Veterinary Microbiology 98 (2004) 169–174

Avian circovirus diseases: lessons for the study of PMWS

Daniel Todd∗

Department of Agriculture and Rural Development for Northern Ireland, Veterinary Sciences Division,Stoney Road, Stormont, Belfast BT4 3SD, UK

Abstract

The diseases associated with psittacine beak and feather disease virus (BFDV), pigeon circovirus (PiCV) and goose circovirus(GoCV), which can be classified with porcine circovirus type 2 (PCV2) as members of the genusCircovirus of the familyCircoviridae, have clinico-pathological features in common with post-weaning multisystemic wasting syndrome (PMWS),with which PCV2 infection is causally associated. Intracytoplasmic botryoid inclusions within macrophages and depletionof T and B lymphocytes are common histopathological features, and, in each case, affected animals usually exhibit ill-thriftand a predisposition to secondary infections, that is suggestive of an underlying immunosuppression. Although these aviandiseases have been the subjects of relatively little research, their study can provide directly applicable lessons in the areas ofdiagnosis, epidemiology, pathogenesis and disease control for those charged with investigating PMWS. In keeping with itstaxonomic separation as the only member of the genusGyrovirus, the disease caused by chicken anaemia virus (CAV) differshistopathologically from the other circovirus-associated diseases. Most notably, the target cells of CAV have been identified ashaemocytoblasts and precursor T lymphocytes, with lymphocyte depletion, which affects T cells only, occurring in cells directlyinfected with the virus. Nonetheless, CAV is the best-researched circovirus and provides excellent examples of both virus-inducedimmunosuppression and virus–virus interactions. The study of CAV-induced disease can therefore provide valuable, if lessdirectly applicable lessons.© 2003 Elsevier B.V. All rights reserved.

Keywords: Circoviruses; Avian circoviruses; Porcine circovirus; Pigeon circovirus; Goose circovirus; BFDV; CAV; PMWS

1. Introduction

TheCircoviridae is a recently established virus fam-ily, which comprises small, non-enveloped, spheri-cal viruses that are unique among animal viruses inhaving circular, single-stranded DNA genomes (Toddet al., 2000). Three avian viruses are currently clas-sified as members of the family. Psittacine beak andfeather disease virus (BFDV) is grouped with porcine

∗ Tel.: +44-2890-525814; fax:+44-2890-525773.E-mail address: daniel.todd@dardni.gov.uk (D. Todd).

circoviruses types 1 and 2 (PCV1, PCV2) as mem-bers of the genusCircovirus. Chicken anaemia virus(CAV) was recently classified in a separate genus,Gy-rovirus, on the basis of its unique genome organisa-tion (Pringle, 1999), and pigeon circovirus (PiCV) isclassified as a tentative member of the family.

In the 1990s, histopathology- and electron micro-scopy-based evidence indicated that additional avianspecies are infected with circovirus-like viruses. Theseinclude Senegal doves, canaries, finches, ostriches,geese and Southern black-backed gulls (Todd, 2000).Recent genome isolation and sequencing studies in our

0378-1135/$ – see front matter © 2003 Elsevier B.V. All rights reserved.doi:10.1016/j.vetmic.2003.10.010

170 D. Todd / Veterinary Microbiology 98 (2004) 169–174

laboratory have confirmed that pigeons, canaries andgeese are infected with novel circoviruses that can nowbe added to the list of members comprising the genusCircovirus, and preliminary results suggest that thisis likely to be the case with the circovirus-like virusthat infects Southern black-backed gulls (Todd et al.,2001b,c). Given that avian species comprise five of thesix animal species recognised as being infected withcircoviruses, it would be unwise for those involved inthe study of the role of PCV2 in post-weaning multi-systemic wasting syndrome (PMWS) to ignore workon the diseases with which these avian circovirusesare associated.

This paper will be review-like in nature and, in themain, the references cited will be relevant review arti-cles. It comprises two major parts. In the first, aspectsof the diseases associated with BFDV, PiCV and goosecircovirus (GoCV), which are grouped or likely to begrouped with PCV2 in the genusCircovirus, will besummarised. These diseases have features in commonwith PMWS and, as a consequence, there are directlyapplicable lessons for those studying PMWS. In thesecond part, disease-related features of CAV infec-tions will be summarised. CAV has been the subject ofconsiderable research, and much of its epizootiologyand pathogenesis are understood. However, in keepingwith the taxonomic separation of this virus, diseaseassociated with CAV differs in a number of importantfeatures from PMWS.

2. Diseases associated with avian circoviruses ofthe genus Circovirus

Relatively little is known about the diseases withwhich avian members of the genusCircovirus areassociated. This is largely because they have beenrecognised very recently, they do not infect commer-cially more important species such as chickens orturkeys and, to date, they cannot be propagated in cellculture. The best-studied disease is psittacine BFD,which was first described in Australian Cockatoosin the 1980s (Gerlach, 1994). BFD is confined topsittacine birds with over 60 species of free-rangingor captive birds known to be affected. Typically, thedisease is detected in young birds of less than 3 yearsand is clinically characterised by a chronic, progres-sive loss of feathers, and, in some species, deformities

to the beak and claws can also occur (Pass and Perry,1984). Although affected birds may live for manyyears, more frequently the condition lasts for sev-eral months up to a year, with death usually result-ing from secondary bacterial, chlamydial or fungalpathogens. Circovirus-associated disease of pigeonswas first reported in 1993, has since been describedin North America, Australia and many Europeancountries, and probably has a worldwide distribution(Woods and Shivraprasad, 1997). In general, youngbirds between 1 and 12 months are affected. Clin-ical signs include ill-thrift, lethargy, anorexia, poorrace performance and a variety of other signs thatare thought to be due to infections with concurrentbacterial, fungal and viral pathogens. The diseasecaused by circovirus-like viruses in geese was thesubject of a single report in 1999 (Soike et al., 1999).The commercial goose farm in which circovirus wasdiagnosed, exhibited runting, adverse production andincreased mortality, which were due, in part, to infec-tions with Riemerella anatipestifer and Aspergillusfumigatus.

Although the major histological lesions associatedwith BFDV were detected in the developing feathershaft and feather pulp, lymphoid necrosis and atrophyalso occurred in the thymus and bursa of Fabricius(BF) of affected birds (Pass and Perry, 1984). The mainhistological changes associated with circovirus infec-tions of non-psittacine birds are those of the primaryand secondary lymphoid tissues (Woods and Latimer,2000). These are commonly observed in the BF andmay range from lymphofollicular hyperplasia to lym-phoid necrosis, cellular depletion, and cystic atrophy.In psittacine and non-psittacine species, the frequentdetection of globular or botryoid, basophilic intracy-toplasmic inclusions within macrophages also appearsto be a characteristic feature. Electron microscopy re-vealed that these inclusions consisted of aggregates,some paracrystalline, of tightly packed virus-like par-ticles. In situ hybridisation (ISH), performed in ourlaboratory, has recently shown that, in naturally in-fected birds, PiCV can be detected in a wide rangeof tissues including BF, spleen, thymus, liver, kidney,brain, crop and intestine, and often when microscopiclesions are not evident (Smyth et al., 2001). Using im-munohistochemical procedures, BFDV has also beenshown to exhibit a widespread tissue distribution indiseased birds.

D. Todd / Veterinary Microbiology 98 (2004) 169–174 171

From this summary of clinico-pathological char-acteristics, it is evident that diseases associated withBFDV, PiCV and GoCV share common features withPMWS, and, on this basis, examination of the researchperformed with these avian circoviruses and their dis-eases may be useful in an attempt to identify usefullessons relating to areas such as diagnosis, epidemiol-ogy, pathogenesis and disease control of PCV2 infec-tions.

With the exception of psittacine BFD, which isdiagnosed on the basis of feathering abnormalitiesand demonstration of virus antigen or nucleic acid,circovirus infections of other avian species are cur-rently diagnosed using histology and electron mi-croscopy. The availability of nucleotide sequencefor PiCV and GoCV has allowed our laboratory todevelop DNA-detecting tests (D. Todd, unpublishedresults), with which useful data about the prevalenceof infections have been produced. PCR and ISH,performed with BF samples, have demonstrated thatPiCV infections are highly prevalent, occurring at asubstantially higher level than that detected by histol-ogy. Virus can be detected in birds that are clinicallyand histopathologically normal, suggesting that, aswith BFDV infections, many PiCV infections may besubclinical. This calls into question the usefulness ofPCR as a tool for diagnosing disease as opposed toinfection. Dot blot hybridisation (DBH) assays havealso indicated that tissue samples from some pigeonsand geese contain relatively large amounts of virusDNA, and these findings prompt the question as towhether virus load correlates with the severity ofcircovirus-induced disease.

Regarding epidemiology, results to date indicatethat different avian species (or groups of relatedspecies, e.g. psittacine species) are infected by dif-ferent circoviruses, suggesting that circoviruses arehost-specific and arguing against the involvement ofanother animal species, especially birds, as vectorssupporting virus replication in the spread of PCV2.Little published information exists as to whether anyof the avian circoviruses comprise more than oneserotype or pathotype. Although the presence of largeamounts of BFDV in affected feather material, faecesand crop washings make it likely that most infectionsresult from oral transmission, there has been some ev-idence suggesting that vertical (egg) transmission alsooccurs. So far, there is little information regarding the

transmission routes used by PiCV and GoCV (Woodsand Shivraprasad, 1997; Soike et al., 1999).

Pathogenesis studies have been confined to BFDand here, there may be valuable lessons for thoseinvestigating PMWS (Gerlach, 1994). Evidence sug-gests that BFDV is epitheliotropic, targeting the basalepithelial layer of the feather and feather follicles.The detection of intranuclear inclusions in thesecells is consistent with expectation that the smallcircovirus genome, which possesses a very limitedprotein-coding capacity, will be highly dependent onthe host cell’s DNA replication machinery, and thathigher levels of virus replication will occur in activelydividing cells, such as those of the basal epithe-lium. Virus-containing inclusions are also commonlydetected in the cytoplasm of macrophages and uncer-tainty exists as to their origin. Because circovirusesare thought to require cellular DNA replication, it hasbeen suggested that most of the accumulated virusin the inclusions is more likely to have resulted fromthe phagocytosing activity of these cells rather thanfrom endogenously replicated virus. The mechanismby which lymphoid organs such as the thymus andBF become depleted in BFD-affected birds is also amatter for speculation. Given that virus antigen is notcommonly detected in lymphocytes, this depletionis unlikely to have resulted from direct destructionof the cells by the virus, but is more likely to haveoccurred by an indirect mechanism perhaps involvinga cytokine-mediated reduction in lymphoid tissue de-velopment. The lymphoid depletion appears to persist,predisposing BFD-affected birds to opportunistic sec-ondary bacterial, chlamydial, and fungal infections.Some of the features of BFD pathogenesis are gener-ally similar to those exhibited by circovirus diseasesof pigeons and geese, suggesting that their study maybe useful for those investigating PMWS, where sim-ilar uncertainty concerning viral pathogenesis exists.

Attempts to prevent and control diseases causedby avian members of the genusCircovirus haveagain been limited to BFD (Gerlach, 1994). Cap-tive psittacine birds can be kept free from infectionby restricting their exposure to infected birds orvirus-contaminated environments such as cages. Thisstrategy cannot be applied to free-ranging birds suchas pigeons. Given that infections with circovirusesare likely to be prevalent and these viruses are prob-ably highly resistant to inactivation, eradication is

172 D. Todd / Veterinary Microbiology 98 (2004) 169–174

unlikely to be regarded as an option for disease con-trol. Preliminary work has shown that maternal anti-body, induced by experimental inactivated vaccines,can be protective against BFDV-associated disease.Since BFDV cannot be grown in cell culture, such vac-cines were produced by inactivating virus, which wascollected from disease-affected birds. On the groundsthat a more efficient antigen production is requiredand that circoviruses are difficult to inactivate, it isreasonable to expect that a subunit vaccine based onvirus proteins produced by recombinant DNA-basedexpression would be worthy of evaluation.

3. Disease associated with chicken anaemia virusof the genus Gyrovirus

Chicken anaemia virus is the best-studied memberof the family Circoviridae and, despite their rela-tively recent recognition in 1979, much is now knownabout its role as a disease-causing agent (McNulty,1991; Bülow and Schat, 1997; Adair, 2000; Toddet al., 2001a). A major reason for the rapid progressmade was the demonstration that the virus couldbe serially propagated in the Marek’s disease virustransformed chicken lymphoblastoid cell lines suchas MDCC-MSB1. This finding facilitated the iden-tification and characterisation of the causal agentand allowed the development of virus neutralisationand indirect immunofluorescence tests for detectingvirus-specific antibodies in chicken sera. Serologicaltesting demonstrated that CAV infections are highlyprevalent worldwide, and that the chicken is probablythe only host species. Cell culture isolates, originat-ing from throughout the world, have been shown tobelong to a single serotype. However, antigenic dif-ferences have been detected using virus-neutralisingmonoclonal antibodies. Comparisons indicate that thegenome nucleotide sequences of CAV isolates arevery similar irrespective of country of origin, and ata level similar to that exhibited by PCV2 isolates.

CAV infections are associated with clinical diseasein very young birds and subclinical disease forms inolder birds (McNulty, 1991). In the field, the acuteclinical disease occurs when in-lay breeders becomeinfected and virus is transmitted vertically via the eggto progeny chicks. The progeny chicks are normal athatching, the first signs of the disease being observed

when the birds are between 7 and 14 days old. Affectedbirds are generally depressed and mortalities of up to60% have been reported, although levels of 5–15%are more usual. The associated gross lesions includeanaemia, pale bone marrows, and thymic, splenic andbursal atrophy, which may vary in severity dependingon the presence of other pathogens. In addition, char-acteristic skin lesions, which are prone to secondarybacterial infections and gangrenous dermatitis, are acommon occurrence. Thus, according to the present-ing signs and lesions, the disease has been variouslycalled anaemia-dermatitis syndrome, blue wing dis-ease, infectious anaemia and haemorrhagic syndrome.The disease can be reproduced experimentally by inoc-ulating 1-day-old specific pathogen free (SPF) chicksusing parenteral routes. At 14 days post-infection,post-mortem examination shows that the majority ofthe birds are anaemic (haematocrit values of less than27), have pale bone marrows and thymic atrophy. Al-though differences in the pathogenicities of Japaneseisolates have been reported (Yuasa and Imai, 1986),published work indicates that all field isolates arepathogenic when tested experimentally.

The experimental disease model has provided thebasis for the current understanding of CAV patho-genesis (McNulty, 1991; Bülow and Schat, 1997).The major histopathological changes were atrophyof the haematopoietic elements of the bone marrowand severe depletion of lymphocytes in the thymusfollowed by hyperplasia of reticular cells. The largestamounts of CAV antigen were detected immunohis-tochemically in the thymus, spleen, bone marrow,proventriculus and ascending duodenum. Based onimmunocytochemical, ultrastructural and virologicalinvestigations, most evidence supports the view thatthe main sites of virus replication are the haemocyto-blasts in the bone marrow and precursor lymphocytesin the thymus. Following experimental infection of1-day-old SPF chicks, the cell populations of theblood and thymus begin to recover by about day16 and are fully restored by days 32–36. Althoughvirus can be recovered from some tissues as late asday 49, histological lesions and virus antigen cannotbe detected after 26 days, suggesting that diseaseis transient. Infection of SPF chicks at ages greaterthan several days rarely results in anaemia, and it hasbeen argued that this age resistance occurs becausethe older chickens display higher degrees of immune

D. Todd / Veterinary Microbiology 98 (2004) 169–174 173

competence and can therefore mount more rapid im-mune responses. Results from our laboratory indicatethat, although anaemia was not induced, histologicallesions have been detected in the thymuses from SPFchicks that were experimentally infected by the oralroute at 3-week-old (J.A. Smyth, unpublished results),and this suggests that the T cell function is adverselyaffected when older birds are infected with CAV.

The adverse effects of CAV infection on the host’simmune function has been demonstrated experimen-tally (Adair, 2000). Splenic lymphocytes recoveredfrom SPF birds, that were infected at either 1-day-oldby the intramuscular route or at 3-week-old bythe oral route, exhibited decreased transformationresponses to the mitogen Concanavalin A and de-creased interferon and T cell growth factor (TCGF,interleukin 2) production. In addition, macrophagesrecovered from similarly infected birds exhibited de-creased interleukin 1 production, bacteriocidal andphagocytic activity. Further experimental evidencethat CAV infections of chicks are immunosuppressiveincludes the apparent enhancement of pathogenicityof a range of co-infecting agents such as Marek’sdisease virus (MDV), infectious bursal disease virus(IBDV), reovirus, Newcastle disease virus,Staphy-lococcus aureus, and Cryptosporidium (Bülow andSchat, 1997). Reduced responses to vaccines againstMarek’s disease, Newcastle disease and infectiouslaryngotracheitis have also been reported. There isalso strong clinical and laboratory-based evidencethat the pathogenicity of CAV is also exacerbated byco-infections with viruses such as MDV, IBDV, re-ovirus and reticuloendotheliosis virus (REV). For ex-ample clinical anaemia can be induced in older birdsfollowing co-infection of CAV with reovirus or IBDV.It has been postulated that destruction of the chicken’sB lymphocytes by reovirus, IBDV and MDV impairsits ability to mount an effective antibody response,thereby allowing CAV to exhibit greater pathogenicity.

The ability of viruses belonging to the genusCir-covirus including BFDV, PiCV, GoCV and PCV2 topredispose their hosts to secondary infections and/orto exacerbate the pathogenicity of other pathogens ina manner similar to CAV is already established. How-ever, whether adverse effects on the host’s immunefunction by other infectious agents can increase thepathogenicities of these circoviruses in the way thatthe pathogenicity of CAV can be enhanced has yet to

be determined. Comparison of CAV-induced immuno-suppression with what is known about the immuno-suppressions induced by BFDV, PiCV and PCV2 mayprove fruitful to those investigating PMWS. Whilethe primary effect of CAV is depletion of the T cellsin the thymus with minimal effect on the B cells inthe bursa of Fabricius, the other circoviruses wouldappear to deplete both the B and T cell populations.This difference may influence their pathogeneses.Thus, the largely intact B cell population that existsafter CAV infection may allow the infected host tomount an effective neutralising antibody responsethat can quickly reduce the levels of circulating virusand produce what can be described as a transientinfection. In contrast, the depletions of the B and Tcell populations, that occur in the diseases inducedby PCV2, BFDV and PiCV, may severely impair thehost’s ability to mount an antibody response. As aconsequence these viruses may not be cleared effec-tively and a longer-term disease state will persist.

Examination of CAV-induced disease can provideadditional lessons in relation to disease control. Thewidespread occurrence of CAV infections, the highlevel of virus excretion and the ability of the virus toresist environmental degradation, all suggest that erad-ication of infections is not a practical option (McNulty,1991), and, for the same reasons, this is likely tothe case with all circoviruses. A vaccination-basedstrategy has been introduced for controlling the clin-ical disease in young birds. Although the majorityof breeder flocks seroconvert to CAV as a result ofnatural infections with horizontally acquired virus, se-roconversion in breeders can also be achieved throughthe use of live non-attenuated or partially attenuatedvaccines. Attention is focussing increasingly on theproblems caused by immunosuppressive, subclinicalCAV infections that occur in older birds includingbroilers. One possible approach would be to extend theduration of maternal antibody by vaccinating breed-ers with an inactivated vaccine. The development ofa cell culture produced inactivated vaccine has beenreported but is not in widespread use. This may reflectthe poor immunogenicity of non-replicating virus andthe relatively high costs of producing virus antigen.Similar difficulties will probably exist with inactivatedvaccines prepared for the other avian circoviruses,if cell culture propagation systems can be identified,and possibly with PCV2. With CAV, a recombinant

174 D. Todd / Veterinary Microbiology 98 (2004) 169–174

baculovirus-derived subunit vaccine has also been in-vestigated, but given that multiple inoculations wererequired to elicit sufficient antibody, this approach isunlikely to be commercially successful (Todd et al.,2001a). This may not be the case with PCV2, whereit has been reported that the single capsid proteincan be produced in immunogenic form using recom-binant baculovirus-expression systems (Nawagitgulet al., 2000). Whether young pigs can be protectedagainst PCV2-induced PMWS by vaccinating sowswith inactivated or subunit vaccines to boost maternalantibody has yet to be determined. Similarly, whetherattenuated PCV2 vaccines have a role to play in thecontrol of PMWS is unknown. With CAV it has beenpossible to introduce partial attenuation by passagingin chick embryos and, in our laboratory, substantialattenuation by subjecting the Cux-1 isolate to multi-ple passages in MDCC-MSB 1 cells. This traditionalapproach to producing attenuated vaccines may alsowork with PCV2, in view of the fact that the relatedPCV1 can be grown in other continuous non-porcinecell lines and that antigenic changes are introduced asa consequence. In addition, an understanding of CAVmolecular biology has allowed strategies for rationalattenuation to be developed. Although the molecularbiology of PCV2 is very different from that of CAV,from studies with PCV1, it is now much better under-stood and rational manipulation of the cloned replica-tive form DNA followed by transfection to produceinfectious virus remains an option for producing can-didate attenuated vaccines, should they be required.

In conclusion, the major lesson from this paper isthat a number of diseases with characteristics in com-mon with PMWS, exist in avian species in associationwith viruses such as BFDV, PiCV and GoCV, whichresemble PCV2 in being grouped as members of thegenusCircovirus.

Acknowledgements

The author gratefully acknowledges contributionsby J.H. Weston, B.M. Meehan, N.W. Ball, B.J. Borgh-mans and J.A. Smyth (Veterinary Sciences Division,Belfast, UK), D. Soike (Federal State Institute forVeterinary Diagnostics & Food Testing, Potsdam, Ger-many), J.P. Duchatel (University of Liege, Belgium),and V. Palya (CEVA-Phylaxia, Budapest, Hungary).

References

Adair, B.M., 2000. Immunopathogenesis of chicken anemia virusinfection. Dev. Comp. Immunol. 24, 247–255.

Bülow, V., von Schat, K.A., 1997. Chicken infectious anemia. In:Calnek, B.W. (Ed.), Diseases of Poultry, 10th ed. Iowa StateUniversity Press, Ames, pp. 739–756.

Gerlach, H., 1994. Circoviridae—psittacine beak and feather di-sease virus. In: Ritchie, B.W., Harrison, G.T., Harrison, L.R.(Eds.), Avian Medicine: Principles and Practice (Viruses).Wingers Publishing Inc., Lake Worth, Florida, Chapter 32,pp. 894–903.

McNulty, M.S., 1991. Chicken anaemia agent: a review. AvianPathol. 20, 187–203.

Nawagitgul, P., Morozov, I., Bolin, S.R., Harms, P.A., Sorden, S.D.,Paul, P.S., 2000. Open reading frame 2 of porcine circovirustype 2 encodes a major capsid protein. J. Gen. Virol. 81, 2281–2287.

Pass, D.A., Perry, R.A., 1984. The pathology of psittacine beakand feather disease. Aust. Vet. J. 61, 69–74.

Pringle, C.R., 1999. Proceedings of the XI International Congressof Virology on Virus Taxonomy, Sydney, Australia. Arch. Virol.144, 2065–2070.

Smyth, J.A., Weston, J.H., Moffett, D.A., Todd, D., 2001. Detectionof circovirus infection in pigeons by in situ hybridisationusing cloned DNA probes. J. Vet. Diagn. Invest. 13, 475–482.

Soike, D., Kohler, B., Albrecht, K., 1999. A circovirus-like infec-tion of geese related to a runting syndrome. Avian Pathol. 28,199–202.

Todd, D., 2000. Circoviruses: immunosuppressive threats to avianspecies: a review. Avian Pathol. 29, 373–394.

Todd, D., McNulty, M.S., Mankertz, A., Lukert, P.D., Dale,J.L., Randles, J.W., 2000. Family Circoviridae. In: Murphy,F.A., Fauquet, C.M., Bishop, D.H.L., Ghabrial, S.A., Javis,A.W., Martelli, G.P., Mayo, M.A. (Eds.), Virus Taxonomy.Classification and Nomenclature of Viruses. Seventh Reportof the International Committee of Taxonomy of Viruses.Academic Press, New York, pp. 299–303.

Todd, D., Adair, B.M., McNulty, M.S., Allan, G.A., 2001a. Animalcircoviruses. Adv. Virus Res. 57, 1–70.

Todd, D., Weston, J.H., Ball, N.W., Borghmans, B.J., Smyth, J.A.,Gelmini, L., Lavazza, A., 2001b. Nucleotide sequence basedidentification of a novel circovirus of canaries. Avian Pathol.30, 321–325.

Todd, D., Weston, J.H., Soike, D., Smyth, J.A., 2001c. Genomesequence determinations and analyses of novel circovirusesfrom goose and pigeon. Virology 286, 354–362.

Woods, L.W., Shivraprasad, H.L., 1997. In: Calnek, B.W. (Ed.),Diseases of Poultry, 10th ed. Iowa State University Press,Ames, pp. 1050–1053.

Woods, L.W., Latimer, K.S., 2000. Circovirus infectionsof nonpsittacine birds. J. Avian Med. Surg. 14, 154–163.

Yuasa, N., Imai, K., 1986. Pathogenicity and antigenicity of elevenisolates of chicken anaemia agent (CAA). Avian Pathol. 15,639–645.