A quantitative decision theory of animal conflict

Abstract

Interactions between individuals are thought to shape evolution and speciation through natural selection, but little is known about how an individual (or player) strategically interacts with others to maximize its payoff. We develop a simple decision-theoretic model that generates four hypotheses about the choice of an optimal behavioral strategy by a player in response to the strategies of other players. The golden threshold hypothesis suggests that 62% is the critical threshold determining the transition of a larger player's strategy in reaction to its smaller dove-like partner. Below this critical point, the larger one exploits the smaller one, whereas above it, the larger one chooses to cooperate with the smaller one. The competition-to-cooperation shift hypothesis states that a larger player never cooperates with a smaller hawk-like player unless the former is reversely surpassed in size by the latter by 75%. The Fibonacci retracement mark hypothesis proposes that, faced with a larger dove-like player, a smaller player chooses to either cooperate or cheat, depending on whether its size relative to the larger player is less or more than 38%. The surrender-resistance hypothesis suggests that, in reaction to a larger hawk-like player, a smaller player can either gain some benefit from resistance or is sacrificed by choosing to surrender. We test these hypotheses by re-analyzing body mass data of full-sib fishes that were co-cultured in a common water pool. Pairwise analysis of these co-existing fishes broadly suggests the prediction of our hypotheses. Taken together, our model unveils detectable yet previously unknown quantitative mechanisms that mediate the strategic choice of animal behavior in populations or communities. Given the ubiquitous nature of biological interactions occurring at different levels of organizations and the paucity of quantitative approaches to understand them, results by our decision-theoretic model represent an initial step towards the deeper understanding of how biological entities interact with each other to drive their evolution.