Abstract: There is a wide range of multicellularity that spans from simple cell clusters to plants and animals. Recent experiments show that simple forms of multicellularity can easily evolve de novo. However, the evolutionary trajectory towards increased complexity such as cell differentiation is not clear. Moreover, since these early forms of multicellularity are just a few mutations away from a unicellular ancestor, evolution might cause the groups to collapse back to unicellularity. Here we study the evolutionary trajectory from a unicellular ancestor to differentiated multicellularity in the context of an environment that periodically contains an abiotic stress. In environments where the stress is present cells are prevented from growth and die, but in environments without the stress cell populations grow exponentially. In response to this stress, cell populations may improve their fitness by either evolving multicellularity or differentiating into two phenotypes specialized to the two environmental states. Importantly both of these responses have associated costs via a time delay in switching between phenotypes or in forming multicellular groups. We use a combination of dynamical models and simulations to study the evolutionary trajectory of unicellular populations adapting to abiotic stress. We find that these trajectories are heavily influenced by historical contingency and often included iterations where complexity is repeatedly gained and lost. Thus we find that the initial selective driver for the evolution of multicellularity may not be enough to sustain it towards ever increasing complexity.
