Goals and Limitations of Modeling Collective Behavior in Biological Systems

Local social interactions among individuals in animal groups generate collective behavior, allowing to adjust changing conditions. Historically, scientists from different disciplines have taken approaches modeling behavior. We describe how each can contribute the goal of understanding natural systems. Simple bottom-up models that and their directly demonstrated local far equilibrium states. However, such simple are not likely accurately actual mechanisms play any real biological system. Other classes top-down group-level behavior been proposed for where function is understood. Such cannot necessarily explain why or functions emerge first principles. Because strengths weaknesses no single approach will always be best, we argue aimed at systems should formulated address specific questions allow validation. As examples, discuss four forms differ both produce ecological context, thus require very frameworks. 1) Harvester ants use consisting brief antennal contact, which one ant assesses cuticular hydrocarbon profile another, regulate foraging activity, modeled as a closed-loop excitable 2) Arboreal turtle form trail networks canopy tropical forest, using pheromone; detects volatile chemical another has recently deposited. The process maintains repairs trail, distributed algorithm, constrained by physical configuration network vegetation they travel. 3) Swarms midges interact acoustically non-locally, well described agents moving an emergent potential representative swarm whole rather than individuals. 4) Flocks jackdaws change effective depending on topological distance when traveling but metric mobbing. research about these led approaches.