Today at lab meeting we discussed the recent Nature paper “Experimental evolution of bet hedging” by Beaumont and colleagues. Nature has an short summary of it, and my colleague Will Ratcliffe at the University of Minnesota has blogged about it.

It’s a cool result. The authors passaged bacteria through a set of alternating environments and at each step selected out the first kind that produced a different colony type than the rest of the population. In effect they were selecting for bacteria that looked different. In a couple of their evolving populations, bacteria evolved that could reversibly switch between two colony types, a trait that contributes to virulence in some bacterial pathogens. They went on to characterize what mutations occured in the evolving population, determine which one caused the switching phenotype, and show that it only increased fitness if some of the earlier mutations were already present in the genome.

So that’s cool and all, but after reading the paper I still found myself wondering what we know about how evolution works that we didn’t know already. The authors basically did an evolution experiment with the expectation that this phenotype might evolve, and then found that it sometimes did. It’s more like a proof of principle. But it doesn’t really tell us much about when we should expect organisms to adapt to environmental variation through these kinds of bet hedging strategies instead of directly sensing the environment or just evolving rapidly to each one in turn.

I guess what I’m reacting to is that this paper is about discovery and not about testing hypotheses. It’s motivated by theory, I guess, but only in a general sense. It’s not trying to distinguish between alternative explanations for some evolutionary phenomenon. It’s more saying: this can happen. Don’t get me wrong—I agree that discovery is a very important part of science. It’s just not something that tends to figure much into my own thinking about the practice of science. Maybe I should do something about that.