Despotism, Boldness and Female Dominance: a Model

Charlotte K Hemelrijk

Theoretical Biology, Centre for Ecological and Evolutionary Studies, University of Groningen, Groningen, The Netherlands

Artificial life models have shown that simple rules in agents may lead to unexpected and complex collective patterns. Therefore, these kinds of models help us to develop new hypotheses about the connection between different traits. They lead to hypotheses that are simple in terms of their assumptions about cognitive sophistication and complexity of behavioural rules. I will illustrate this with examples that are particularly relevant to the types of societies of primates, and types of personalities in birds and fish. We use a model (called DomWorld) in which agents exclusively group and compete. The effects of victory and defeat are self-reinforcing. Under certain conditions a mutual feedback between dominance and spatial structure arises that has a cascade of effects. These effects resemble phenomena observed among real animals. Increasing the value of a single parameter (intensity of aggression) leads to many emergent phenomena that are also observed in despotic and egalitarian societies of primates: groups loosen, behaviour becomes dominance-orientated, and dominant individuals occupy the centre. A similar effect can be obtained by changing the distribution of food and, thus, the degree of cohesion of a group. By adding the sexes (which differ in aggression intensity and body size) we show what factors increase female dominance over males and, at the same time, male tolerance towards females. Besides, the hierarchical differentiation becomes stronger among females than males despite the male\x{2019}s greater intensity of aggression. All this resembles findings in primates. Personality types are represented in the model by different strategies to attack ('bold' personalities by 'obligate'-attack versus 'cautious' ones by 'risk-sensitive'-attack). It appears that the dominance values of individuals that attack obligatorily are bi-modally distributed and are extremer than those of risk-sensitively attacking individuals. This resembles observations on personality style in great tits. Furthermore, dominance-assorted spatial structure may lead to differences in experiences with predators between dominants in the centre of the shoal and subordinates at its periphery. This may lead to correlations between the frequency of aggression and of predator inspection that have been found among fish in certain lakes. It will be explained how all these phenomena emerge in the model and in what respects explanations from the model are simpler and more general than the conventional ones.