Analysis of evolution of expression plasticity
To test for the evolution of plasticity in gene expression responses, we examined the subset of transcripts with significant interaction effects in our differential expression analysis (see above). We grouped genes with significant interaction effects into categories for the evolution of plasticity outlined by Renn and Schumer (2013) based on post hoc differences between rearing environments within a population (p<0.05): (1) Assimilated: plasticity in the ancestral high-predation population but a loss of plasticity in the derived low-predation population; (2) Accommodated: a change in the degree, but not the direction, of plasticity in the derived as compared to the ancestral population; (3) Reversed: opposing directions of plasticity in high- versus low-predation populations; (4) Evolved plastic: no plasticity in the ancestral high-predation population but an emergence of plasticity in the derived low-predation population; and (5) Unclassified: genes that had a significant main interaction effect but no significant post hoc rearing differences and could therefore not be unambiguously classified into one of the other categories. All statistical tests and data visualization were performed in R (version 3.5.1; The R Foundation for Statistical Computing).