Statistical decomposition of passive and active phenotypic plasticity in traits under homeostatic regulation

Traits subject to homeostatic regulation exhibit both active and passive phenotypic plasticity, where trait dynamics are shaped by passive effects of the environment and active physiological regulation. We present a model that decomposes the temporal dynamics of such traits into parameters that describe the passive effect of the environment, the rate at which active regulation is adjusted (i.e., rate of plasticity), and the asymptotic magnitude of active regulation (i.e., capacity for plasticity). We apply this model to a dataset comprising 653 experiments documenting time-course changes in ion concentrations and osmolality of aquatic organisms following salinity shifts. Our model captures the diversity of trait responses, and meta-analyses reveal strong phylogenetic signals in all three parameters. Ray-finned fishes had faster regulatory responses, weaker passive plasticity (i.e., diffusion), and greater magnitude of active regulation than crustaceans. Trait-specific differences were also evident. Active regulation of magnesium was faster and of larger magnitude than the other ions and osmolality, implying strong selection for precise regulation of magnesium, which may play a key role in several physiological pathways. By disentangling the passive vs. active components of homeostatic trait regulation, our approach provides new opportunities for studying novel ecological and evolutionary aspects of phenotypic plasticity.