Abstract
Juvenile plaice Pleuronectes platessa ( n = 1281) were tagged and released at two locations 300 m apart on a 1 km long sandy beach. Most (>90%) of the fish were recaptured within 100 m of the release site (shown by the colour of the tag), with very few caught >200 m distance after 6 weeks. The changing spatial distribution of marked fish was adequately reproduced by a simple dispersal model with a single parameter: a 78% probability of remaining in a 100 m wide zone from one day to the next, with a 22% chance that fish move into an adjacent zone. In a subsequent similar study at the same beach, fish were either released at the point of capture ( n = 881) or transplanted to the alternate site ( n = 910) 100 m distant. After 6 weeks, transplanted fish moved along the shore towards their sites of original capture. Fish replaced at the point of capture showed no such movement along the shore. Further modification of the dispersal model to allow for a distinction between dispersal from home sites and from sites away from the original point of collection, was sufficient to reproduce the behaviour of the populations of both transplanted and control treatment groups. The likelihood of dispersal from home sites was much less than that seen at sites away from home. Juvenile plaice thus have a degree of long‐shore site fidelity not expected of a fish with strong depth‐related migration behaviour in a relatively homogenous habitat.