Individual-level interaction rules for the coordination of collective motion

5 July 2019, 12.00 PM - 5 July 2019, 12.45 PM

Andrea Perna

University of Roehampton, UK

The movement of an animal across the environment is usually modulated by interactions with other animals. A flock of birds, a school of fish, a predator tracking a prey or an ant following a pheromone trail are just a few among the many possible examples of such collective interaction phenomena. If we want to understand how animals move within a group, we need to simultaneously understand how they interact with each other, and how the group as a whole will move. Methods and models inspired by statistical physics have played a major role in shaping this research field of understanding collective motion, as they allowed characterising complex group-level patterns and predicting the broad conditions under which different patterns can emerge. However, biologists are often interested in the evolution and the adaptive value of behaviour, which requires addressing the study of collective motion with a focus on individual animals and individual differences of behaviour: which individuals in a group play a major role in determining collective decisions about the direction to take or the activity to perform? Which specific signals and cues are used for communication by animals of a given species? In my talk I will illustrate with some examples two complementary approaches that can be used to characterise group interactions at the individual-level: (1) a "local" approach, in which we study the movement of individuals directly from tracking data, and we make predictions of how they move based of the local configuration of the environment and other individuals around them and (2) a "global" approach, in which we identify regularities in the group configuration or dynamics and we use these regularities to propose possible rules of interaction between individuals. I will show that the two approaches can provide complementary information about how animals move and interact.


Back to Statistical Physics Meets Movement Ecology

Edit this page