“Most experiments to study the evolution of bacteria and their population dynamics are performed with well-stirred test tubes. Bacteria from different populations are put in there and mixed to shambles. Ultimately, always one population takes over the complete tube; the others are exterminated. Such situations are not at all true to nature.
If you look in the soil or in your intestines you'll see that the natural habitat of bacteria is very structured with lots of cavities. If you want to understand how it is possible that different populations coexist in nature, then you need to mimic the living conditions of the bacteria.
That's what we do in our laboratory. We work like urban planners. Our default ecosystems consist of a series of 85 microscopic chambers linked to one another by a small channel. In such an ecosystem on a chip I study the competition between three E. coli strains. One of these bacterial strains produces antibiotics that kill the bacteria from the so-called sensitive strain. The third strain is the resistant strain; it doesn't produce antibiotics but it can't be killed by them either.
Put these three together and you have a version of the rock-paper-scissors game in which each strain can defeat one of the others. The sensitive one is vulnerable when exposed to the virulent strain, but it too has an advantage. In absence of antibiotics, it grows faster than the non-sensitive strain, since it doesn't invest energy in protection measures.
During my experiments I use a microscope to take a picture of the chambers every ten minutes, in order to visualize the population dynamics. Here in this video you can see very cool patterns of the three strains chasing each other.
We also want to study evolution. You know that famous experiment done by the American microbiologist, Richard Lenski, who grew 40,000 generations of E.coli? This experiment showed sequential evolution. In sequential evolution, one population is replaced by the next, which, due to a mutation, is better adapted. Our hypothesis however is that with spatially segregated coexisting populations, evolution will go much faster. To become very fit, as a bacterium, you sometimes need several mutations, which means that you might first have to go through a low to become fitter in the end. In a flask or test-tube, this isn't possible, since that bacterium would be wiped out immediately. In our on-chip ecosystems however different subpopulations of bacteria can simultaneously explore different evolutionary paths.”