Fig. 3

Oviposition preference by female adults and larval performance on the six host-plant species in the study. Box-and-whisker plots of the total number of eggs laid per female on each host-plant species a. Host-plant species in the study are on the x-axis along with their classification based on plant use (core or extended). Ranking of plants according to female preference for oviposition b. Two kinds of rank-ordering of plants per female butterfly were performed; one based on proportions and the other on cumulative frequency of preferred plant for each paired choice (preference scores). Results presented are based on multiple comparison tests (k = 6, n = 12) after significant Friedman-rank sum (RS) based on proportions (RS prop) or frequencies (RS freq). Groups (Group prop and Group freq) are designated by letters so that plants with rank sums without the same letter are significantly different from each other determined by post-hoc correction (P < 0.04). Ranking of plants ranged from 1 (lowest RS) to 6 (highest RS) is shown above each plot in A. Performance of V. cardui caterpillars on six plants selected based on their history of interaction with nymphalids through evolutionary time c. Growth rate to pupation [log(mg)*days-1 ± SE] per plant. Rates without the same letter are different from each other according to multiple after a significant effect of plant on growth rate (Χ2 = 59.06, df = 5, P = 1.901e-11)