First published 2 July 2012
Robustness and evolvability of cooperation
Antoine Frénoy, François Taddei, Dusan Misevic
Robustness and evolvability are indirectly selected properties of biological systems that still play a significant role in determining evolutionary trajectories. Understanding such second order evolution is even more challenging when considering traits related to cooperation, as the evolution of cooperation itself is governed by indirect selection. To examine the robustness and evolvability of cooperation, we used an agent-based model of digital evolution, Aevol. In Aevol individuals capable of cooperating via costly public good secretion evolve for thousands of generations in a classical tragedy of the commons scenario. We varied the cost of secreting the public good molecule between and within individual experiments and constructed and evaluated millions of mutants to quantify the organisms' position in the fitness landscape. Populations initially evolved at different regimes selecting against secretion, and then continued the evolution at a reasonably low cost of secretion. The populations that experienced a very strong selection against cooperation evolved less secretion than the ones that initially experienced a less drastic selection against cooperation via a high secretion cost. The mutational analysis revealed a correlation between the number of mutants with increased secretion and the secretion level across all costs of secretion. We also evolved several clones of each population to highlight a strong effect of history in general on cooperation. Our work shows that the history of cooperative interactions has an effect on evolutionary dynamics, a result likely to be relevant in any cooperative systems that are frequently experiencing changes in cost and benefit of cooperation.