout.
- The “state-based, predictive” approach to theory for adaptive decisions by individuals. How do animals make decisions, such as selecting habitat, that affect both their growth and their mortality risks? We found no satisfactory methods even though modeling such decisions realistically is essential to the success of many IBMs. Many behavior models simply assume that animals make decisions to maximize their growth, whereas we know that (1) avoiding risk is critical and (2) the importance of growth varies drastically over an animal’s life cycle.
We developed a new approach to modeling decision-making that is conceptually simple and computationally compatible with IBMs (Railsback et al. 1999, Ecological Modelling). For the habitat selection decision by trout, we assume fish move to maximize their fitness, where fitness is the probability of surviving (both predation and starvation, a function of food intake) and growing to sexual maturity over a specified number of upcoming days. Limitations of similar, previous approaches (e.g., the Unified Foraging Theory of Mangel, M. and C. W. Clark, 1986, Ecology 67: 1127-1138) are avoided by assuming individuals make a simple prediction of future environmental and internal conditions. There are many other potential applications of this state-based, predictive approach in individual-based modeling.
- An overall conceptual framework for IBMs. The developing field of Complex Adaptive Systems offers many appealing features as a framework for designing IBMs. The following concepts from Complex Adaptive Systems provide a list of issues to consider in designing individual-based model. This topic was explored in a presentation to the 1999 SwarmFest meeting and by Railsback (2001, Ecological Modelling).
- 1. Emergence: what behaviors should emerge from the model’s mechanistic representation of key processes vs. being imposed on the model as empirical relations? How should individual traits be modeled so that realistic population responses emerge?
- 2. Adaptation: given the model’s temporal and spatial scales, what adaptive processes of individuals should be modeled? What mechanisms do individuals use to adapt in response to what environmental forces?
- 3. Fitness and strategy: what measures of fitness are appropriate to use as the basis for modeling decision making? Should fitness measures change with life history state?
- 4. State-based responses: how should decision processes depend on an individual’s state?
- 5. Prediction: anticipating decision outcomes appears essential for modeling many behaviors; what are realistic assumptions about how animals predict the consequences of decisions?
- 6. Computer implementation: what user interfaces are necessary to make the model, and especially individual behaviors, observable and testable? How will the model’s full design and computer implementation be documented and tested so results are reproducible and valid?
- Using IBMs as a testbed for conventional ecological theory. An IBM can provide a rich environment for testing ecological theories that are difficult or impossible to test in natural systems. For example, “habitat selection modeling” is widely used in ecological management, and based on a theory very difficult to test in nature: that the habitat types where animals most commonly occur provide the best fitness. But are habitats where you see the most animals really the best habitats? If not, is habitat selection modeling of any value? We addressed both of these questions in our trout IBM, with surprising results (Railsback et al. 2003, Ecological Applications).