An illustration of experimental selection, as a proportional shift between families during the experiment. The change in relative representation of individuals between the seven families defined as ‘unknown’, at different stages of the trials. A: Initial number of copepodids from each family at the infection. B: Number of lice present at termination of the common-garden trial (34 DPI). C: Final proportions of surviving individuals after the EB trial. Families are ranked by their performance in the EB-trial and plotted in falling order, from the darkest slice at 12 o’clock, representing the family scoring highest, and descending clockwise.