Individual movement regulates many important ecological and evolutionary processes, and so it is essential to understand why animals move. Matching habitat choice (MHC), a process where individuals select and occupy a habitat best suited to their phenotypic traits through functional effects on ecological performance, is one mechanism thought to influence dispersal. Theory emphasizes the potential for MHC to associate phenotypes with habitats in populations occupying heterogenous environments. However, tests of MHC in natural populations are logistically challenging, and few studies are available. Furthermore, studies are biased toward short-term habitat selection responses, raising concerns over the long-term individual and population consequences of MHC behaviours. In a previous study, tagged pumpkinseed sunfish (Lepomis gibbosus) were experimentally translocated across littoral and pelagic lake habitats to disrupt pre-existing phenotype-habitat associations. Recapture site information demonstrated that individuals made movement decisions predicted by an interaction between their phenotype and the release habitat that subsequently re-established the association over a single summer. Using this same population, we investigate i) whether individual movements reform phenotype-habitat associations one and two years following experimental translocation and ii) evaluate the costs of phenotype-habitat mismatch to individuals.
We found evidence of a long-term effect of MHC behaviour, especially for fish released into the pelagic habitat. The direct effects of MHC on dispersal decisions, such as emigration from a release site and immigration to a new site, appear to decline over years. However, MHC may indirectly affect movement decisions through effects on other homing behaviours over these longer time scales. Contrary to expectations and other prior results, we found no evidence that individuals phenotypically mismatched to recapture habitat suffered any growth costs here, raising questions over the complexity of measuring individual performance under natural conditions. Pumpkinseed populations that inhabit deep postglacial lakes are often composed of coexisting littoral and pelagic ecotypes distinguished by a variety of traits. By contributing to the association between phenotype and habitat over a range of time scales, MHC behaviour contributes to reducing gene flow between ecotypes along with other mechanisms that will tend to promote their adaptive divergence.