There is increasing pressure to develop alternative control strategies against the pathogen Gyrodactylus salaris, which has devastated wild Atlantic salmon Salmo salar in Norway. Hyperparasitism is one option for biological control and electron microscopy has revealed two ectosymbionts associated with G. salaris: unidentified rod-shaped bacteria, and the protist, Ichthyobodo necator. No endosymbionts were detected. The flagellate I. necator occurred only occasionally on fish suffering costiosis, whereas bacterial infections on the tegument of G. salaris were observed throughout the year, but at variable densities. Bacteria were seldom observed attached to fish epidermis, even when individuals of G. salaris on the same host were heavily infected. Wounds on salmon epidermis caused by the feeding activity of bacteria-infected G. salaris did not appear to be infected with bacteria. On heavily infected gyrodactylids, bacteria were most abundant anteriorly on the cephalic lobes, including the sensory structures, but no damaged tissue was detected by transmission electron microscopy in the region of bacterial adherence. Furthermore, transmission and survival of infected G. salaris on wild salmon did not appear to be influenced by the bacterial infection. The lack of structural damage and impact on G. salaris biology indicates that these bacteria are not a potential agent for control of gyrodactylosis. However, this may not be the case for all gyrodactylid–bacterial interactions and a review of bacterial infections of platyhelminths is presented.
(Accepted May 26 2006)