Size-assorted fish shoals and the majority's choice

Similarity among group members may serve as a defence against visually hunt ing predators that preferentially attack individuals who are phenotypically different from the group majority. The presence of such odd individuals in an otherwise homogeneous group may, however, increase the vulnerability of the other group members as well. Individuals might thus be expected to for m uniform groups in order to decrease predation risk, not only in trying to avoid being odd in a group, but also when attempting to avoid being accomp anied by odd individuals. This hypothesis was tested with small and large t hree-spined sticklebacks, Gasterosteus aculeatus. Focal fish were offered t he choice between a shoal consisting of conspecifics that were all similar in body length to the focal fish (matching shoal) and a shoal in which one or a few individuals differed in body length from the focal fish (non-match ing shoal). In the control experiment, all individuals in the nonmatching s hoal differed in size from the focal fish. The control confirmed that indiv iduals preferably joined the matching shoal when the alternative option was to be odd in another one. However, when the alternative for size-assortati ve shoaling was to belong to the majority in a mixed shoal, the shoal choic e of individuals appeared on average to be random. Visual contact with a li ve pike, Esox lucius, did not affect the shoal choice pattern. Furthermore, despite the frequency-dependent nature of; the oddity effect, varying the number of odd individuals in the non-matching shoal did not have a signific ant effect on individual's shoaling decisions. These results suggest that s ize-assortativeness in fish shoals is not a. result of individuals avoiding being among the majority in a mixed group.