Search

supplementary figures and tables for Leaders of War: Modelling the evolution of conflict among heterogeneous groups

Social behaviours cannot be fully understood without considering the network structures that underlie them. Developments in network theory provide us with relevant modelling tools. The topology of social networks may be due to selection for information transmission. To investigate this, we generated network topologies with varying proportions of random connections and degrees of preferential attachment. We simulated two social tasks on these networks: a spreading innovation model and a simple market. Results indicated that non-zero levels of random connections and low levels of preferential attachment led to more efficient information transmission. Theoretical and practical implications are discussed.

Social learning can be adaptive, but little is known about the underlying mechanisms. Many researchers have focused on imitation but this may have led to simpler mechanisms being underestimated. We demonstrate in simulation that imitative learning is not always the best strategy for a group-living animal, and that the effectiveness of any such strategy will depend on details of the environment and the animal's lifestyle. We show that observations of behavioural convergence or "traditions" might suggest effective social learning, but are meaningless considered alone.

Mimicry is a poorly understood phenomenon; we present a simulation of the evolution of both Batesian (parasitic) and Muellerian (mutualistic) mimicry. In the model, multiple species are preyed upon by a single abstract predator; the appearance of each prey species can evolve but their palatability is fixed. Batesian mimicry evolves regardless of the initial phenotypic distance between the two species, whereas Muellerian mimicry requires an initial resemblance, except when a third (Batesian) species is present.