Metabolic plasticity in response to different environmental conditions is widespread across taxa. It is reasonable to expect that such plasticity should be adaptive, but only few studies have determined the adaptive significance of metabolic plasticity by formally estimating selection on metabolic rate under different environmental conditions. We used a model marine colonial invertebrate, Bugula neritina to examine selection on metabolic rate in a harsh and a benign environment in the field, then tested whether these environments induced the expression of different metabolic phenotypes. We conducted two experimental runs and found evidence for positive correlational selection on the combination of metabolic rate and colony size in both environments in one run, whereas we could not detect any selection on metabolic rate in the second run. Even though there was no evidence for different selection regimes in the different environments, colonies expressed different metabolic phenotypes depending on the environment they experienced. Furthermore, there was no relationship between the degree of plasticity expressed by an individual and their subsequent fitness. In other words, we found evidence for phenotypic plasticity in metabolic rate, but there was no evidence that this plasticity was adaptive. In the absence of estimates of performance, changes in metabolic rate should not be assumed to be adaptive.

We estimated survival, growth, and reproduction of each individual by checking colony absence/presence and counting the number of bifurcations (colony size) and the number of ovicells (external offspring bearing brood chambers; every ovicell contains a single larva) every two weeks over a period of six months.