Stigmergy for Biological Spatial Modeling

Abstract

Complex adaptive systems can be characterized as systems that are comprised of groups of agents following simple rules that, collectively, produce emergent, complex behavior. The key to these emergent properties are the interactions—the exchanges of information—between the agents. Many biological systems can be studied using a complex adaptive systems approach, such as colonies of bees or ants. In some of these biological systems, the communication between individual agents is indirect. This type of communication is termed stimergy: a relatively small amount of information being shared through the environment, rather than directly from agent to agent. This information is nonetheless crucial to the self-organizing properties of the system, and is used by the agents to inform decision making, such as when ants follow a trail of pheromones left by other ants. In this chapter we describe computer simulations of two such systems, created and used to conduct experiments on various types of stimergy: collaboration within a predator-prey system, and angiogenesis in cancer growth. The first utilizes a cellular automata model, and the second a multiscale agent-based model. Further, this paper defines various options of communications for these simulations, and examines the use of similar communication paradigms in these two different types of models. Results support that stigmergy can be adapted to a variety of situations. Also, that awareness of the speed of algorithmic decisions can increase its usefulness in biological modeling. These ideas can be adapted to many other modeling situations other than the classic examples of self-organization like bees or ants.