'Individual rules for social animals: integrating state with decisions about interactions' - by Dr Sean Rands

5 Sep 2012 - 14:00
Date: Tuesday 11 September 2012
Time: 13h00
Venue: Niven Library
Speaker: Sean Rands, University of Bristol, UK (

Synopsis: Organisms have to make decisions about their actions throughout their lives, and these decisions can take many forms. Some can be choices related to key events in their life histories, whilst others are choices about what behaviours to perform from moment to moment. The optimal choice for an organism to perform will depend upon both its internal state (such as its physiology), as well as the state of the environment. Within behavioural ecology, well-characterised techniques exist for identifying consecutive optimal decisions, such as dynamic programming. The decisions made by many organisms are also dependent upon the state and actions of other individuals, which are themselves attempting to optimise their behaviour. Here, I will explore a series of dynamic games that consider the social behaviour of pairs of animals, looking at cases where interacting individuals have a common interest and similar rewards (such as social foragers). If time permits, I will also briefly consider what happens when interacting individuals have differing and conflictatory interests, looking at the uneasy alliance between a parasite and its host. These dynamic games give an optimal rule-set for each individual to follow given that it knows its own state and that of its colleagues. I will also consider what happens if this rule-set breaks down, and ask whether simpler rules-of-thumb can be used to approximate optimal behaviour.