Extended Data Fig. 1: Bacterial fitness during and resulting from experimental evolution.

a, b, Bacterial fitness during the evolution experiment. a, Bacterial fitness in host across cycles of the evolution experiment measured as colony forming units (CFU) per worm population after 3.5 days of exposure to Ce_MY316. b, In the negative control, bacterial fitness was assessed on nematode growth agar in absence of the host. For each data point, bacteria were collected at the bottleneck time point of the noted cycle. Replicate populations (n = 6) are shown as separate thin lines, with the mean shown as a thick line. c, Mean CFU per individual host in a worm population for the evolved bacterial populations of cycle 10. Five L4 C. elegans larvae proliferated on evolved or ancestral bacterial lawns for 3.5 days (reaching F1 generation) and CFUs were extracted from the whole worm population. CFUs per population were divided by the number of worms in the population. Overall, results are shown as boxplots, with boxes indicating 25% above and below the median, which is given as the thick line within boxes; replicate populations (n=6) are indicated as individual data points. d, e, Dynamic changes in morphotype composition during the free-living phase of the host-associated life cycle for bacterial populations from the end of the evolution experiment. c, Results for the replicate populations from the host-associated evolution treatment. Box plots show median (center line), upper and lower quartiles (box limits) and the interquartile range (whiskers). d, Results for the replicate populations from the control treatment. Proportions of the different colony morphotypes (see graphical legend) is shown across time of the host-associated life cycle. Time point 0 is at the end of the host-associated phase, when bacteria are transferred to the free-living phase, which itself lasts 168 hours. Fdr-corrected beta-regressions were used to predict proportions and test for a change in proportions over time (see Supplementary Table 2).