I investigate collective behaviour using a wide range of theoretical and experimental approaches. Individual-based (Lagrangian) computer modelling is used extensively to reveal how individual movement and interactions result in group characteristics. This technique is used to gain insight into the structured patterns of movement within human crowds and the development of trail networks by ants. These models reveal the importance of interactions among individuals to density-dependent group behaviour. A simulation of animal groups in three-dimensional space reveals the existence of several robust collective patterns. Simulated groups show similar group-level behaviour and internal structure to natural groups. The model also reveals how differences among individuals influence group structure, and how individuals employing simple, local rules of thumb, can accurately change their relative position within a group (for example, to move to the centre, or to the periphery) without necessitating information regarding their current position within the group.
New techniques in computer vision are introduced that can facilitate the automatic analysis of collective motion. This software can simultaneously track and analyse the movement of a large number (hundreds) of organisms. Computer vision is used to reveal the spatio-temporal patterns of activity in ant colonies for the first time. I also show how it can record detailed aspects of individual behaviour, including the movement of, and production of honeydew and offspring by, aphids. This technique is used in a detailed analysis of ant exploratory behaviour, revealing temporal and spatial information about the movement patterns of individual ants, and the relationship between individual behaviour and collective exploration. Simultaneous digital tracking of organisms is a powerful technique that in the future is likely to provide insight into the behaviour of many animal groups.
|Date of Award||3 Dec 1999|