Built using the Modular Agent-Based Evolution (MABE) library.
Experiments reported in this paper used vTODO of this software and commit 67b9575de1af89742ffdb2a02a52cc58448d786b of MABE.
data, tutorials, and writeup @ https://osf.io/ydtvs/
Stigmergy is an indirect mechanism of coordination where an action alters the environment and the changed environment, in turn, alters the likelihood of future actions. Pheromone-marker-based stigmergy is the most common form of stigmergy used in social insect groups. The ant species Monomorium pharaonis produces pheromones that decay after several minutes, whereas Atta columbica ants produce pheromones that persist for several years. The respective transience and permanence of these ants' foraging paths is thought to be related to their pheromones' volatility. Several mathematical and computational models have demonstrated that relatively moderate pheromone decay rates are correlated with increased foraging success in stigmergic agents. However, the implications of pheromone volatility on the evolution of navigation behavior are not well-studied. Here we use digital agent-based evolution to show that moderate pheromone decay rates correspond to an increase in the extent to which agent foraging behavior relies on stigmergic cues. This result suggests that the properties of an organism's pheromones play a vital role in the organism's evolutionary trajectory, driving some towards, and others away from, stigmergic foraging strategies. These results help shed light on how elements of an organism's environment, such as the stability of its food source, relate to the properties of that organism's pheromones. Understanding the evolution of stigmergic group coordination strategies, an instance of collective cognition, provides a foundation for addressing the larger questions surrounding the evolution of intelligence.