DAO 72:193-199 (2006)  -  doi:10.3354/dao072193

ABSTRACT: A prospective longitudinal study of salmonid alphavirus infection in farmed Atlantic salmon Salmo salar L. was initiated in post-transfer smolts on a UK farm in July 2004 and continued for 320 d. Sampling was concentrated on a single caged population (C4) with serum and tissue samples collected and tested for viraemia, virus neutralising (VN) antibodies and viral nucleic acid by real time RT-PCR and by histopathology; 380 sera collected between Days 0 (D0) and 139 (D139) were consistently negative for both viraemia and VN antibodies. The first evidence of infection was detected on D146, when 4 out of 20 fish were found to be viraemic and 1 of 20 to be antibody-positive. On D153 only 2 of 20 fish was viraemic and 1 antibody positive. At the next sampling (D158) no viraemic or antibody positive fish were detected. Thereafter, one or two viraemic fish were detected on 6 occasions, including on D320. The prevalence of antibody-positive fish remained low (0 to 5%) until D192 after which time it rose irregularly to a peak of 57.9% on D320. Real time RT-PCR testing of sera was more sensitive than screening for viraemia, detecting a peak of 35% positive on D153 before declining. Histological lesions diagnostic for pancreas disease (PD) were observed at D146 and D153 only. In addition, mild cardiac and to a lesser extent brain lesions were frequently found after virus was detected, but not in earlier samples. No clinical signs or mortalities attributable to PD occurred throughout the study. This is the first detailed report of sub-clinical infection and highlights the usefulness of longitudinal surveys and the detection of virus and antibodies as diagnostic and epidemiological tools.

KEY WORDS: Salmonid alphavirus · Pancreas disease · Longitudinal survey · Sub-clinical infection · Viraemia · Serology · RT-PCR