Does plasticity enhance or dampen phenotypic parallelism? A test with three lake–stream stickleback pairs

Parallel (and convergent) phenotypic variation is most often studied in the wild, where it is difficult to disentangle genetic vs. environmentally induced effects. As a result, the potential contributions of phenotypic plasticity to parallelism (and nonparallelism) are rarely evaluated in a formal sense. Phenotypic parallelism could be enhanced by plasticity that causes stronger parallelism across populations in the wild than would be expected from genetic differences alone. Phenotypic parallelism could be dampened if site-specific plasticity induced differences between otherwise genetically parallel populations. We used a common-garden study of three independent lake-stream stickleback population pairs to evaluate the extent to which adaptive divergence has a genetic or plastic basis, and to investigate the enhancing vs. dampening effects of plasticity on phenotypic parallelism. We found that lake-stream differences in most traits had a genetic basis, but that several traits also showed contributions from plasticity. Moreover, plasticity was much more prevalent in one watershed than in the other two. In most cases, plasticity enhanced phenotypic parallelism, whereas in a few cases, plasticity had a dampening effect. Genetic and plastic contributions to divergence seem to play a complimentary, likely adaptive, role in phenotypic parallelism of lake-stream stickleback. These findings highlight the value of formally comparing wild-caught and laboratory-reared individuals in the study of phenotypic parallelism.