Collective action in an RNA virus

A recent empirical study by Turner and Chao on the evolution of competitive interactions among phage virus strains revealed that a strain grown at high rates of co-infection evolved towards lowered fitness relative to an ancestral strain. The authors went on to show that the fitness pay-off matrix between the evolved and ancestral strain conforms to the prisoners' dilemma. In this paper, I use Turner and Chao's data to parameterize a simple model of parasite collective action. The prisoners' dilemma is based on pairwise interactions of a discrete cooperate/defect nature. In contrast, the collective action model explicitly deals with individual-group interactions where the extent of cooperation is a continuous variable. I argue here that the 'collective action' modelling approach is more appropriate than the prisoners' dilemma for the biology of virus evolution, and hence better able to form a predictive framework for further work on related strains of virus, linking mixing ecology, cooperative phenotype and fitness. Furthermore, the collective action model is used to motivate discussion on the evolutionary ecology of viruses, with a focus on the 'levels of selection' debate and the evolution of virulence.