Summary

Animal contests were the focal topic that brought game theory to the attention of behavioural ecologists, giving rise to evolutionary game theory. Game theory has remained by far the most popular method of deriving theoretical predictions ever since, although it nowadays coexists with other methods of analysis. Here I review the developments to date and highlight similarities and differences between models. There is a clear progression from simple two-player models with fixed payoffs to explicit tracking of fitness consequences in a population context. In many cases this development has helped to discover that some of the early predictions may have been misleading. Despite the large number of current models, there are still gaps in the theoretical literature: sometimes simplifying assumptions have been relaxed in one context but not another. I hope that by highlighting these gaps theoreticians will be provided with new research ideas, and empiricists will be encouraged not only to distinguish between existing models but to be able to point out assumptions that are essential for deriving a result yet may be violated in existing systems, thus directing new modelling in the most useful direction.

Introduction

Until the mid 1960s, animal contests were viewed using group selectionist thinking. Julian Huxley (1966) thought that ritualised fights evolved to limit intraspecific damage, partly based on Konrad Lorenz's (1964, 1965) ideas that species need to evolve mechanisms that limit aggression in species that possess dangerous weapons for other reasons, e.g. as adaptations for capturing prey. Following George C. Williams’ (1966) book Adaptation and Natural Selection: A Critique of Some Current Thought, however, biologists became aware of the need to distinguish between explanations that are based on benefits to the individual versus those that rely on benefits accruing to a group (or a species).