Fig. 2: Selection is divergent in synonymous environments.

Six replicate populations of E. coli, founded from a single clone, were evolved in three different, yet synonymous sugar environments – glucose-galactose mixture, lactose, and melibiose. After 300 generations of evolution, growth assays were performed in triplicate for each of the evolved populations and the ancestor to measure growth rate(r) and yield (K). Triangular data points show r and K for each of the replicate populations, and the error bars correspond to the standard deviations obtained from technical replicates. The square blocks show the average r and K for the evolved populations in each of the environments (obtained as the average of the mean of the six replicates), and the error bars correspond to the standard deviation (obtained as the standard deviation of the mean of the six replicates). The circular blocks show the ancestor’s average r and K in each of the evolution environments, and the error bars correspond to the standard deviation obtained from technical replicated. For all the eighteen evolved populations, changes in r and K depended on the exact composition on the evolution environment. Clearly, selection acting on a population changes even with a minor change in the environment.