Fig. 2: Attraction to each bacterial strain correlates with its impact on fitness.

a Experimental scheme to estimate impact on fitness: A single worm was placed in the middle of a food patch, and surrounded by a copper ring to prevent it from escaping. Five hours later, the number of eggs was counted. b Number of eggs laid after 5 h on a food patch of DA1885, as a function of the density of the food patch (open circles). Dotted horizontal line: average number of eggs when no food was present. Dashed horizontal line: Threshold chosen to determine when the number of eggs increases with respect to the no-food baseline. Solid dot: Point at which the number of eggs crosses the threshold, which defines \({D}_{{{\mbox{fitness}}}}\). c Same as (b), but for 11 bacterial strains and without errorbars. d Density at which worms start laying more eggs (\({D}_{{{\mbox{fitness}}}}\)), versus density at which worms start being attracted to a food patch (\({D}_{{{\mbox{attract}}}}\)). Line indicates perfect proportionality between the two variables (\({D}_{{{\mbox{fitness}}}}=0.2{D}_{{{\mbox{attract}}}}\); the value of the proportionality constant has little consequence, since it depends on the thresholds chosen to define \({D}_{{{\mbox{attract}}}}\) and \({D}_{{{\mbox{fitness}}}}\)). Color and shape of all markers identify bacterial strain, following the legend in Fig. 1. All errorbars show 95% confidence intervals, calculated via bootstrapping; see Supplementary Table 1 for sample sizes and Supplementary Data 1 for the data and computer code that generate this figure.