Abstract

Evolutionary biologists are increasingly interested in the dynamics of multilevel selection, or selection acting simultaneously at more than one level in a hierarchy of reproducing entities (e.g., gene, chromosome, organelle, cell, organism, group, community, ecosystem). Systems of linear equations are the usual tool for studying evolution, but may be poorly suited to capturing important dynamics of multilevel selection. Here we use an agent-based model to study the evolution of cooperation in spatially structured populations. This work addresses the long-standing controversy over the role of “group selection”, or natural selection between versus within groups of interacting individuals. In an ecologically plausible setting, cooperative individuals with lower rates of food consumption compete reproductively against selfish individuals with higher rates of consumption. The results show that changing the spatial distribution of food, and thus the distribution of the individuals seeking it, can determine whether or not cooperation evolves. In this model cooperation evolved under a fairly wide range of parameter values, even without the kinship effects and discrete mixing phase that are sometimes thought to be necessary. We suggest that integrating equation-based analysis tools into agent-based models is a powerful way to study selection in systems with complex dynamics.