Individual-based ecological models: Adjunctive tools or experimental systems?

• Individual-based computer modeling (IBCM) can be either theoretical or applied.
• Proposed methods of model evaluation tend to emphasize empirical calibration.
• Theoretical IBCM tends to forgo empirical calibration owing to generality of aims.
• Evaluating theoretical IBCM that focuses on predictive validation is considered.

The role that individual-based computer modeling (IBCM) should play in the field of integrative ecology needs to be clarified in light of the continuing concern with the empirical validation of individual-based computer models. Though perfectly legitimate and understandable, the requirement of empirical validation has been taken to extremes. The result of doing so is that potentially useful scientific investigations involving computer simulations risk being thwarted on the grounds that they are not empirically calibrated or perhaps not historically validated. We shall argue that the role that IBCM plays in integrative ecology depends on whether one is doing applied ecology with concerns such as species conservation or whether one is doing theoretical ecology. In the former case, computer modeling should incorporate real-world elements and actual experimental data if the goal is to model existing ecosystems and to make long-term predictions about these systems. In this case, IBCM functions more like an investigative tool for scientific inquiry. On the other hand, if one's concerns are more theoretical, then IBCM has value in its own right in terms of high generality and equally high predictive power. Although the modeling should be realistic in a broad sense that is consistent with species generally that evolve in a world with predation, pathogens and fluctuating resources, simulations for more theoretical investigations need not incorporate experimental data – especially in light of the fact that these are not always obtainable in the field. They are experimental systems in their own right and not merely adjunctive tools.