Discussion

Diseases of concern

Iridovirus consistent with a Megalocytivirus was diagnosed in four submissions of cichlids including red Oscars in two states. In one of the submissions of oscars there were several other pathogens noted and it is conceivable that immunocompromised fish with chronic iridoviral infection are more likely to become heavily infected with other pathogens and opportunists. Published reports of Megalocytivirus in ornamental fish usually report the disease in gourami (Go et al. 2006), however, the disease was not diagnosed in gouramis in this survey despite several previous cases having been received from fish in AQIS quarantine premises (Appendix 4). This virus was identified as a high priority risk in gouramis (Kahn et al. 1999). The severity of disease in the affected cichlids and the number of cases suggest that the two week quarantine period for gouramis and cichlids is warranted. The import risk assessment for these viruses should be reviewed in light of the cases seen during this survey and the occurrence of a similar virus in Murray cod which was attributed to infection with a virus originating in imported fish (Go et al. 2006). Implementing similar risk management practices to those for imported goldfish (ICON-AQIS) may be appropriate under the circumstances.

Epizootic ulcerative syndrome (EUS) was reported in seven cases of gouramis from a number of states (Table 1 and 2). This disease is listed as a disease of significance by the World Organisation for Animal Health (OIE) and is reportable in Australia to the Department of Agriculture, Forestry and Fisheries. EUS is endemic to Australia and, as such, was not considered in the risk assessment of the 1999 Import Risk Analysis (Kahn et al. 1999). Gourami in this survey appeared to have longstanding, subclinical infections and it is easy to imagine the dissemination of this organism by imported fish (David Alderman, personal communication, 2002; Lilley et al. 2003). There are no restrictions on the movement of live fish within Australia in relation to EUS and there is no published information about strain differences in the causative agent Aphanomyces invadans. However, it is likely that there are strain variations (Lilley et al. 2003; Heather McLetchie, pers. comm. December 2006, FRDC Projects 2001/621 and 2004/091). Species vary in their susceptibility to EUS and in their immune response following infection (Catap & Munday 2002). Catap found that three spot gouramis were less susceptible to developing severe clinical disease than other species that she studied. It is therefore likely that asymptomatic carrier fish are provided with health certificates in the country of origin because Australia’s current import requirements only require there to be “no clinical signs of infectious disease or pests” prior to export. There is no requirement for histological or microbiological examination of gouramis. Goldfish are the only fish currently required to have a surveillance and monitoring program for specific diseases including spring viraemia or carp and Aeromonas salmonicida (ICON-AQIS). As with most diseases, the stresses of transport, water quality aberrations and assimilation to a new environment before and after arrival in Australia would be sufficient to precipitate overt disease and mortality from EUS. It is also possible that these fish may have concurrent infection with opportunistic pathogens such as Aeromonas hydrophila. Therefore, because of the number of cases seen in this survey, revisiting the risk assessment for this disease seems warranted.

The presence of some unusual parasites such as the coccidian Goussia carpelli in the intestines of goldfish; Myxobolus sp. in goldfish; Sphaerospora sp. in guppies; lesions consistent with Cryptobia iubilans in cichlids; Centrocestur formosanus in goldfish and unidentified microsporidia in a variety of fish, is of interest. The risks associated with these parasites are unknown and thus have generally been interpreted as insufficient to pose a significant biosecurity risk, notwithstanding that the viviparous freshwater monogenean parasite Gyrodactylus salaris is of sufficient concern to be OIE listed. The import risk analysis (Kahn et al. 1999) was based on meeting the criteria of the SPS agreement and there was no attempt to assign risk to potential pathogens of uncertain risk status that may be entering Australia with ornamental fish. Cryptobia iubilans and other species of Cryptobia were not mentioned by Kahn et al. (1999) but there is evidence that Cryptobia species are now common parasites in Australian cichlids and are often diagnosed histologically in subclinically infected fish or in fish that have recovered from infection and have residual granulomas. Evans & Lester (2001) reported 66% of Gyrinocheilus aymonieri (sucking catfish) examined by them were carrying Cryptobia sp. (probably an undescribed species). What species of Cryptobia have established in Australia is unknown. Cryptobia iubilans causes serious disease in young or stressed fish (Yanong et al. 2004) and is one parasite that would warrant further attention during any review of the import risk analysis of ornamental fish. A parasite resembling the digenean Centrocestur formosanus was found in goldfish from a retail outlet in Western Australia in 2006 and has been reported in Turkey in imported goldfish (Yildiz 2005). The significance of the spread of this and other parasites will not be known until there is evidence of their establishment or non establishment in other regions or species.

A large number of bacteria were isolated from the fish but none of these are considered to be primary pathogens and none cause significant or reportable diseases. They are more likely to be opportunistic infections following the stresses of transport and less than optimal water quality.

Subclinical disease may go undetected and this is a risk that warrants further investigation in risk assessment. One disease that was identified as a biosecurity risk and for which disease management strategies were implemented after the 1999 import risk analysis was goldfish haematopoietic necrosis virus (GFHNV). Goldfish haematopoietic necrosis virus appears to have entered Australia in goldfish in the early 2000s since several laboratories in Australia diagnosed the disease at around the same time (Stephens 2002). This disease is caused by a Herpes-like virus, and fish appear to be carriers without showing clinical signs. When carrier fish are stressed they can develop clinical or subclinical disease and produce virus particles which can enter the environment and infect other goldfish. Another feature of this disease is that it targets the haematopoietic tissue that is largely responsible for the determining the efficiency of the immune system of fish. To illustrate this point, in 2004 the Western Australian Department of Fisheries received goldfish from an ornamental fish retailer for diagnosis. The fish had been imported from Malaysia seven weeks earlier and had a steady slow mortality throughout the seven week period despite the application of several anti-parasitic and antibiotic treatments. At the time of submission 80% of the fish had died. Histologically lesions in the haematopoietic tissue and gill were consistent with goldfish haematopoietic necrosis virus, but there were also dactylogyrid-like parasites and bacteria similar to Flavobacteria sp. in the gills and Aeromonas hydrophila was cultured from the blood of one fish. It was likely that the fish were dying from secondary infections resulting from immunosuppression caused by either subclinical goldfish haematopoietic necrosis or from an outbreak of clinical disease that had occurred prior to importation. It is possible that fish with this disease entered Australia as carriers or were subclinically infected or not suffering mortalities that were high enough to trigger action by AQIS inspectors.

Commonly used diagnostic techniques such as histology are unlikely to detect carriers of this and most other viral and bacterial disease and for this reason screening tests that can detect carriers need to be developed for a number of diseases if health certification and quarantine (both pre- and post-) are to be effective. GFHNV illustrates some of the deficiencies in the current AQIS risk management policy. Goldfish are required to have additional certification based on inspections by the Competent Authority in the country of origin every six month to ensure that there are no clinical signs of this disease. How “clinical signs” of disease are determined is not specified. It may be merely and visual inspection fish from the side of a pond in which case the disease is unlikely to be detected because diseased fish often display no gross external lesions. Histological examination of sick fish during clinical outbreaks of disease or a demonstrated absence of the infectious agent by regular surveillance using molecular techniques such as the one published by Goodwin et al. (2006) is needed if health certifications of freedom from infection are to have validity.

Additionally goldfish must be treated with an effective parasiticide within 7 days of export to Australia to manage infestations of the gill parasites Dactylogyrus vastator and Dactylogyrus extensus. Trichlophon, formaldehyde and sodium chloride are suggested as suitable treatments. Whilst these treatments will certainly remove some of these parasites, and also other parasites, their efficacy is likely to be extremely variable. These parasites are very small, infested fish show few clinical signs except increased respiratory effort during heavy infestations, are difficult to remove and eradicate from a population of fish (Thoney & Hargis 1991). They can produce a large number of eggs and populations can build up rapidly under suitable conditions, especially when fish are stressed. During the survey 5 submissions of goldfish reported as an incidental finding that dactylogyrid-like parasites were present on the gills. This finding is evidence that D. vastator and D. extensus could be entering Australia undetected.

A suggested course of action to address some of these concerns with the health of goldfish would be to implement a more detailed study of the disease status of goldfish in both imported fish in quarantine and in populations already in Australia. This should include molecular testing for GFHNV and morphologic identification to species of Monogenea found on gills.

Pattern of submissions

Logistically some issues arose during the survey. In some states there was confusion over whether fish presented for diagnosis were species that were targeted as part of the ornamental fish survey. This occurred most often in Western Australia where an ongoing diagnostic service is offered to AQIS for imported ornamental fish. As a result, some species were initially included that did not satisfy the selection criteria and these results were removed at a later date.

Analysis of trends is difficult because submissions showed a pattern but there were not enough submissions to assess whether this pattern was a true indicator of the level of health of imported ornamental fish. Perhaps the trend is a reflection of the number of imports received into each state, the species that were imported at particular times or a seasonal pattern in health of imports. For example, why were most of Queensland’s submissions during the first two months of the project and why was there only one submission from South Australia? AQIS could use its records to clarify some of these possibilities. Alternatively there may have been certain AQIS officers who submitted more samples because they were personally more committed to the project.

Tank records sometimes showed information on other tanks at the same quarantine facility and it was not uncommon to see that mortalities of target species in other tanks had exceeded the 25% needed to trigger submission as part of the present survey but no fish had been submitted. There were no comments on the sheets to explain why these fish had not been submitted. AQIS officer’s sometimes explained to us that all of the fish had died when they visited the premises. At other times they thought that management problems such as aeration failures had caused the high mortality but the records did not note such events. Such events should be recorded on tank records and more adequately investigated to rule out the possibility of disease contributing to the problem.

Quality of diagnostic reports

It was clear from the comments on the diagnostic reports received that a number of the pathologists had limited experience or training in fish pathology. For example, supporting a diagnosis of “mild parasitic branchitis”:

When checking the gills for ectoparasites an irregular pattern of melanin distribution along the filaments was noted. In addition, instead of the filaments being thin and 'filamentous', these were quite thick and appeared to be full of refractile 'globules'.

The appearance of these gill filaments seemed 'unusual', but it is a species of fish I have not seen before and am unsure whether this is 'normal' or not.

At times it was difficult to interpret findings. For example, a histological finding of “Moderate infestation of the gills with flukes with significant numbers of encysted trematodes” was followed in the comments section by “Flukes are small worm-like monogenean trematodes measuring up to 2mm in length……Gill flukes…..”. It is unclear whether the parasites noted in the gills histologically were Digenea that can correctly be described as flukes or trematodes, have a lifecycle involving more than one host species and can have encysted stages in fish tissues or whether the parasites were Monogenea such as Dactylogyrus sp. These gill parasites are often seen in ornamental fish and are colloquially referred to as “flukes” but this term can cause confusion when used by fish health professionals.

A case reporting findings in discus noted chronic granulomas in the spleen, skin and stomach and possible fungal hyphae in the spleen. There was no reference to the use of special stains or other techniques to confirm a diagnosis and it remains unclear whether the granulomas were the result of Mycobacteria, fungi or infection with Cryptobia sp. (a common cause of granulomas in the stomach wall of cichlids).

Another report described hepatic lipidosis at length without being able to report any other significant findings. Fatty liver is common in fish reared in aquaculture and is generally not noted as a significant finding unless other aspects of the history or analysis suggest that it may be contributing to ill health.

It was apparent that pathologists in laboratories without a designated specialist fish pathologist had access to photographs or diagrams of fish parasites and a good understanding of mammalian pathology. They extrapolated this knowledge to fish, however, there were limitations to this process because the significance of lesions, such as gill lesions, is difficult to assess without experience or training. One method of increasing the effectiveness of diagnoses would be to send histology slides or photographs of parasites seen in wet preparations to more experienced peers for comment.

This issue is of relevance to Biosecurity Australia in that diseases entering Australia may be misdiagnosed; but it is highly probable that overseas veterinary pathologists may also be struggling to interpret aquatic pathologies. The results of this survey suggest that Australia is not well placed to withstand scrutiny of routine aquatic animal veterinary diagnostic capability by overseas auditors.

Inconclusive fish origin

It would be interesting to traceback the red albino oscars with iridovirus and the gouramis with lesions typical of EUS but under the current AQIS import conditions database (ICON-AQIS) there is no requirement to trace these species any further than the premises immediately prior to export to Australia. It would be interesting to know whether these fish came from the same farm or whether there is a much larger problem with the majority of fish of certain species being infected globally. Since these problems were readily detected by histology in Australian laboratories, the basis on which the health certificates or additional health certificates are provided needs to be re-assessed in light of a review of the risk analysis for these disease agents.

Comparison of survey results and other Western Australian ornamental fish submissions

Appendices 4 and 5 compare the diagnoses from ornamental fish in quarantine with those from ornamental finfish submitted from other sources such as hobbyists, aquarium shops, aquaculturists and aquarium maintenance businesses in Western Australia prior to the present survey. There were three cases of EUS (all in gouramis), one of GFHNV, six of cryptobiosis and four of iridovirus (all in gouramis) from fish in quarantine but one, seven, six and nil respectively from fish not in quarantine. Comparison of this historical data with results from the present survey present some interesting findings that are worthy of further consideration. Diagnosis of the latter three diseases can be difficult and requires some previous fish diagnostic experience and/or electron microscopy. Goldfish recovering from goldfish haematopoietic necrosis virus are often infected with other agents because of their immunocompromised state and this can compound the difficulties. Cryptobia organisms are difficult to see histologically (hence the name, from the same root as “cryptic”) but produce characteristic granulomas, disease signs and clinical history. Iridovirus inclusions can be mistaken for amoebae.

It would seem that some diseases including GFHNV and cryptobiosis of cichlids are now established in ornamental fish in Australia. The relatively large numbers of cases of GFHNV seen by the Fish Health Unit in Western Australia between 2002 and 2005 may reflect outbreaks of disease when the virus entered previously naïve populations in Australia. Cryptobiosis is usually diagnosed in cichlids in Western Australia after some weeks of ill thrift following a change of ownership or as an incidental finding. Many cases of “Malawi bloat” or “bloat” in cichlids have histological lesions typical of cryptobiosis as well as infection with Spironucleus sp. Occasionally cryptobiosis is seen in juvenile fish in Western Australia. In the US it was reported in juvenile discus (Yanong et al. 2004) and there are other less comprehensive reports of the disease in cichlids from other parts of the world. This suggests that the relatively short time of holding in quarantine is more likely to result in ill thrift than mortality from this disease. Both cryptobiosis and goldfish haematopoietic virus now seem more likely to be diagnosed in fish outside AQIS quarantine facilities In Australia. In some cases the fish were previously in quarantine but were presented for diagnosis after failure to thrive in the days or weeks after leaving quarantine.

A submission of a large number of species from a retail outlet with an AQIS import licence during this survey found a number of unusual pathogens. These included Digenea, possibly Centrocestus formosanus, in goldfish gills. Similar lesions were also found in a submission from Queensland a few weeks later. Digenea have at least one intermediate host but there is the possibility that a suitable intermediate host would be present in Australia. There were also Dactylogyrid-like parasites in the gills of two varieties of goldfish and protozoan-like cells in the brain stem and spinal cord.

Decision making on fate of fish in quarantine

When laboratories receive fish from AQIS because of concern about high mortalities during the quarantine period there has been confusion and concern over who should make decisions about the fate of the fish, and their legal power to do so. Pathologists in some laboratories make recommendations to AQIS officers that fish should be destroyed when notifiable or other infectious diseases of concern such as EUS or iridovirus are diagnosed. Other pathologists are justifiably reluctant to provide such advice. The legal ramifications of destroying fish that are not affected by a notifiable disease need to be considered and some guidelines should be presented to diagnostic laboratories. Fortunately there has not been a major objection to the destruction of fish to date, perhaps because almost all of the fish in the consignment have died from the disease before the diagnosis is made.

Time delay between sample taking and diagnosis

It is reasonable to expect that provisional diagnoses will take up to one week and definitive diagnoses may take a considerable time longer or often may not be possible at all. Because most species have a mandatory quarantine period of only 1 or 2 weeks the diagnostic process will often not be completed by the end of the quarantine period. Perhaps this one reason why importers may be reluctant to provide samples for diagnosis and may prefer to destroy all fish in a tank.

Variation in diagnostic techniques

Diagnoses made during this survey varied considerably in their specificity. In most cases further testing could be have done to improve the accuracy of the diagnosis. For example all five diagnoses of EUS were made on the basis of histology with granulomatous myositis and large fungal hyphae in histological section. Culture or PCR were not used to confirm the diagnosis. Similarly with diagnoses of iridovirus infections the appearance of typical basophilic inclusions in histological sections, sometimes backed up by the finding of iridovirus-like particles in transmission electron microscopy, was the basis of diagnosis. PCR was not performed.

Similarly the identity of a few metazoan parasites were confirmed morphologically by specialists with an interest in the specific group of parasites, but more commonly a diagnosis was made from the appearance of parasites in histological sections and their similarity to those in the same fish species in published articles.

If Biosecurity Australia is to make any significant improvement in the accuracy of health certificates of imported ornamental fish it is imperative that the method of diagnosis is stated. Additionally Australia needs to supply details of acceptable test methods, as it has done for infectious viraemia of carp. As new and improved diagnostic methods become available these need to be included as the required test methods. Examples of this deficiency are the goldfish haematopoietic necrosis virus and Dactylogyrus in goldfish. There are currently no diagnostic methods stated and it is unlikely whether “gill flukes” found on goldfish have been identified by a specialist in the morphology of Monogenea. Likewise routine testing of goldfish histologically is unlikely to find evidence of goldfish haematopoietic necrosis virus. The recently published PCR for detection of this virus (Goodwin et al. 2006) might provide the ability to identify populations that are infected with or free of the virus. Similarly, as diagnostic tools are developed that should be validated and assessed for their potential use in screening imported fish.

Areas of concern

1. The Import Risk Analysis for gouramis and cichlids, especially in relation to EUS and Megalocytovirus (iridovirus) would appear to be in need of review in the light of recent publications and the results of this project.
   
   
2. In several laboratories there were no pathologists with specialised training in diagnosis of aquatic diseases and pathogens. This means that diseases entering Australia may be misdiagnosed, but it is highly probable that overseas veterinary pathologists are also struggling to interpret aquatic pathologies. Australia is not well placed to withstand scrutiny of routine aquatic veterinary diagnostic capability by overseas auditors.
   
   
3. At present, diagnostic techniques, sampling protocols and treatment methods are not specified for diseases for which management strategies and health certificates are required. Under the current Import Conditions (ICON) database these includes goldfish haematopoietic necrosis virus, Dactylogyrus extensus, D. vastator and Aeromonas salmonicida (atypical strain). The importation of asymptomatic carriers of disease will continue to be a problem until robust, rapid diagnostic techniques preferably based on PCR (for viruses) are developed and used as a basis of health certification.
   
   
4. Kahn et al. 1999, p.119. recommended that the Competent Authority, in countries producing large numbers of goldfish that are imported into Australia, should be audited. This audit process may need to be reviewed given that GFHNV is now in Australia and that fish are entering with dactylogyrid–like parasites.
   
   
5. Prior to this survey it would appear that testing of imported fish at the point of entry was rarely undertaken in Australia. Increase testing of batches of ornamental fish that suffer high mortality in quarantine and the provision of instructions to officers and laboratories on how to manage these batches after diagnosis would be useful.

Previous page | Contents | Next page