The evolution of cooperation and the production of public goods in the context of predator inspection in the Trinidadian guppy

Abstract

As animals grow, survive and reproduce, they often must interact with others. Such interactions can involve conflict, but are also often characterised by mutual benefits. The evolution of such cooperative behaviour is an interesting problem for evolutionary biology because individuals that behave cooperatively are often at risk of being exploited by others. In this thesis, I use a combination of theory and experiments to explore three mechanisms by which individuals might avoid exploitation by others while benefitting from the rewards of cooperation: living in small groups, leading others into cooperative action, and forming partnerships. I investigate these mechanisms in the context of predator inspection in Trinidadian guppies, a natural example of a public goods game in which individuals can cooperate to produce a benefit for the whole group. In Chapter 1, I introduce the concepts and theoretical frameworks that I will use throughout the thesis. In Chapter 2, I describe an experiment in which I tested a prediction from existing theory that individuals in large groups should cooperate less frequently than those in small groups. I found that guppies in large groups – contrary to prediction – inspected more frequently than those in small groups. This could result from individual phenotypic differences amongst individuals, which are not included in current models of predator inspection. In Chapter 3 I therefore developed a game theoretic model allowing individuals to differ in their propensity to lead their groupmates into inspection, and the amount of risk that they take on when they do. This showed that while leading could evolve depending on whether leaders were likely to be followed, there was no stable level of risk that leaders should evolve to adopt. I then experimentally tested whether guppies showed evidence of having leaders, and whether leaders took on more or less risk than their followers during inspection. I found evidence of leadership in the guppies but not during close inspection, and leadership was not related to the amount of risk that guppies took. Animals’ participation in public goods production probably also depends on their relationships with others in their group, and cooperative animals often form partnerships that can promote cooperation. In Chapter 4, I investigated the effect of information about the behaviour of others on the stability of partnerships using a Bayesian learning model. This showed that the more information individuals had about the population, the more they switched partners (thus, the less stable their relationships). Such partnerships are probably important for cooperation if individuals are influenced by the behaviour of their partners. In Chapter 5, I explored these potential consequences of partnerships for public goods production, using an evolutionary simulation model. This model showed that simply being influenced by other individuals can promote the production of public goods, even when there is selection against the behaviour. Finally, in Chapter 6, I discuss these findings in the context of previous literature on the evolution of cooperation, and suggest directions for future research.