Multi-agent systems: the ecology of swarming societies

Iain Couzin

Department of Ecology and Evolutionary Biology, Princeton University, USA
& Department of Zoology, University of Oxford, UK

In recent years there has been an expanding interest in pattern formation in biological systems. The study of pattern formation covers a wide range of areas, including attempting to explain foetal development, patterns on the coats of mammals, and the collective swarms of bacteria, army ants and locusts. In particular there is a growing interest in the relationship between individual and population-level properties. One of the fundamental problems is related to spatial scale; how do interactions over a local range result in population properties at larger, averaged, scales, and how can we integrate the properties of aggregates over these scales? Many group-living animals exhibit complex, and coordinated, spatiotemporal patterns which, despite their ubiquity and ecological importance, are very poorly understood. This is largely due to the difficulties associated with quantifying the motion of, and interactions among, many animals simultaneously. It is on how these interactions scale to collective behaviours that I will focus here. Using a combined empirical approach (using novel computer vision techniques) and multi-agent computer models, I investigate pattern formation in both invertebrate and vertebrate systems. This includes self-organized lane formation and the organization of traffic flow in army ants, as well as collective memory, self-organized group structure, information transfer and leadership in animal groups.