Abstract

Most models describing the evolution of animal cooperative behaviour are based on the principle of reciprocal altruism. However, according to the "advertising hypothesis", based on Zahavi's handicap principle, an altruistic act is a signal indicating an individual's quality and is profitable even if its recipient never reciprocates. In this paper, this hypothesis has been tested in three computer simulations of the evolution of altruistic behaviour. A model of a population has been studied, where individuals' fitness was calculated as a product of two parameters: "energy" and "prestige". It was assumed that as a result of an altruistic act its donor loses energy and gains prestige, and a recipient gains energy and loses prestige. Individuals played a game based on decision-making: to cooperate (i.e. to behave altruistically) or not in a particular round. An individual's strategy influenced its fitness, i.e. the number of its offspring. As offspring inherited their parent's strategy, frequency of "good" strategies increased in the population, and "bad" ones disappeared. The results of these simulations show that if altruistic acts are costly and bring social prestige, cooperation can persist in the population under a wide range of conditions, despite the lack of direct reciprocity. However, besides the strategy of cooperation with a frequency proportional to an individual's quality, as predicted by the advertising hypothesis, other strategies, even unconditional altruism, can persist in the population in some conditions.