Further observations on the epidemiology and spread of epizootic haematopoietic necrosis virus (EHNV) in farmed rainbow trout Oncorhynchus mykiss in southeastern Australia and a recommended sampling strategy for surveillance

Epizootic haematopoietic necrosis virus (EHNV) is an iridovirus confined to Australia and is known only from rainbow trout Oncorhynchus mykiss and red fin perch Perca fluviatilis. Outbreaks of disease caused by EHNV in trout p opulations have invariably been of low severity, affecting only 0+ post-hat chery phase fingerlings <125 mm in length. To date the virus has been demon strated in very few live in-contact fish, and anti-EHNV antibodies have not been found in survivors of outbreaks, suggesting low infectivity but high case fatality rates in trout. During an on-going study on an endemically in fected farm (Farm A) in the Murrumbidgee River catchment of southeastern Ne w South Wales, EHNV infection was demonstrated in 4 to 6 wk old trout finge rlings in the hatchery as well as in 1+ to 2+ grower fish. During a separat e investigation of mortalities in 1+ to 2+ trout on Farm B in the Shoalhave n River catchment in southeastern New South Wales, EHNV infection was demon strated in both fingerlings and adult fish in association with nocardiosis. A 0.7 % prevalence of antibodies against EHNV was detected by ELISA in the serum of grower fish at this time, providing the first evidence that EHNV might not kill all infected trout. EHNV infection on Farm B occurred after transfer of fingerlings from Farm C in the Murrumbidgee river catchment. Wh en investigated, there were no obvious signs of diseases on Farm C. 'Routin e' mortalities were collected over 10 d on Farm C and EHNV was detected in 2.1 % of 190 fish. Tracing investigations of sources of supply of fingerlin gs to Farm B also led to investigation of Farm D in Victoria, where the pre valence of anti-EHNV antibodies in 3+ to 4+ fish was 1.3 %. The results of this study indicate that EHNV may be found in trout in all age classes, nee d not be associated with clinically detectable disease in the population, c an be transferred with shipments of live fish, can be detected in a small p roportion of 'routine' mortalities and may be associated with specific anti bodies in a small proportion of older fish. Sampling to detect EHNV for cer tification purposes should be based on examination of 'routine' mortalities rather than random samples of Live fish. Antigen-capture ELISA can be used as a cost effective screening test to detect EHNV on a farm provided that sampling rates conform with statistical principles.