We recently showed that, contrary to clear theoretical predictions, shape and motility have evolved independently in a large group of bacteria. So how did we arrive at this conclusion?

After finishing my PhD I had the opportunity to work on a multidisciplinary project called 'form and function in a microbial world', developed by Prof. Stuart Humphries. I found the project exciting, first because it integrates evolutionary patterns and biophysical constraints in order to try to understand microbial shape and its function. Secondly, because it offered me, for the first time, the prospect of challenging my biologist's ideas about shape with the perspectives of a biophysicist (Stuart), a physical ecologist (Oscar) and a mathematical modeller (Rudi).

While the study of functional morphology is commonplace in eukaryotes, the same cannot be said for prokaryotes. Many evolutionary biologists have asked why a giraffe has a long neck, but why should we care about the shape of a bacterium?

Given that for my PhD I worked mostly on protozoa, I never really thought about bacterial shape diversity (as many biologists don't!), except to know that they could be rods, spheres or spirals. I was really surprised when I discovered the myriad of shapes that bacteria can exhibit.

Knowing why they have so many different morphologies is the driving aim of our project. However, in order to understand why different bacteria have different shapes, we needed to know which bacteria had which shape. With no datasets available, Stuart and I saw a clear need to collect morphological and ecological data on bacteria. Somehow, this job fell to me...

I spent several months collecting morphological and ecological data for 325 species from Firmicutes bacteria, and that task continues now with other groups.

Our analysis, in collaboration with Dr. Chris Venditti, took advantage of a new generation of phylogenetic comparative methods. We fully expected to confirm a widely held belief, backed by strong theoretical predictions, that rod-shaped cells tend to be motile, while coccoid cells do not. Our expectation was that shape and motility in the Firmicutes (and probably more widely in bacteria) had co-evolved. To our great surprise we didn't find any association between the two traits. Rebutting a theory with data turns out to be surprisingly difficult, and in order to both satisfy ourselves and others, we used a number of approaches within two statistical frameworks, both Bayesian and maximum likelihood (frequentist), always getting the same answer. We eventually accepted what the results were telling us, and we hope that you will be just as fascinated by our results as we still are.

The paper is published in Nature Ecology & Evolution here: http://go.nature.com/2eYPZDD