infectious pancreatic necrosis

International Common Names

• English: aquabirnavirus infection; aquatic birnavirus infection; infectious pancreatic necrosis and associated aquatic birnaviruses

English acronym

• IPN

Infectious pancreatic necrosis (IPN) is a highly contagious viral disease of young fish of salmonid species held under intensive rearing conditions (Hill, 1982; Wolf, 1988; Wolf et al., 1960). The disease most characteristically occurs in rainbow trout (Oncorhynchus mykiss), brook trout (Salvelinus fontinalis), brown trout (Salmo trutta), Atlantic salmon (Salmo salar), and several Pacific salmon species (Oncorhynchus spp.). Susceptibility generally decreases with age, with resistance to clinical disease in salmonid fish usually being reached at about 1500 degree-days (value obtained by multiplying the age in days by the average temperature in degrees centigrade during the lifespan) (Dorson and Torchy, 1981), except for Atlantic salmon smolts, which can be affected normally within the first weeks after transfer from fresh water to seawater, but also before seawater transfer (Smail et al., 1989). The first sign of an outbreak in salmonid fry is frequently a sudden and usually progressive increase in daily mortality, particularly in the faster growing individuals. Clinical signs include darkening pigmentation, a pronounced distended abdomen and a corkscrewing/spiral swimming motion. Cumulative mortalities in a tank may vary from less than 10% to more than 90% depending on the combination of several factors, such as virus strain (McAllister and Owens, 1995) and infectious dose (Okamoto et al., 1984), host and environment (Dobos and Roberts, 1983). For further details see reviews by Hill (1982), Reno (1999) and Wolf (1988).

The disease is transmitted horizontally via the water route. Vertical transmission has been demonstrated for brook trout and rainbow trout, but this has never been successfully demonstrated in Atlantic salmon (Ahne et al., 1989; Ahne and Negele, 1985; Dorson and Torchy, 1981). Surface disinfection of eggs is not entirely effective in preventing vertical transmission (Bullock et al., 1976).

The disease has a wide geographical distribution, occurring in most major salmonid-farming countries of North and South America, Europe, Asia and Oceania (Anderson, 1996; Crane et al., 2000).

IPN virus (IPNV), or viruses showing serological relatedness to IPNV, have been reported to cause diseases in some farmed marine fish species, such as yellowtail (Seriola quinqueradiata) (Nakajima et al., 1993), turbot (Scophthalmus maximus) (Castric et al., 1987; Mortensen et al., 1990; Novoa et al., 1993b), dab (Limanda limanda) (Olesen et al., 1988), halibut (Hippoglossus hippoglossus) (Mortensen et al., 1990, Rodger and Frerichs, 1997) and Atlantic cod (Gadus morhua). Subclinical covert infections have been detected in a wide range of estuarine and freshwater fish species, such as loach (Misgurnus anguillicaudatus) (Chou et al., 1993), pike (Esox lucius) (Ahne, 1978) and numerous other species in the families Anguillidae, Atherinidae, Bothidae, Carangidae, Catostomidae, Cichlidae, Clupeidae, Cobitidae, Coregonidae, Cyprinidae, Esocidae, Moronidae, Paralichthydae, Percidae, Poecilidae, Sciaenidae, Soleidae and Thymallidae (Reno, 1999). It has been suggested to reserve the term IPNV for when the virus is isolated from a salmonid fish or has been shown to be able to produce IPN in salmonids through challenge experiments. Otherwise the term aquatic birnaviruses or aquabirnavirus is used.

The causative agent, IPNV, is the type species of the genus Aquabirnavirus of the Birnaviridae virus family. Birnaviruses have a bi-segmented (called segment A and B), double-stranded RNA genome, which is contained within a medium-sized, non-enveloped, single-shelled, icosahedral capsid (see review by Dobos 1995). Isolates display wide antigenic diversity (Dorson and Torchy, 1981; Hill and Way, 1995; Melby et al., 1994; Okamoto et al., 1983) and fall into two serogroups that do not cross-react in serum neutralisation tests (Ahne et al., 1989; Olesen et al., 1988; Underwood et al., 1977), with the majority of strains belonging to serogroup A, which comprises at least nine serotypes (Hill and Way, 1995). Isolates also show marked differences in degrees of virulence (Hill, 1982; Hill, 1992; McAllister and Owens, 1995).

Control methods rely on the implementation of control policies and of hygiene practices in salmonid husbandry, through the avoidance of the introduction of fertilised eggs originating from IPNV-carrier broodstock, and the use of a protected water supply (e.g. spring or borehole pond) where the ingress of fish, particularly possible virus carriers, is prevented. In outbreaks, a reduction in the population density (‘thinning out’) may help to reduce the overall mortality. There are vaccines available aimed to induce protection of Atlantic salmon smolts after seatransfer (Christie 1997) but there are few publication of results allowing for reliable judgement about vaccine efficacy, and little information is provided to validate the protection under field conditions.

Infectious pancreatic necrosis (IPN) is a disease which causes high mortalities in young salmonid fishes; the aetiological agent of the disease (IPNV) was the first virus to be isolated from teleosts (Wolf et al., 1960). In the time since that work, agents biochemically and serologically identical to IPNV - the aquatic birnaviruses - have been detected in a wide variety of both diseased and non-diseased fishes and other aquatic animals.

The early work on IPNV was mainly responsible for the development of fish virology. Techniques for the isolation of viruses from teleosts were modified from methods used for the isolation of mammalian viruses.

The virus itself has sufficiently unusual biophysical characteristics to warrant its place as the archetype of a new family of viruses, the birnaviruses (Dobos et al., 1979).

Serologically and biochemically related viruses induce a variety of disease syndromes in taxonomically diverse host groups, which also include aquatic invertebrates. The worldwide distribution of these viruses in cultured as well as wild fishes and shellfish provides significant potential for economic damage.

Infectious pancreatic necrosis and aquatic birnavirus-induced diseases can cause a significant economic impact on cultured fishes, especially salmonids. The cost can be as a consequence of lethal disease in young-of-the-year fish, or the destruction of infected stocks even in the absence of disease, or in Atlantic salmon smolts after sea transfer. The virus may be vertically transmitted; therefore, the detection of virus in broodstock often means the destruction of these animals. In the late 1980s, there were instances in the USA where 7 million young salmonids were destroyed to eliminate IPNV from a single hatchery (P. Walker, Colorado, 1988, personal communication) and others where detection of the virus in returning coho salmon resulted in the destruction of 2 million eggs (Olson et al., 1994); similar occurrences were also reported in Canada (B. Larson, Alberta, 1988, personal communication). In addition, regulations currently in place to reduce the dissemination of diseases of salmonids often result in restriction of the movement of IPNV carrier fish between local, regional and/or national boundaries. These restrictions have attendant adverse economic consequences to aquaculturists. In many instances, the slaughter of infected fish and/or the restriction of movement of fish may be more economically detrimental than the loss associated with direct mortalities.

In some countries there are mandatory destruction of salmonid stocks and restriction of movement after detection of IPNV. This is controversial. Since swim-up fry and fingerlings are generally affected, the impact of the disease can be ameliorated by spawning additional brood fish to compensate for the potential loss caused by the disease. Another perspective is the dissemination into watersheds of a virus unresponsive to therapeutic agents

which can readily infect susceptible native or cultured stocks in adjacent drainages. Given the myriad of variables which come into play in the generation of IPN and the potential for extensive, irreparable damage to stocks, it would seem prudent to adhere to more stringent rather than lax requirements. Resolution of the problem of IPNV management will be difficult and achieved only through a thorough knowledge of the epizootiology of the disease and an adequate database on both the disease and the virus.

Infectious pancreatic necrosis virus may also have an economic impact on aquaculture because of fish health inspections mandated prior to the movement or sale of fish across local, regional and national boundaries. Tests required for the detection of IPNV and other teleost viruses often represent an expense for aquaculturists. Again, this can be a contentious issue because the regulations are not often applied uniformly, even within regions regulated by one agency.

Control of the disease worldwide will also be difficult, given the ubiquitous nature of the agent.

Table. Natural diseases caused by, or associated with, aquatic birnaviruses. Included are isolations of aquatic birnaviruses in the absence of substantive proof of causal association with disease.

Diseases for which Rivers' postulates have been fulfilled

Diseases for which birnaviruses have been isolated from moribund animals

1, Wood et al., 1995; 2, Castric et al., 1987; 3, Sorimachi and Hara, 1985; 4, Sano et al., 1981; 5, Stephens et al., 1981; 6, Kusuda et al., 1989; 7, Bonami et al., 1988; 8, Schutz et al., 1984; 9, Chen et al., 1990; 10, Diamant et al., 1988; 11, McAllister et al., 1984; 12, Nagabayashi and Wolf, 1979; 13, Giorgetti, 1990; 14, Lo et al., 1991; 15, Hill et al., 1982.

Little has been done with natural IPN disease to determine the portal of entry, dissemination of virus to target organs or effects on the target cells, except histopathologically.

In experimental infections, the disease is manifest by 1 week postinfection and generally has run its course within another week or so (Wood et al., 1955; Reno, 1976; Reno et al., 1978). The acute nature of the disease is characteristic of all aquatic birnavirus-induced diseases, including Japanese flounder and yellowtail ascites, turbot disease, nephritis of eels and spinning disease of menhaden (Sano et al., 1981; Bonami et al., 1983; Stephens et al., 1983; Sorimachi and Hara, 1985).

Swanson and Gillespie (1981, 1982) experimentally infected 6-month-old or yearling brook trout with IPN (strain VR-299) by intraperitoneal injection and followed the distribution of virus in the gastrointestinal tract, kidney and blood. While no clinical disease was produced , high viral titres were found in the gut and kidney, and focal necrotic lesions were detected in the pyloric caeca within 3 days postinfection and in the kidney within 5 days postinfection. Indirect fluorescent antibody tests indicated the presence of foci of IPNV antigen in pyloric caeca, kidney and liver and the virus was detected in both the serum and the mononuclear cell fractions of the blood. Hornich et al. (1986) indicated that fry undergoing acute IPN had the most intense fluorescence in the gut epithelial cells. In contrast, in natural carrier brook trout no virus was detected in the plasma, although virus was detected in the leucocytic fraction (Yu et al., 1982). In addition, McAllister et al. (1987) found that the cellular phase of the ovarian fluid of carriers contained up to 400 times more virus than the fluid portion. The distribution of virus within the viscera is not uniform, with distinct tissue tropisms.

Transmission of the disease and epidemiology

In the classic Culture and Diseases of Game Fishes, Davis (1953) makes the following reference to M’Gonigle’s information (1940) on alleviating acute catarrhal enteritis, now assumed to be IPN: "M’Gonigle recommends that affected fish be planted in small streams where they can get natural food such as insect larvae. Since the disease is not due to infection by animal parasites or bacteria, liberation of the fish in natural waters can do no harm." It is unknown how many fish farmers adhered to this practice, which was promulgated before the discovery of virus infections in teleosts. It is possible that the panzootic nature of aquatic birnaviruses may be due in part to this type of remedy for a disease of unknown aetiology.

The epizootiology of IPN is complex and is not thoroughly understood. The disease affects young fish and postsmolts, the virus may be retained in the body after primary infection without causing clinical signs in older fish, and the virus may be transmitted vertically.

IPNV may be introduced to a hatchery by eggs, fry, human, fomites, different animals, and water. The impacts of these possible introductory alternatives may vary.

The prevalence of IPNV and IPN disease is not easily appraised, although the virus appears to be present at rather high levels in some instances.

Table. Reported IPNV carrier rates in various aquatic animals.

ATS, Atlantic salmon; BKT, brook trout; BNT, brown trout; RBT, rainbow trout. S, spleen; K, kidney; P, pyloric caeca; G, gonads; I, intestine; L, liver; F, faeces; PW, peritoneal wash; OF, ovarian fluid; B, blood.

1, Yamamoto, 1975b; 2, Wolf et al., 1961; 3, Reno, 1976; 4, Wolf et al., 1968b; 5, Frantsi and Savan, 1971b; 6, Hedrick and Fryer, 1982; 7, Billi and Wolf, 1969; 8, Munro et al., 1976; 9, Dorson, unpublished; 10, Dorson 1980, 1982; 11, Dorson, 1981, 1982; 12, Bucke et al., 1979.

While there is considerable information in the literature on prevalence levels, substantial variations exist within even small geographical locales and, temporally, even at the same facility. This diversity precludes easy analysis or compilation of information. It is generally accepted that, if the virus is present in a facility, the infection will remain for long periods in the absence of active measures to eradicate the agent. IPNV seems to be persistent once introduced into a hatchery.

For example, in one hatchery, the disease was present in the 1940s and mortalities were in excess of 90% in brook trout; by the late 1970s, the disease was still present, but mortality had dropped to around 40% (L. McCullogh, personal communication). Conversely, a brook trout hatchery known to have IPNV present at enzootic levels for many years has had no virus-induced mortalities for more than 25 years (P.W. Reno, personal observations). It is interesting to note that this hatchery has distributed IPNV-infected brook trout to many other aquaculture facilities, but the virus has not necessarily remained with all of the fish that were moved. For instance, one fish farm received fish from the facility in 1975 and inspection of the stocks indicated the presence of the virus. The only measures taken were to house the fish at the tail-end of the impoundments. When these fish were tested annually, starting in 1987, no virus was detected in them. This is one of only a few instances in which IPNV has been eliminated from a population of carrier fish. Bucke et al. (1979) report that in 1975 IPNV was detected in grayling at the outfall of an IPNV-contaminated hatchery, but there was no evidence of virus within 2 years in the same population of fish.

There is accumulating evidence that the mode of transmission within a hatchery may be a combination of vertical and horizontal transmission. It is most likely that the initial source of infection is from previously infected stocks or from homeothermic vertebrates or invertebrates. Horizontal transmission in hatcheries is most probably due to contact with the virus in faeces and urine of infected fish (Billi and Wolf, 1969; Frantsi and Savan, 1971b).

Little is known of the epizootiology of the other diseases caused by aquatic birnaviruses (branchionephritis of eels, whirling disease of menhaden and yamabe ascites of flat-fish). Much more work in this area is needed.

The aquatic birnaviruses are not restricted to hatchery environments or only associated with fish released from contaminated facilities. The virus is in the viscera of wild stocks in remote regions that have had no known plantings or contacts with hatchery-reared fish. Souter et al. (1986) isolated IPNV from nearly half of 229 anadromous Arctic char, Salvelinus alpinus, prespawners collected from the Mackenzie river delta and rivers in the Yukon Territory, where there was no known contact with hatchery-reared fish. The high level enzootic nature of the virus in this location indicates that natural infections may be widespread and that the maintenance of the carrier state in fish is efficient in the absence of the high viral densities found in rearing facilities. In another remote region of Canada (Hudson Bay), IPNV was enzootic in brook trout during the late 1970s and early 1980s (A.J. Sippel, personal communication); likewise, the virus was detected in wild Atlantic salmon in Loch Awe in Scotland (Munro et al., 1976). In addition, aquatic birnaviruses have been isolated from wild brown trout at the outfalls of IPNV-contaminated hatcheries in England and Wales

(Bucke et al., 1979). In Japan, yellowtail ascites virus was also detected in 15% of wild yellowtail fingerlings some distance from mariculture sites; nearly half of infected fish reared in the laboratory developed clinical disease (Isshiki et al., 1989). However, it cannot be assured that these populations were isolated from potential virus carriers in the sea or from fish upstream of the mouth of the bay. Routine culture of yellowtail in Japan employs feral fingerlings as mariculture stocks and these fish are probably the source of virus which causes the epizootics in cultured yamabe. Thus, it appears that the aquatic birnaviruses are indigenous to some fish populations, even wild stocks.

It appears that IPN disease is extremely contagious in hatchery facilities. Fomites and humans are apparently capable of transmitting the virus within and between fish-rearing facilities. The virus is transmitted both vertically and horizontally. In most instances, it appears that if no recourse is taken to eliminate IPNV from a trout culture facility, the virus will remain at the facility for years. Many of the facilities from which IPNV was originally isolated still have virus present (P.W. Reno, unpublished observations). The perennial pattern of infection and disease in facilities that harbour IPNV led to early suspicion that the virus was transmitted vertically. Wolf et al. (1968a) provided evidence that this was the case with brook trout. This was later confirmed (Sano, 1971; Yamamoto, 1975a; Reno, 1976; Hedrick and Fryer, 1982; Luqi and Zhizhuang, 1988). The exact location of the virus during embryonation is unknown, but it is extremely difficult to isolate virus from the egg. Fijan and Georgetti (1978) detected the virus at low levels in the chorion of the fertilized ovum. However, the virus is readily demonstrable after hatching (Reno, 1976; Hedrick and Fryer, 1982).

Other animals may also be involved in transmission of the virus within and between hatcheries. Sonstegård et al. (1972) infected several mammalian and avian species being able to vector IPNV and found infectious virus in the faeces of mink (Mustela sp.), great horned owl (Bubo virginianus) and chickens (Gallus domesticus) after 1 week. The virus was not detected in the faeces of great blue herons, common eider, or common merganser. Similarly, Eskildsen and Vestergård-Jørgensen (1973) isolated IPNV from gulls (Larus ridibundus) administered virus per os. Peters and Neukirch (1986) demonstrated that herons (Ardea cinerea) which were fed virus either by direct intubation or by ingestion of IPNV-injected trout harboured the virus at fairly high levels for about 30 days after the last feeding. While these findings lend credence to the possibility of transferring the virus by other animals, the consensus is that transfer in hatcheries most often occurs either horizontally within the same facility or watershed or through dissemination of contaminated eggs. It was also demonstrated that freshwater crayfish (Astacus astacus) were susceptible to IPNV, retained the virus in their tissues and haemolymph and shed virus constantly into the water (Halder and Ahne, 1988). Transmission from crayfish to trout eggs occurred, as did the infection of the crayfish from ingestion of IPNV-infected trout. There appeared to be no diminution of virus titre in the crayfish haemolymph or tissues for at least 6 months, but there was no indication of viral replication in the crayfish. These findings indicate that crustaceans and other aquatic invertebrates may be involved in the transmission of aquatic birnaviruses. Trout-rearing facilities whose water-supply is from open sources or sources that are not treated with ultraviolet irradiation or ozonation could have aquatic birnaviruses disseminated by both fish and crustacean vehicles, such as crayfish or copepods.

Mixing populations from several hatcheries, mode of transportation, and size of smolts at sea transfer are associated with increased risk of IPN outbreaks in seawater (Jarp, 1999). Intensified rearing conditions in the freshwater phase (superoxygenation, low water supply and high density) are shown to increase risk of IPN outbreaks in seawater (Jarp, 1999).

The movement of fish and especially eggs, in the international market must have contributed to the dissemination of the virus worldwide, although no thorough retrospective analysis has been done. Widespread movement of infected fish may have spread the virus to non-salmonid animals. However, it must be pointed out that non salmonids, especially invertebrates, show little or no evidence of disease when infected with aquatic birnaviruses and it is possible that the virus can be transmitted from these animals to salmonids, which under artificial rearing conditions were susceptible to disease. In this context, it has been reported by Hill (1982) that at least eight of the aquatic birnaviruses isolated from marine shellfish species were capable of producing characteristic signs of clinical IPN in experimentally infected rainbow trout.

Retrospective epizootiological studies of the diseases and infections caused by aquatic birnaviruses are difficult, due to a lack of consistency and accuracy in record keeping. Molecular markers for different virulence variants of IPNV within different serotypes can be revealed by RT-PCR and nucleotide sequencing (Bruslind et al., 2000; Santi et al., 2004; Shivappa et al., 2004). This implies that more precise characterization of IPNV isolates and the distribution of different variants in the cultivated and wild fish populations are needed. Such information is important for further understanding of IPN, risk factors for spread of the infection and for the implementation of proper control measurements.

Viruses belonging to the He and Tellina serotypes (types A4 and A5 of Hill and Way) have been detected exclusively in Europe, and Canada 1, 2, and 3, and Jasper (types A6-A9) have been confined to North America (Hill and Way, 1988b). This is interesting because it is known that trout infected with the WB serotype (A1) were often transferred to Europe from North America in the 1960s, but there is a single report of a WB serotype virus isolated in Europe (Novoa et al., 1993a). There are, however, some instances where the presence of disease in areas previously free from virus could be traced directly to the importation of infected eggs. Such importations occurred in South Africa (Bragg and Combrick, 1987a) and in Japan (Sano, 1971). In both instances, virus accompanied eggs imported from the USA. In other instances, there was circumstantial temporal evidence that the disease arrived with eggs imported into France from the USA (Besse and de Kinkelin, 1965), but subsequent serological evidence indicated that it was probably an indigenous strain (Wolf and Quimby, 1971). A heightened awareness of the disease in France by fish health workers, rather than the importation of virus in fish, was probably responsible for the detection of these strains. In other instances, there have also been indications that aquatic birnaviruses were transmitted to indigenous and cultured species from imported fish or eggs: to Chile from the USA (McAllister and Reyes, 1984; Espinoza et al., 1985), to Japan from Europe with European eels (Sano et al., 1981) and to Taiwan from trout imported from Japan (Hedrick et al., 1983). It is very difficult, however, to accurately differentiate between ‘indiginous’ aquatic birnaviruses and those that were brought into a country or region. The lack of efficient record-keeping at facilities, the lack of universally applied inspection regimes for viruses and the limited usefulness of the serological classification make epidemiological studies less than optimal.

Except for ‘classic’ IPN disease in freshwater trout, all of the diseases associated with aquatic birnaviruses affect either stenohaline or euryhaline animals: yellowtail and Japanese flounder ascites, sea bass, menhaden, eels and possibly clams.

The virus is quite stable, even under conditions that inactivate most other fish pathogens. The infectivity of the virus survives for nearly a year at 4°C and nearly 2 months at 15°C in buffer (Dorson, 1982) and also for long periods in seawater, brackish water and unsterilized fresh water (Moewus-Kobb, 1965; Toranzo and Hetrick, 1982). This stability increases transmission of the virus from carrier or clinically ill fish in rearing facilities to susceptible fish upstream or downstream in the watershed, or even to susceptible hosts in seawater. The ‘spread’ of the virus among freshwater, brackish and seawater environments is facilitated by the stability of the virus, the migration of anadromous and catadromous fishes and the predator-prey relationships. The latter is supported by the finding of a birnavirus-like agent in mass cultures of rotifers, Branchionus plicatilis (Comps et al., 1991), which could potentially be transmitted to larval sea bass (D. labrax), since rotifers are the primary food source for this species in culture. It was suggested that rotifers could be responsible for the epizootics of the disease in maricultured sea bass in France (Bonami et al., 1983).

Vertical transmission and the carrier state

Vertical transmission is usually defined as the transmission of virus from one generation to the next. The virus can either be within the contents of the gametes, often referred to as true vertical transmission, or alternatively on the surface of gametes or in ovarian and seminal fluids and mucus.

Vertical transmission of the virus via the reproductive products of salmonids has been substantiated (Wolf et al., 1968a; Sano, 1971; Yamamoto, 1975a; Reno, 1976; Hedrick and Fryer, 1982; Luqi and Zhizhuang, 1988); however, the actual method by which this occurs remains obscure. The perennial nature of the disease in salmonid hatcheries facilities indicates that this is an important transmission mechanism. This makes sense with trout which are reared in freshwater facilities, often with a well-water source, which would preclude infection via a water-borne route, although the possibility of contamination by birds or transfer by aquatic mammals cannot be ruled out. However, this cannot explain the recurring disease problems in salmonid seawater culture, which indicate that transmission from an environmental source is important. This can be compared to infectious haematopoietic necrosis, which was originally postulated to be vertically transmitted but now appears to be horizontally transmitted (Leong and Munn, 1991).

Wolf et al. (1968a) demonstrated that the virus was transmitted via reproductive products. Eggs and milt were collected from IPNV-positive fish and three pairings were made. The virus was isolated from two matings and frank disease was noted in one batch of the offspring. Bullock et al. (1976) found that the virus was transmitted to an uninfected facility via fertilized brook trout eggs transferred from a contaminated hatchery. However, the presence of the virus in the parent’s reproductive products or progeny does not invariably lead to the development of clinical disease. Reno (1976) found low levels (10^1.4 TCID₅₀ g^-1) of IPNV in all eight groups of progeny from matings of naturally infected brook trout which harboured fewer than 10^2.4 TCID₅₀ ml^-1 of IPNV in milt or ovarian fluid. Virus was not detectable during the embryonation period or until feeding commenced, but IPNV was detected in postfeeding fish for only 1 month. Similarly, in other species, vertical infection has been transient and sporadic. In Atlantic salmon, Smail and Munro (1989) detected virus for less than 1 h in eggs from brood fish exposed to IPNV at titres of 10⁶-10⁹ plaque-forming units (pfu) kg^-1 (injection) or 10⁷ pfu kg^-1 (immersion), and the resultant fry were negative for virus. However, when fry were infected experimentally, the virus was isolated even after 2 years.

Dorson and Torchy (1985) experimentally infected rainbow trout 4.5 months prior to spawning. In addition, at spawning, they infected milt and ova with high doses of IPNV Sp. Three of six infected females had detectable IPNV in ovarian fluid at spawning, whereas none of the males had virus in the milt. There was no evidence of IPN disease in the progeny. When milt was infected, however, IPN disease developed in the fry. Similarly, Ahne and Negele (1985) demonstrated that experimental infection of rainbow trout and Arctic char resulted in consistent isolation of virus from the chorion of the eggs during embryonation and posthatch. In fry, virus was isolated after only 1 month posthatch, but there was no evidence of disease in these fish. This indicates that eggshells may be involved in transmission of virus from one generation to the next and important for the horizontal transmission of IPNV to first feeding fry. The variations in results between the different experiments may be due to the species differences; brook trout are regarded as more susceptible to disease than other salmonids. However, different virulence variants of viruses, especially in experimental infections using cell culture-derived virus may explain differences in disease development. Only a few passages of virus in cell culture may cause attenuation of the virus (Santi et al., 2004).

Survivors of IPN epizootics remain carriers for up to 6 years (Wolf, 1966; Yamamoto,1975a; Reno, 1976). In a given population, the proportion of fish from which virus can be isolated varies drastically. In naturally infected brook trout, the prevalence ranged from 99% to less than 5% (Reno, 1976; unpublished observations). The latter is the prevalence used in fish health inspections for the detection of viruses in salmonids (Amos, 1985). The potential for the dissemination of the virus through the faeces has been demonstrated (Billi and Wolf, 1969; Frantsi and Savan, 1971b; Reno, 1976) and this ‘shedding’ of virus, albeit at low levels, may be responsible for the maintenance of the virus in susceptible fish. Carrier states have also been reported in survivors of epizootics associated with other aquatic birnaviruses: striped bass, turbot, yellowtail, eels and menhaden (Sano et al., 1981; Stephens et al., 1983; Sorimachi and Hara, 1985; Wechsler et al., 1987a).

There are no reported instances of zoonotic infections contracted from any aquatic birnaviruses.

Traditionally, IPNV was a disease of fry and fingerlings, with brook and rainbow trout being considered to be more susceptible than brown trout, lake trout (Salvelinus namaycush), coho and Atlantic salmon (Salmo salar). However, mortality varies considerably between outbreaks and has been related genetic susceptibility of the host (Ozaki et al., 2001 and Silim et al., 1982) and differing level of environmental stress (Frantsi and Savan, 1971 and McAllister and Owens, 1986). IPNV also infects non-salmonid species including eels, molluscs and crustaceans, all of which may act as carriers. The economic impact of IPN mortality of marine-reared Atlantic salmon has been of increasing concern, partly because of the size of the fish (Jarp et al., 1995).

Infection of susceptible species in fresh water can result in high mortality. Virus-associated mortality is rapid between 10 and 14°C, and at lower temperatures is prolonged. Water temperature, fish age and the virus strain affect the severity of the disease, as well as the establishment of covert infections in fish. IPNV replicates in kidney, pancreas, gonad, spleen and intestinal epithelium and may be shed from carrier fish via faeces, as well as through the seminal and ovarian fluids. Virus shedding appears to be dependent on stress. Although stress may induce a recurrence of IPN in fish 6-11 months old, older fish show no clinical signs.

The apparent increased incidence of IPN is considered to be the result of widespread fish and egg movements between countries, and because of increased sensitivity of diagnostic methods, resulting in improved surveillance practices.

IPNV is transmitted horizontally and vertically (Reno, 1999). However, vertical transmission has been confirmed only in brook trout and rainbow trout (Ahne et al., 1989; Ahne and Negele, 1985; Dorson and Torchy, 1981). The exact mechanism of egg entry or location of the virus within the egg is still unclear. Bebak et al. (1988) infected rainbow trout with IPNV by immersion challenge. The fish started to excrete the virus within 2 days of infection and shedding increased and then declined in less than 12 days post-exposure. From this study it was estimated that within 14 days, more than 75% of the population can be infected (Bebak et al., 1988). This gives rise to a rapid spread of IPNV.

Ethanol, methanol, iodophore and chlorine inactivate IPAV (Inouye et al., 1990), but it retains more than 90% of its infectivity after treatment with chloroform or ethyl ether, at pH 3.0 for 60 min. IPNV has been shown to be infective for several years at -70°C and for several months at 4°C. Smail et al., (1993) found that IPNV was not deactivated by an acidic pH unless the sample (silage) was heated for at least 2 h at 60°C. This confirms that IPNV is a robust virus with long survival in the environment.

This disease is not a zoonosis.