Wolinsky, E.,Libby, E.
Multicellular organisms coordinate growth and differentiation from single cell starting points with developmental programs. While the evolutionary origins of these programs are unknown, it is likely that they are closely tied to the evolution of regulated-not stochastic-phenotypic expression. To determine how such regulation might arise, we consider experimental populations of Pseudomonas fluorescens which evolved stochastic switching in the lab. This switching is directly coupled with environmental oscillations generated by the bacteria themselves. This unique example of niche construction provides reliable information that organisms may incorporate into regulation of phenotypes. We use mathematical models to investigate the success of two forms of regulation that rely on sensing either external or internal information. We find that both strategies can outcompete stochastic strategies for certain combinations of parameters. In particular, external sensing mechanisms are very effective over a large range of parameter space-including parameters that correspond to poor sensing of the extracellular signal and gradual responses. We show with evolutionary simulations that this robustness makes them more likely to evolve from initially stochastically switching populations rather than internal sensing mechanisms which require more tuning of parameters. These results demonstrate that, within this oscillating system, if regulatory mechanisms can evolve to incorporate environmental information then their selective advantage is sufficient for them to fix in the population.