University of York Interdisciplinary TRANSIT Seminar Series (funded by EPSRC)

Given by David Sumpter, University of Uppsala

The movement of groups of fish, birds and insects are classic examples of decision-making by animal groups. Here I present some recent experimental and modelling results relating to the movement decisions made by these groups.

I will first give a couple of simple arguments about why wide consultation leads to better decisions, even between a group of poorly informed individuals with extreme views. Empirical studies do not indicate that individuals widely consult others in making decisions. Instead, animals tend to copy the choices of a small number of neighbouring individuals. This leads me to describe the process of decision-making in terms of positive feedback. I show that although positive feedback can lead to sub-optimal decisions, provided copying is made according to the 'correct' functional response to others, it does on average lead to better decision-making. Using this approach I show that even pigeons and fish can make accurate 'consensus' decisions, without explicit consultation between each other.

David Sumpter's research aims to understand the behaviour of animal groups. In particular, he uses mathematical models to predict how interactions between individuals produce patterns at the level of the collective. Study organisms include ants, fish, birds, locusts and humans. He currently works at the mathematics department in Uppsala Univeristy, Sweden, and has previously worked in Oxford, Umeå and Manchester (where he did his PhD).