The history of bluetongue and a current global overview

Transcription

1 Vet. Ital., 40 (3), Global situation The history of bluetongue and a current global overview T.E. Walton Centers for Epidemiology and Animal Health, Veterinary Services, Animal and Plant Health Inspection Service, United States Department of Agriculture, 2150-B Centre Avenue, Fort Collins, Colorado , United States of America Summary Bluetongue (BT) was first reported more than 125 years ago when European breeds of sheep were introduced into southern Africa. BT viruses (BTV) have been identified in many tropical and temperate areas of the world. BT, the disease, is a phenomenon of ruminants in the temperate zones. There is little clinical disease in the tropical and subtropical areas of the world. At least 24 serotypes of BTV have been described. While the viruses are classified antigenically and taxonomically as BTV, each serotype is unique and may not cause BT, the disease. The BTVs are transmitted among ruminants by competent vector species of the genus Culicoides, i.e. biting gnats or midges. BTV serotypes exist with vector species of Culicoides in predictable, but finite, geographic and ecological cycles or ecosystems around the world. Despite the almost certain movement of livestock and Culicoides species between these ecosystems, there is little evidence that introduced BTV serotypes have been established in these ecosystems. Rather, periodic cyclic extensions and remissions of these virus-vector ecosystems permit the viruses and the disease to move into and recede from adjacent non-endemic areas in a pattern characteristic of many other known arthropod-borne viruses (arboviruses). Earlier publications suggested that a carrier state occurred in cattle infected as foetuses with BTV. No subsequent natural experiences or research support the hypothesis which has not been validated. The conclusions of the research are not accepted by the scientific community. It is logical, therefore, to propose that regulatory restrictions against the movement of cattle from BTVaffected countries be relaxed or eliminated. Keywords Bluetongue Culicoides Epidemiological cycle Non-tariff trade barriers Regulatory actions Vectors Virus-host interactions. Bluetongue, bluetongue viruses and bluetongue virus vectors in North America Bluetongue (BT) was first reported more than 125 years ago with the introduction of European breeds of sheep into southern Africa (12). Non-native sheep experienced a severe febrile disease with high morbidity and high mortality. The viral aetiology of the disease was demonstrated in Bluetongue virus (BTV) strains have been identified in many tropical and temperate areas of the world since that time from a latitude of approximately 40 N to 35 S. However, BT, the disease, is a phenomenon of ruminants in the temperate zones. There is little, if any, clinical disease reported in the tropical and subtropical areas of the world, except when non-native ruminants are introduced into a virus-endemic area. At least 24 serotypes of BTV have been described internationally. While the viruses are classified antigenically and taxonomically as BTV, each serotype is unique and may not cause BT, the disease. The BTVs are transmitted among ruminants by competent vector species of the genus Culicoides, which are known as biting gnats or midges. With the evolution of improved virological and serological techniques, BTVs were discovered outside Africa. With such discoveries and the extension of clinical BT into temperate areas previously considered to be free of BT, it was presumed that BTVs were emerging from Africa (7). More recently, it has been shown that BTV serotypes exist with vector species of Culicoides in predictable geographic and ecological cycles or ecosystems around the world. Despite the Veterinaria Italiana, 40 (3),

2 Global situation almost certain movement of livestock and Culicoides species between these ecosystems, there has been little evidence that BTV serotypes have been moved between these ecosystems and persisted. Rather, periodic cyclic extensions and remissions of these virus-vector ecosystems permit the viruses and the disease to move into and recede from adjacent nonendemic areas in a pattern characteristic of many other known arthropod-borne viruses (arboviruses). While clinical descriptions compatible with a diagnosis of BT were reported earlier, BTV serotype 10 was first isolated in the United States of America (USA) from sheep in California in 1952 (9). Despite the perception that the disease was emerging from Africa into other countries of the world, it was recognised early in the history of BT that, in fact, BT was primarily a disease problem when susceptible ruminants were introduced into an area with endemic virus activity (7, 12). While there is concern about the movement of BTV via infected animals and via infected insects, there are sufficient examples to suggest that movement of infected Culicoides, rather than of infected ruminants, has been the more serious threat (5, 27). Subsequently, other serotypes of BTV were isolated in North America. Serotypes 2 (1983), 10 and 11 (1955), 13 (1967) and 17 (1962) have been isolated in the USA. Culicoides sonorensis is the primary vector of serotypes 10, 11, 13, and 17, while the infrequently reported isolations of serotype 2 have been associated with populations of C. insignis that lie at the northernmost limits of the distribution of this southern species, central Florida (31). In Canada, serotype 11 has been isolated in a discrete focus in the Okanagan Valley of British Columbia, from cattle with mild clinical disease (8). Serotypes 10, 11, 13 and 17 have been identified in Mexico, north of Mexico City to the USA (29). C. sonorensis has been shown to be the primary vector of these serotypes (31). C. occidentalis has been demonstrated to be a probable vector of North American BTV serotypes in discrete and limited geographic foci in saline environments in the western USA. Clinical BT has been observed in sheep, cattle, bighorn sheep (Ovis canadensis), and white-tailed deer (Odocoileus virginianus) through much of the southern range of C. sonorensis in temperate climatic zones of the USA. While BTV infection in the USA causes severe and frequently fatal disease in sheep and white-tailed deer which may experience a peracute, lethal, haemorrhagic disease, the clinical disease in cattle is mild and infrequently observed. During the 1970s, severe epizootics of BT involving multiple BTV serotypes circulating in infected flocks at the same time were reported in sheep in the western states annually. In recent decades, for unknown reasons, epizootics have been less frequent and less severe with a decrease in the frequency of multiserotypic outbreaks. The predominant species of Culicoides found in the eastern provinces of Canada and the north-eastern and New England states of the USA is C. variipennis, which is considered to be a poor vector of BTV (31). Therefore, these areas of Canada and the USA are considered to be BT-free and BTV-free because they are vector-free. While the prairie provinces of Canada and the northern tier of the USA as far west as Montana have populations of C. sonorensis (previously identified as C. albertensis in Canada (10), these areas are considered low risk for, or free of, BTV. Indigenous populations are C. sonorensis genetically and morphologically, but there is no overt phenotypic expression of the genetically controlled oral susceptibility to the viruses and the vectorial capacity, e.g. maturation time of eggs and insect densities, which are environmentally influenced, is low. Clinical BT and transmission or isolation of BTV serotypes have not been demonstrated from cattle in these areas. From Guatemala (and presumably from southern Mexico and Belize) south-east through Central America to Panama, and in the islands of the Caribbean Sea, distinct ecological cycles have been identified with BTV serotypes 1, 3, 4, 6, 8, 12, 14 and 17 (31, 32). The primary vector of these serotypes is C. insignis. Clinical BT has not been described in ruminants found in these subtropical and tropical climatic zones. Despite the almost certain movement of livestock and Culicoides species between the BTV- C. insignis ecosystem of the Caribbean Basin countries and the continental BTV-C. sonorensis ecosystem of northern Mexico, the USA and Canada, there has been no indication that BTV serotypes, with the possible exception of serotype 17, have been moved and sustained between the ecosystems. While there is sparse serological and virological evidence of BTV activity in South America, little coordinated research has been published to define the virus-vector situation. It is presumed that C. insignis and the Caribbean Basin BTV serotypes are found in South America. In Australia and Oceania, serotypes 1, 3, 9, 15, 16, 20, 21 and 23 have been isolated. C. brevitarsis, and perhaps C. wadai, C. actoni, and C. fulvus, are the presumed primary vectors. There has been no clinical BT reported in cattle but, in Australia, there is evidence that some of the indigenous BTV 32 Veterinaria Italiana, 40 (3), 2004

3 Global situation serotypes are pathogenic for sheep. Likewise, C. brevitarsis appears to be the primary vector of serotypes 1-3, 9, 12, and 23 in South-East Asia. In Africa, C. imicola is the reported vector of serotypes 1-15, 18, 19, 22, 24 and 25. While C. imicola is distributed in the southern Mediterranean countries of Europe, incursions of BTV have been infrequent and periodic until recently. Past outbreaks through the Middle East into south-eastern Europe, the Iberian Peninsula, and recently, southern Italy and adjacent islands, have occurred. C. pulicaris has been reported to be a vector during recent BT outbreaks in Italy. The vector status of naturally occurring populations of a common Culicoides species in most of Europe, C. obsoletus, has not been established clearly, but historic evidence would suggest that this species is of low vectorial competence and capacity. It is anticipated that a much clearer picture of the international BTV and vector situations will be painted during the regional presentations. Vectors and vectorship The bluetongue chapter of the OIE Terrestrial animal health code (1) refers to Culicoides by genus, rather than by species, thus conveying the impression that all species of Culicoides are competent vectors of BTV serotypes. The logic that follows from this impression is that all countries could be considered at risk for BTV transmission since it appears that only Antarctica can be considered free from Culicoides. The fact is that in the absence of confirmed vector status, or competence of indigenous Culicoides species, it is illogical to imply that every country with Culicoides is at risk if imports of ruminants are permitted from countries in which livestock are considered infected with BTV. In the absence of a competent vector, the importation of infected or viraemic ruminants serves as no threat to an importing country. Restrictions do, however, place a potential trading partner at a competitive disadvantage that is not based upon science. Similarly, seropositivity of a candidate ruminant for proposed import in no way conveys an indication that the animal is viraemic and does not justify legislation against the importation of such animals. Seropositivity simply documents a previous historic experience with a BTV antigen and has no relevance to current infection or infectivity of the host. Furthermore, and perhaps more importantly, these animals are immune to the infecting and closely related viruses. The presence of a Culicoides species, and even isolation of BTV from a species is not evidence of vectorship or the vectorial capabilities of a species. Analogous to Koch s postulates for establishing the relationship of a micro-organism to a disease (23), there are similar basic criteria required to prove vector status of haematophagous insects. Just as finding an organism in a diseased tissue is not sufficient proof that the organism is the cause of that disease, isolation of a virus from an insect is insufficient evidence for differentiating true vectors from those species that are only incidentally infected because of the high titres of virus in the infected host. To prove vector status, four criteria must be met, as follows: 1) the isolation of the disease-producing agent from wild-caught specimens 2) the demonstration of its ability to become infected by feeding upon a viraemic host 3) the demonstration of its ability to transmit by bite 4) the confirmation through field evidence of the association of the infected arthropod with the vertebrate population in which the infection is occurring (30). Presence of culicoid species of unknown vectorial capability in a country considered BTV-free because it is north of 40 N or south of 35 S is insufficient justification for denying access by the livestock industry of BTV-affected countries to the markets of a BT-free country. Prior studies in cattle not validated Publications during the period from 1970 through to mid-1980 suggested that: persistent BTV viraemia and a BTV carrier state concurrent persistent circulating BTV and antibody activation of BTV viraemia by feeding of noninfected C. sonorensis chronic BTV excretion in semen occurred in cattle infected as foetuses by feeding of C. sonorensis infected with some North American BTV serotypes on pregnant cows (15). Subsequent studies in the same laboratory and similar studies by internationally recognised scientists at the same and in other laboratories were unable to reproduce these results. No subsequent natural field experiences and no experimental research have supported the original conclusions. The hypothesis has not been validated and the conclusions of the original research are no longer accepted as valid by the scientific community. Acknowledgement that the Veterinaria Italiana, 40 (3),

4 Global situation previously published results and conclusions were not valid was made at the Second International Symposium on bluetongue, African horse sickness and related orbiviruses (34) in which it was reported that numerous experimental and field studies failed to duplicate and validate the earlier conclusions (35). During the First International Symposium on bluetongue and related orbiviruses (2), it was confirmed that BTV infections of the bovine and ovine foetuses could produce developmental defects that resulted in death or deformities in offspring and poor viability of the newborn. BTV was not isolated from any foetus or newborn animal and it was considered unlikely that the deformities would be compatible with survival of the young bovid or serve as a threat for BTV persistence and transmission (16). Transitory excretion of BTV has been demonstrated in the semen of some experimentally infected bulls during the periods of highest viraemia, but BTV was never detected in the semen unless it was present concurrently in the blood (4). It was suggested that the presence of BTV in the semen was the result of infected erythrocytes contaminating the semen at collection rather than infection of cellular components (gametes) in the semen (11). Persistent or chronic BTV excretion was not confirmed and seroconversion was demonstrated in the infected bulls (4). In contrast, however, coincidental excretion of BTV in the semen of a single field study seronegative donor bull (26) and other retrospective, anecdotal field reports of reproductive problems controverted these controlled experimental studies (13, 14). In their evaluation of the available data, the WHO/FAO Working Team on Pathology at the First International Symposium observed that BTV infections of cattle in many parts of the world did not cause disease and concluded that the role of infected vectors was more important than was vertical transmission by transplacental transfer in the spread and persistence of BTV in a region (37). The WHO/FAO team concluded that: the possibility of a carrier state in cattle for BTV infection was not supported by the evidence in the literature during viraemia it was not unusual for BTV to be shed in the semen, as is the case with many other blood-borne virus infections field and other experimental data did not support the hypothesis that foetal developmental problems and carrier animals occurred naturally. In the Round-Table discussion on the international regulatory aspects of BT, five principles were used as a framework for the discussions as follows: 1) the movement of livestock and germplasm internationally is in the best interests of mankind 2) it is the first responsibility of regulatory agencies to protect the livestock industries they represent from losses caused by importation of pests or disease agents 3) regulatory policies must be developed uniformly, comprehensively and intelligently 4) regulatory policies work best when they are based upon incontrovertible scientific evidence 5) the scientific method underpinning regulatory decisions demands rigorous proofs (19). Evaluating the risk of importing BTV to a BTV-free country or of importing new serotypes to a BTV-infected country must be based upon sound science for the benefit of society and humanity rather than upon self-serving, protective attitudes or inaccurate or incorrect information. The latter results in unsupportable regulatory application of nonscientific non-tariff trade barriers to unrestricted international animal movement. The First International Symposium concluded with the recommendation that confirmation of the presence or absence of chronic infections with BTV in cattle and sheep must receive the highest priority (24). It was noted that failure to confirm chronic infections would remove BTV as a major international nontariff trade barrier. Likewise, it was suggested that if chronic infections in ruminants did not occur, there was no alternative explanation for virus overwintering and virus persistence through adverse environmental conditions, thus requiring a new, more plausible hypothesis. This recommendation was the basis for much of the research published in the Proceedings of the Second International Symposium in which it was concluded that the hypotheses related to persistent infections and carrier cattle were not supported by sound science and, therefore, that the BTV-related regulatory restrictions on international movement of cattle were not valid (35). A wealth of new knowledge about BTV and other orbiviruses, including the absence of adverse effects of BTV on pregnant cattle and the bovine foetus was presented at the Second International Symposium (35). An explanation for overwintering of BTV was not perceived to be a problem due to the constant renewal of the susceptible host/vector pools (22). It was observed that most amplification of BTV occurred as epizootics when infected vectors were transported by prevailing winds and by movement of infected vectors to areas in which susceptible hosts 34 Veterinaria Italiana, 40 (3), 2004

5 Global situation and competent arthropod vectors co-existed. The hypothesis was advanced that BTV serotypes evolved independently rather than by evolving from the original viruses first identified in Africa. Cattle are considered to be the reservoir hosts of BTV because the viraemia is prolonged and the majority of infections are subclinical (17). The hypothesis that persistent infections or a carrier state responsible for BTV persistence in an area was produced in cattle by foetal infection was not substantiated. While BTV may be shed in semen, it is transient in viraemic bulls and, therefore, vertical transmission is unlikely in the epidemiology of BT. In a recent study using a serotype 17 strain from the USA, it was shown that while viraemia in cattle can be prolonged, it was detectable by virus isolation techniques for no more than 49 days and by blood feeding by competent C. sonorensis for only 21 days (3). An analysis of the published data on >500 cattle infected with BTVs concluded that there was a >99% probability that detectable viraemia terminated in 63 days (28). While viral nucleic acids can be detected by reverse transcriptase polymerase chain reaction techniques for up to 222 days, this signal does not reflect viable virus, but fragments of virus remaining in the metabolically inactive erythrocytes. Using bulls and pregnant cattle infected naturally in field studies by Australian BTV serotypes, it was demonstrated that BTV was not excreted in the semen of infected bulls, even during viraemia, and there was no evidence of foetal infection in pregnant cows (18). Inoculation of calves in utero during the first trimester of gestation with North American BTV isolates produced brain lesions that were not compatible with life and, therefore, the calves could not contribute to the spread of BTV (33). Late gestational infections caused mild lesions and the production of protective antibody to eliminate the infecting virus. In an extensive study on BTV infection of pregnant cows using North American BTV isolates, all cows became infected, were viraemic, and seroconverted (25). It was concluded that BTV infection of pregnant cows did not produce transplacental infection of the bovine foetus and that induction of immunotolerance and latent or persistent infections did not occur in the calves. The BTV studies in the Caribbean Basin confirmed that: BTV occurs as a ubiquitous inapparent infection in young ruminants of tropical and subtropical countries BTV serotypes are isolated frequently from clinically normal ruminants there is no evidence to support extension of the Caribbean serotypes to northern Mexico, the USA and Canada or northern serotypes to the Caribbean Basin, despite the movement of livestock and insects (6, 41). Recognising the important role of insect vectors and sharing of insect habitats across borders and ecosystems, there is still a stable separation of BTV serotypes between the tropical and temperate ecosystems in the Americas. There are no unequivocal examples and there is no confirmed epidemiological evidence of the introduction to and establishment of BTV in a new area that can be attributed to international trade or movement of infected livestock and livestock products (6, 20, 39). It was noted that BTV distribution is consistent with insect habitat, not geopolitical boundaries, and that control of animal movements exceeds the usefulness of this control to prevent distribution of BTV serotypes (6, 21). Clearly, embryo transfer and the use of semen collected with appropriate mitigations pose no risk for transmission of BTV (38). Movement of seronegative animals or seropositive animals under appropriate mitigations does not pose a risk of introducing BTV into a BTV-free area and there is no evidence that it occurs (40). Persistence of BTV in cattle and the postulated carrier state have not been validated as a threat to importing countries (20). Therefore, it is logical and reasonable to recommend that the epidemiological, virological, vectorial and ecological realities of trading partners must be evaluated before making regulatory policies (36). The following recommendations for regulatory policy consideration were modified from those presented at the Second International Symposium, but they are as valid today as they were at that time (6): 1) A review of the assignment of BT to List A of the Terrestrial animal health code or a modification of the reporting requirements for diseases must be initiated. 2) Special consideration needs to be given in the Terrestrial animal health code to the status of arboviruses which are not restrained by international borders and for which animal movements play a minor or no role. 3) A review of the International animal health code and the Working Team recommendations from the Second International Symposium will facilitate bilateral or inter-regional discussions between Veterinaria Italiana, 40 (3),

6 Global situation trading partners and place concerns about BTVs and BTV vectors in the proper scientific perspective. 4) Continued international co-operation and dialogue between scientists and the regulatory community are needed to permit the study and understanding of arbovirus infections which span ecological regions to contribute to promulgation of responsible regulatory policies. It has come as a surprise to many of us in the scientific community that 12 years after the Second International Symposium appeared to dispel concerns about BTV carrier states in cattle, the international regulatory position and attitudes have not changed to reflect the reality of current science. There is an overwhelming lack of support for the hypothesis of BTV carrier cattle. It is hoped that another 18 years will not pass after this Third International Symposium before the five principles proposed during the First International Symposium by F.A. Murphy (19) and supported by the scientific contributions of the First and Second International Symposia (2, 35) are incorporated into international regulatory policy considerations. References 1. Anon. (2003). Bluetongue, Chapter In Terrestrial animal health code, 12th Ed., Office International des Épizooties, Paris. 2. Barber T.L., Jochim M.M. & Osburn B.I. (1985). Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, 746 pp. 3. Bonneau K.R., DeMaula C.D., Mullens B.A. & MacLachlan N.J. (2002). Duration of viraemia infectious to Culicoides sonorensis in bluetongue virusinfected cattle and sheep. Vet. Microbiol., 88, Bowen R.A., Howard T.H. & Pickett B.W. (1985). Seminal shedding of bluetongue virus in experimentally infected bulls. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, Braverman Y. (1992). The possible introduction to Israel of Culicoides (Diptera: Ceratopogonidae) borne animal disease by wind. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Calderon J.E & Homan E.J. (1992). Policy considerations for bluetongue in the Caribbean Basin. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Callis J. (1985). Bluetongue in the United States. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, Dulac G.C., Sterritt W.G., Dubuc C., Afshar A., Myers D.J., Taylor E.A., Jamieson B.R. & Martin M.W. (1992). Incursions of orbiviruses in Canada and their serologic monitoring in the native animal population between 1962 and In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Hardy W.T. & Price D.A. (1952). Sore muzzle of sheep. J. Am. Vet. Med. Assoc., 120, Holbrook F.R., Tabachnick W.J., Schmidtmann E.T., McKinnon C.N., Bobian R.J. & Grogan W.L. (2000). Sympatry in the Culicoides variipennis complex (Diptera: Ceratopogonidae): a taxonomic reassessment. J. Med. Entomol., 37, Howard T.H., Bowen R.A. & Pickett B.W. (1985). Isolation of bluetongue virus from bull semen. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, Howell P.G. & Verwoerd D.W. (1971). Bluetongue virus. In Virology monographs, Vol. 9 (S. Gard., C. Hallaver & K.F. Meyer, eds). Springer Verlag, New York, Jaggers R. & Anderson G. (1985). Bluetongue in western Nebraska: an area herd study. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, Kvasnicka W.G. (1985). Reproductive problems associated with bluetongue virus activity in Nebraska. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, Luedke A.J. (1982). Effect of bluetongue virus on reproduction in sheep and cattle. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, MacLachlan N.J. Osburn B.I., Stott J.L. & Ghalib H.W. (1985). Orbivirus infection of the 36 Veterinaria Italiana, 40 (3), 2004

7 Global situation bovine fetus. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, MacLachlan N.J., Barratt-Boyes S.M., Brewer A.W. & Stott J.L. (1992). Bluetongue virus infection of cattle. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Melville L.F. & Gard G.P. (1992). Investigations on the effects of natural infection with orbiviruses on reproduction in cattle. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Murphy F.A. (1985). Round-table discussion on current policies regulating the international movement of animals and germplasm. Chairman s comments. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, Osburn B.I. (1992). Summary of the Second International Symposium on Bluetongue, African horse sickness, and related orbiviruses. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Ozawa Y. (1992). Summary of the Second International Symposium on Bluetongue, African horse sickness and related orbiviruses. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Parsonson I.M. (1992). Overview of bluetongue virus infection of sheep. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Prescott L.M., Harley J.P. & Klein D.A. (1990). Microbiology. W.C. Brown, Dubuque, Iowa, Reeves W.C. (1985). Perspectives and recommendations for future research. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, , Roeder P.L., Taylor W.P., Roberts D.H., Wood L., Jeggo M.H., Gard G.P., Corteyn M. & Graham S. (1992). Failure to establish congenital bluetongue virus infection by injecting cows in early pregnancy. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Schultz R.D., Rhodes R., Panangala V.S. & Kaproth M. (1985). Unusual observations on a serologically negative bluetongue virus infected bull. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, Sellers R.F. (1992). Weather, Culicoides and the distribution and spread of bluetongue and African horse sickness viruses. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Singer R.S., MacLachlan N.J. & Carpenter T.E. (2001). Maximal predicted duration of viremia in bluetongue virus-infected cattle. J. Vet. Diagn. Invest., 13, Stott J.L., Blanchard-Channel M., Osburn B.I., Riemann H.P. & Obeso R.C. (1989). Serologic and virologic evidence of bluetongue virus in cattle and sheep in Mexico. Am. J. Vet. Res., 50, Sudia W.D. (1972). Arthropod vectors of epidemic Venezuelan equine encephalitis. In Venezuelan encephalitis. Proc. Workshop-Symposium on Venezuelan encephalitis virus, Washington, DC, September Pan American Health Organization, Washington, DC, Scientific Publication No. 243, Tabachnick W.J. (1996). Culicoides variipennis and bluetongue virus epidemiology in the United States. Ann. Rev. Entomol., 41, Thompson L.H., Mo C.L., Oviedo M.T., Homan E.J. & Interamerican Bluetongue Team (1992). Prevalence and incidence of bluetongue viruses in the Caribbean Basin: serologic and virologic findings. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Waldvogel A.S., Anderson G.A. & Osburn B.I. (1992). Strain-dependent variations in the pathogenesis of fetal infection with bluetongue virus serotype 11. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Walton T.E. (1992). Foreword. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, v-vi. 35. Walton T.E. & Osburn B.I. (1992). Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, pp. Veterinaria Italiana, 40 (3),

8 Global situation 36. Walton T.E., Tabachnick W.J., Thompson L.H. & Holbrook F.R. (1992). An entomologic and epidemiologic perspective for bluetongue regulatory changes for livestock movement from the USA and observations on bluetongue in the Caribbean Basin. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, WHO/FAO Working Team (1985). WHO/FAO Working team report: pathology. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, Working Team on Germplasm (1992). Working team report on germplasm. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Working Team on International Impacts (1992). Working team report on international impacts. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Working Team on Regulatory Issues (1992). Working team report on regulatory issues. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Working Team on Vectors (1992). Working team report on vectors: recommendations for research on Culicoides vector biology. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Veterinaria Italiana, 40 (3), 2004

9 Vet. Ital., 40 (3), Global situation A South African overview of the virus, vectors, surveillance and unique features of bluetongue G.H. Gerdes Agricultural Research Council, Onderstepoort Veterinary Institute, Private Bag X05, Onderstepoort, South Africa Summary The origin of bluetongue (BT) is probably African and the disease was first recognised in South Africa in Merino sheep in the late 18th Century. Diagnostic and research findings for a number of years have been summarised to obtain data relevant to the distribution of BT and its serotypes in the country. The role of ruminant game and cattle as maintenance hosts for BT virus (BTV) is mentioned although cattle appear to have largely replaced antelope in this role. Only about 30% of over game animals tested for export were found to be BT-antibody positive. An outbreak of a bluetongue-like disease in cattle is mentioned as are the BT and epizootic haemorrhagic disease of deer (EHD) isolates in the outbreak. A summary by serotype and province of sheep isolates is given and it is pointed out that the sheep population in a province does not reflect the number of isolates made and the province with the largest sheep population has almost the smallest number of BTV isolates and vice-versa. South Africa currently has 21 of the 24 BTV serotypes with 17, 20 and 21 being exotic to the country. The recent retrospective typing of serotype 17 in South Africa is being investigated, as type 17 crosses strongly with type 20, which is absent and also with type 4 which is present. 1, 3, 4 and 2 were the most common serotypes while 18, 19, 22 and 23 were not found among the isolates. Mention is made of BTV isolates made from Culicoides bolitinos catches during two devastating outbreaks of African horse sickness in an unvaccinated population. A six-year Culicoides monitoring project is mentioned and the many BTV isolates made of a variety of serotypes. BTV is endemic in Africa and in South Africa unvaccinated indigenous breeds appear to have achieved a balance with the virus. Indeed, it is possible to find virus, antibody and lesions in asymptomatic animals in different situations. Bluetongue creates a significant trade barrier but the virus remains interesting among a number of other uniquely African viruses. Keywords Bluetongue virus Culicoides Distribution Game Serotype Survey Wildlife. Introduction Bluetongue (BT) virus (BTV) is probably of African origin. It was first described in South Africa when the disease was encountered in Merino sheep introduced into the Cape Colony in the late 18th Century. Serotypes of BTV have since spread to other countries to the east, west and north of Africa. Bluetongue is described as a disease of domestic and wild ruminants and it is thought that cattle have now largely replaced antelope as a maintenance host of the virus. Figures reflecting BT antibodies in game exported between 1997 and 2000 range from a low 18.8% to a high 43.4% positive. Consignments varied from 1 up to 10 different species of ruminant game. Protocols required export only in vector free/low seasons and the certification by virus isolation that serologically positive animals were not viraemic at time of export. Obviously, animals were sourced from drier areas of the country where they were expected to be serologically negative. A more specific and structured survey was undertaken in 1997 and animals were sourced from Veterinaria Italiana, 40 (3),

10 Global situation areas of different vegetation types and rainfall. These areas varied from semi-desert to forest transition with rainfall of mm and >900 mm, respectively. BT antibodies were present in animals from all regions sampled and 10 out of 18 species tested positive (1). Comparable figures to illustrate the exposure of cattle to BTV do not exist and cattle are not routinely tested or vaccinated. Naive, imported animals do experience clinical disease although bovine isolates are generally treated with caution. However, in 1996, an exceptionally wet year, BT in cattle was reported from various parts of the country (2). Symptoms of stomatitis, coronitis, lacrimation, salivation, nose and teat sloughs and sometimes haemorrhagic diarrhoea, were recorded. Morbidity and mortality were low but a marked drop in milk production was noted. Thirteen BT and nine epizootic haemorrhagic disease (EHD) isolates were made. Serotypes 2, 3, 6 and 8 of bovine BTV were involved, while the EHD isolates remain to be typed. Sheep and cattle serotypes for that specific season did not correspond and additional serotypes were obtained from sheep (Fig. 1). Virus BTV is endemic in South Africa but only wet seasons with large outbreaks serve to raise levels of concern. It is a notifiable disease in terms of the Animal Diseases Act of 1984 but compliance with the act is erratic. Outbreaks reported vary from as few as 21 to almost 100 in the years from 1998 to Isolates Serotypes Ovine Bovine Figure 1 Bluetongue serotypes in sheep and cattle in South Africa, 1996 Twenty-one of the twenty-four serotypes of BTV occur in South Africa and currently types 17, 20 and 21 are exotic to the country. To illustrate distribution of BT in South African sheep, a summary of 20 years of domestic isolations is given below by province and by serotype (Fig. 2). The total sheep population of South Africa is approximately 28.5 million distributed across nine provinces (Fig. 3). Provincial isolation totals do not follow the sheep population by province (Fig. 4). The province with the smallest number of sheep had the second highest number of isolates while the province with the largest number of sheep had the second lowest number of isolates. This is obviously dependent on the breed of sheep and the type of husbandry practised. Isolates Serotypes Figure 2 Isolates of bluetongue virus serotypes in ovines, Veterinaria Italiana, 40 (3), 2004

11 Global situation BTV isolations from asymptomatic animals also occur. In the embryo export programme of sheep and goat donor females, animals were bled and tested at the time of embryo flushing. From 1997 to 2002 all donor females tested BTV-negative, except in 1999 when 12 isolates of 5 different serotypes were made from asymptomatic animals. Figure 3 Sheep population of South Africa, 2001 Serotype totals make the most interesting reading with the low denomination serotypes being isolated more frequently and four serotypes not having occurred in 20 years. The most commonly isolated serotypes were BTV-1, BTV-2, BTV-3 and BTV-4 while BTV-18, BTV-19, BTV-22 and BTV-23 were absent (Fig. 2). Serotype 17, thought to be exotic, was typed in 1985, 1986 and The last two isolates have been retyped and are currently being sequenced to confirm their identity as serotype 17 crosses strongly with both types 4 and 20. As the polymerase chain reaction (PCR) is now also an accepted trade test, selected samples in the donor animal testing programme were subjected to side side isolation and PCR testing. A total of 88 blood samples were tested and all proved to be negative in both tests. The high exposure months of February, March and April were chosen. Vectors BTV has been isolated from a number of Culicoides species although only two, namely C. imicola and C. bolitinos are recognised as being vector competent. In 2001 and 2003, C. bolitinos was investigated in the field during two devastating outbreaks of African horse sickness (AHS) in the Eastern Cape. Culicoides were trapped as part of the AHS investigation and in 2001, BTV-1 and BTV-24 were isolated and in 2003 types 6, 8 and 19 were identified in the C. bolitinos catches made near horse stables. The role of mosquitoes and other biting flies in the mechanical transmission of BTV should not be overlooked. Although not important, 301 mosquito catches were made at Onderstepoort during a sixyear Culicoides project and 21 catches were positive for BTV. These yielded serotypes 2, 3, 4, 5, 11, 12, 14 and 16. During the same period, 300 Culicoides catches were made, also at Onderstepoort, of which 180 yielded BTV of a variety of serotypes (Fig. 5). Surveillance Figure 4 Provincial bluetongue virus isolation totals, Serotypes are distributed randomly in any given area and season although nine are recognised to have a high epidemic potential, six others occur regularly and a further three only occur sporadically (4). The distribution of BTV serotypes in a population of both vaccinated and unvaccinated animals is often surprising and also uneven (Fig. 2). In a single flock of sick sheep it is possible for four out of five bloods to all yield type 9 or even six out of six bloods to yield two type 1s, two type 11s, a 10 and a 4. Since BTV is endemic in South Africa, no surveillance per se is carried out. Reference has been made to a six-year Culicoides survey conducted between 1978 and 1985 encompassing 25 sites in different provinces and various areas (3). It was accidental that BTV was the most frequently isolated virus, as the approach was multifocal and various isolation systems were used. Eight genera of viruses were isolated, together with a batch of unidentified isolates. Bluetongue made up 58.4% of the final total obtained (810 viruses). A second single provincial survey being conducted at present is the Rural Livestock Survey where livestock at diptanks are being targeted. The breeds are mostly Veterinaria Italiana, 40 (3),

12 Global situation Isolates Serotypes Figure 5 Culicoides isolates of bluetongue virus, indigenous or crosses thereof and they are an unvaccinated rural/communal type of population. Sheep and goats are being screened for BT and at present slightly more goats than sheep have been presented for bleeding. Unvaccinated bluetongue seropositive animals total 63.7% in a provincial area, which is low lying, wet and fairly subtropical. Recent figures are presented for seronegative Merino sheep being sourced from a cold, high-lying area where vaccination is not practised because BT is not recognised. These animals are required for projects requiring susceptible sheep. They were bled in Spring and autumn and varied from less than 1% to 84% antibody-positive sometimes much to the surprise of their owners. In conclusion, BT is an Office International des Épizooties (OIE) List A disease and remains a considerable trade barrier. An endemic situation exists in South Africa where the population is exposed to many different serotypes of the virus each season. Some sheep are regularly vaccinated and other breeds do not show clinical disease unless mouth lesions are an incidental finding at deworming. Vaccination is either done by the book with concern about temporary infertility in rams and early pregnancy in ewes with associated contraindications, or a cocktail of all three bottles of vaccine is administered as a single dose. Heavy rainfall will invariably bring about an explosive increase in vectors and localised outbreaks of BT and the virus always remains interesting among the other uniquely African viruses. Acknowledgements Thanks are extended to: the staff of the OIE BT Reference Laboratory for typing the isolates the staff of the Department of Entomology for supplying the Culicoides survey data the staff of the Department of Virology for assistance in the BTV isolations. References 1. Barnard B.J.H. (1997). Antibodies against some viruses of domestic animals in southern African wild animals. Onderstepoort J. Vet. Res., 64, Barnard B.J.H., Gerdes G.H. & Meiswinkel R. (1998). Some epidemiological and economic aspects of a bluetongue-like disease in cattle in South Africa. Onderstepoort J. Vet. Res., 65, Nevill E.M., Erasmus B.J. & Venter G.J. (1992). A six-year survey of viruses associated with Culicoides biting midges throughout South Africa (Diptera: Ceratopogonidae). In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Verwoerd D.W. & Erasmus B.J. (1994). Bluetongue. In Infectious diseases of livestock with special reference to southern Africa, Vol. I (J.A.W. Coetzer, G.R. Thomson & R.C. Tustin, eds). Oxford University Press, Cape Town, Veterinaria Italiana, 40 (3), 2004

13 Vet. Ital., 40 (3), Global situation North Africa: a regional overview of bluetongue virus, vectors, surveillance and unique features S. Hammami Institut de la Recherche Vétérinaire de Tunisie, Rue Djebel Lakhdhar, La Rabta 1006, Tunis, Tunisia Summary Bluetongue virus serotype 2 (BTV-2) appeared in North Africa in December 1999 and caused a total of clinical cases and deaths in sheep. This arthropod-borne viral disease was first reported by the Tunisian veterinary services in 1999 followed by the Algerian authorities in 2000 and has been described in adult sheep only. The overall morbidity and mortality rates were 9% and 3.5%, respectively. Following the initial incursion of BTV-2 in December 1999, Tunisia reported an epidemic in 2000 and a few outbreaks in Neither Morocco nor Libya has reported any clinical cases of bluetongue in recent years, despite surveillance programmes being carried out. In Tunisia, the control strategy was based on mass vaccination of sheep using a live-attenuated monovalent type 2 vaccine, while in Algeria, it was based on vector control. Vector surveillance has proven the presence of Culicoides imicola in Morocco but there are no data available for either Tunisia or Algeria. Keywords Algeria Bluetongue Epidemiology Morocco North Africa Surveillance Tunisia. North Africa (Algeria, Libya, Morocco and Tunisia) with their livestock populations of approximately 60 million small ruminants and 5 million cattle, are threatened by numerous transboundary diseases. This is due in part to integrated global trade and cross-boundary movement of animals. In addition, climatic changes observed during recent years have affected the world distribution of several animal diseases including bluetongue (BT). First described in South Africa in 1870, BT has since been reported over a broad band of countries on all the continents between 40 N and 35 S. Changes in climatic conditions may have contributed to the newly emerging distribution patterns of the disease, extending its range in the Mediterranean region to countries that have not previously reported BT. The disease was diagnosed in the Greek Islands towards the end of 1998 (10). In 1999, cases of BT were reported in countries around the Mediterranean Basin, such as Bulgaria, Tunisia and Turkey (10). The following year, BT re-emerged in Tunisia and epidemics were reported in Algeria, Italy, Spain (Balearic islands), Greece and France (Corsica). In 2001, Italy confirmed the presence of bluetongue virus (BTV) serotypes 2 and 9, the latter also being reported in Bulgaria. New cases were also reported in Greece and Corsica. In 2002, new clinical cases were reported in Tunisia (BTV-2) and Bulgaria (BTV-9) (11). In 2003, the disease re-emerged in Minorca and Corsica and Sardinia (BTV-4). History of the incursion of bluetongue in North Africa and unique features of outbreaks 1956 In 1956, a limited incursion of BT was reported in the southern area of Larache and west of Arbua in Morocco (9). The rapid implementation of a vaccination programme using a polyvalent vaccine combined with unfavourable climatic conditions for the vectors halted the spread of the disease. Typical clinical signs were observed in some animals during this incursion. Only adult sheep were clinically affected with an incidence of 20% (9). No clinical cases have been reported in recent years in Morocco In Tunisia, the autumn of 1999 was characterised by high temperatures and heavy rain. This weather Veterinaria Italiana, 40 (3),

14 Global situation created favourable conditions for BTV vector activity and subsequently the occurrence, for the first time, of BT in the country. During this first incursion, severe clinical signs were observed in affected sheep: high temperature (41-42 C), nasal discharge, salivation, oedema and congestion of the head and the mucous membranes. Only adult sheep were affected; no cases were recorded in cattle or young lambs. Affected sheep flocks were located in the eastern part of Tunisia along the coast. The overall morbidity and mortality rates were 8.35% and 5.5%, respectively. Of the 837 reported clinical cases, 325 died. The serotype of the virus isolated in Tunisia was determined as BTV-2 by the Institute for Animal Health (IAH), Pirbright. During the winter of 2000, a serological survey, using competitive ELISA, was conducted one month after detection of the last clinical case on one farm, comprising 886 sheep, 812 lambs and 400 bovines. Seropositive BTV reactions were recorded in 63.4% of adult bovines, 40.6% of heifers, 27.09% of adult sheep and just 2.1% of young lambs. Overall, the average level of positivity of animals on the farm was 22.6% (7) Tunisia In 2000, 72 outbreaks of BT were reported during the period extending from June to October. A total of clinical cases were diagnosed in sheep, of which died. Outbreaks were reported in 10 districts with most cases diagnosed in the eastern and central parts of the country. Circulation of BTV serotype 2 was confirmed by the IAH in Pirbright (6). Molecular studies comparing genomic segments 2 and 7 of the virus isolated in Tunisia to those of the isolate in Corsica showed no significant difference between them (segment 2: 99.4% homology; segment 7: 100% homology). Therefore, the two isolates were probably of the same origin; the transport and amplification in the vector had no effect on the virus genome sequence. No significant difference was observed when comparing segments of the virus isolated in 1999 and in 2000 in Tunisia (1). Further sequencing studies are needed to better characterise these isolates. Algeria For the first time in its history, Algeria reported 28 outbreaks of BT between July and September 2000 in the north-east of the country. The disease spread after the first cases to affect 24 localities in the district of Jijel. Of susceptible sheep, were clinically affected. The disease continued to spread and by the end of the epidemic, six districts in the eastern and central parts of the country were affected (Skikda: cases; Souk Ahras: 430 cases; Annaba: 500 cases; Guelma: cases; Oum El Bouaghi: 5 cases; Tebessa: 35 cases; and Jijel: 18 cases) (8) In 2002, the Tunisian national veterinary authorities reported a very limited number of new outbreaks of BT. A total of 4 outbreaks with 21 clinical cases and 6 deaths were recorded in the central part of the country in non-vaccinated sheep flocks, indicating that BTV serotype 2 was probably still circulating. The Algerian national veterinary authorities did not report any new outbreaks in The virus was probably not circulating in the other countries in the region (Libya and Morocco). Figure 1 summarises the epidemiological situation in North African countries between 1999 and Figure 1 Distribution of recent bluetongue epidemics in North Africa, Control measures and surveillance programmes Control measures Once BT had been confirmed in Tunisia, the national veterinary authorities implemented a series of control measures. On premises where outbreaks were recorded, flocks were isolated and dead animals buried. Sick animals, animal holdings and surrounding areas were sprayed with insecticide. Surveillance for the detection of new clinical cases in the nearby flocks was initiated and a monovalent type 2 vaccine was ordered from Onderstepoort in South Africa. In 2000, the only vector control measures used were the application of insecticides on the affected premises and spraying of these insecticides over the affected region. By the end of 2000, a campaign of mass vaccination of sheep was conducted with a total of sheep vaccinated. 44 Veterinaria Italiana, 40 (3), 2004

15 Global situation The subsequent annual vaccination campaigns involved and sheep in 2001 and 2002, respectively. A study conducted in the field on six flocks of different local breeds (3 controls and 3 vaccinated flocks of about 300 animals) to evaluate the innocuity and the efficacy of the vaccine, showed that side-effects were not detected and that animals did seroconvert after vaccination. The control strategy adopted by the Algerian national veterinary authorities during the first epizootic was based on vector control and included spraying insecticides over the affected regions. In addition, a surveillance programme aimed at the detection of clinical cases was conducted but at no time was vaccine used (8). In the absence of any clinical cases being either diagnosed or confirmed in Morocco, only serological monitoring was undertaken. Vector surveillance Since the incursion of BT into North Africa, the veterinary authorities of the region have implemented surveillance programmes to detect new clinical cases, circulation of BTV by seroconversion and the presence and the distribution of known vectors of the disease. Tunisia Following the epizootic of African horse sickness (AHS) in 1966, studies have been conducted to detect different species of Culicoides in Tunisia (4). A total of 17 species have been identified with C. circumscriptus, C. coluzzii, C. longipennis and C. puncticollis found most frequently. The distribution of C. circumscriptus was dependent on the degree of salinity of the collection site. In Tunisia, this halophytic species is the most frequently found under different climates. C. puncticollis was retrieved from most samples collected in the northern part of the country and semi-arid zones and it has also been found in the Saharan ecozone where it is commonly associated with C. circumscriptus. These two Culicoides species were adapted to heavily polluted sites. C. coluzzii is the most frequently found in the semiarid, arid and Saharan ecozones. It is often associated with C. circumscriptus even in sites very rich in sodium chloride (3). Studies are being conducted to determine BT vector distribution and to try to detect C. imicola. Morocco Following the epizootic of vector-borne diseases (BT in 1956 and AHS in 1966), extensive field studies were conducted that identified 49 species of Culicoides. Further investigations were conducted following the two epizootics of AHS in and in (2). The presence of the following three Culicoides vector species was confirmed: C. imicola, C. obsoletus and C. pulicaris. The former is the most frequently trapped and was found in all sites covered by the investigation in altitudes varying between 4 m and m. However, the abundance of this vector depends on geo-climatic parameters, thus Marrakech, Arbua and Larache, located in the north-west of the country, were the regions where C. imicola was found all year around. The two other species trapped during the investigation were C. obsoletus and C. pulicaris and have very similar spatial and seasonal distributions. Since 2000, the national veterinary authorities of Morocco have implemented an epidemiovigilance programme. Entomological studies have been conducted for two years aimed at the detection of any new competent Culicoides vectors and to determine their spatial and temporal distributions with C. imicola identified in three of the five sites studied. Serological investigations in the susceptible population and in sentinel herds have been also conducted (5). Conclusions The distribution of BT worldwide is changing. The incursion of BTV serotype 2 into North Africa has been confirmed. The way that BTV entered and spread within the North African region remains unclear although it is likely that the virus will probably continue to circulate for a few years if favourable conditions are maintained. Surveillance programmes have to be implemented to detect circulation of this serotype, incursion of new serotypes and the distribution of competent vectors. These programmes will surely be more efficient with reinforced diagnostic capabilities of national laboratories and collaboration at regional and international levels. Acknowledgement The author would like to thank Chris Hamblin from the Institute for Animal Health, Pirbright, United Kingdom for reading the manuscript and making valuable recommendations. References 1. Ben Fredj S., Bréard E., Sailleau C., Zientara S. & Hammami S. (2003). Incursion de la fièvre catarrhale ovine en Tunisie : caractérisation moléculaire des isolats viraux. Rev. Élev. Méd. Vét. Pays Trop., 56 (3-4), Bouayoune H., Touti J., El Hasnaoui H., Baylis M. & Mellor P.S. (1998). The Culicoides vectors of African horse sickness virus in Morocco: distribution and Veterinaria Italiana, 40 (3),

16 Global situation epidemiological implications. Arch. Virol., 14, Chaker E., Delécolle J.-C. & Kremer M. (1980). Variabilité des caractères morphologiques de Culicoides circumscriptus Kieffer, Mise en synonymie de C. kirovabadicus Dzhafarov, 1964 [original in French] (Morphological variability of Culicoides circumscriptus Kieffer, 1918 with C. kirovabadicus Dzhafarov, 1964 as a probable synonym). Arch. Inst. Pasteur Tunis., 57 (1-2), Chaker E. & Kremer M. (1982). Les Culicoides de Tunisie : particularités morphologiques. Chorologie et écologie des espèces retrouvées [original in French] (Culicoides of Tunisia: morphological characteristics. Chorology and ecology of species found). Arch. Inst. Pasteur Tunis, 59 (4), El Harrak Y. (2001). La fièvre catarrhale du mouton (bluetongue). Données épidémiologiques et système d épidémiosurveillance lancé au Maroc. BEV Maroc, 6, Hammami S., Ben Said M.S. & Bahri S. (2000). La fièvre catarrhale du mouton : éléments de diagnostic et moyens de lutte. BEIV, 17, Hammami S., Seghaier C., Hamblin C., Ben Said M.S. & Bahri S. (2001). Particularités de la fièvre catarrhale du mouton en Tunisie : étude d un foyer. BEIV, 18, Hamida B. (2000). Les pays touchés par la fièvre catarrhale. 67th General Session of the International Committee of the Office International des Épizooties (OIE). OIE, Paris. 9. Placidi L. (1957). La bluetongue au Maroc. Vigot Frères, Eds. Bull. Acad. Vét., XXX (février 1957), Zientara S., de La Rocque S., Gourreau J.M., Grégory M., Diallo A., Hendrikx P., Libeau G., Sailleau C. & Delécolle J.-C. (2000). La fièvre catarrhale ovine en Corse Epidémiol. Santé Anim., 38, Zientara S., Grillet C., de La Rocque S., Gourreau J.M., Grégory M., Hendrikx P., Libeau G., Sailleau C., Albina E., Bréard E. & Delécolle J.-C. (2001). La fièvre catarrhale ovine en Corse en Epidémiol. Santé Anim., 40, Veterinaria Italiana, 40 (3), 2004

17 Vet. Ital., 40 (3), Global situation Bluetongue viruses, vectors and surveillance in Australia the current situation and unique features P.D. Kirkland Virology Laboratory, Elizabeth Macarthur Agricultural Institute, PMB 8, Camden, NSW 2570, Australia Summary While there are dramatic differences between recent bluetongue (BT) developments in Europe and the situation in Australia, there are also a number of similarities. About 25 years ago, as a BT-free country, Australia was advised that a BT virus (BTV) had been identified, though there was no evidence of disease. During the following 15 years, 8 BTV serotypes were identified. Despite the presence of some virulent viruses, Australia remains free of BT disease. Nevertheless, the economic impact is considerable due to disruption to trade. In the last decade, research efforts have focussed on reducing the impact of BTVs on the export of livestock, semen and embryos. In 1993, the National Arbovirus Monitoring Program (NAMP) was established as a co-operative initiative between the livestock industries, national and state governments. The main emphasis of NAMP has been to define the distribution of BTVs and their vectors, together with monitoring annual fluctuations of viruses and vectors. A combination of climatic, geographical, virus and vector monitoring data that have been gathered over more than 25 years, have allowed the accurate delineation of BTV-free zones and zones of possible BTV transmission in accordance with OIE guidelines. These zones are now promoted to trading partners to facilitate trade. Keywords Australia Bluetongue-free zones Bluetongue viruses Culicoides Epidemiology Surveillance National Arbovirus Monitoring Program. While there are dramatic contrasts between the bluetongue (BT) situation in Europe and Australia, there are also similarities. Some European countries, previously free of BT, have recently experienced either significant outbreaks of disease in sheep, or the discovery of the presence of BT viruses (BTVs). About 25 years ago, Australia faced a similar situation. In 1977, Australian animal health authorities were advised that a virus that had been isolated two years previously from insects had just been identified as a BTV (25). This was the first recognition of BTV serotype 20. As a country with a population of more than 140 million sheep, Australia had exotic disease plans to respond to an incursion of a BTV. Consequently, the livestock industries struggled with this news when there was no evidence of disease but the presence of a virus that halted exports of animals and, in some cases, even animal products. Over the next decade, the distribution of BTVs was investigated. Initially, serogroup reactive tests were employed to determine the maximum possible distribution of BTVs. Positive samples were then tested in serotype-specific virus neutralisation (VN) tests. Results of VN tests suggested that the original BTV isolate of serotype 20 was not widely distributed but certain other serotypes, especially serotype 1, may be present (11). In some instances, type-specific results for sera that had given strong serogroup test reactivity were negative. This suggested that there may be additional, as yet unrecognised, serotypes present. Virus isolation studies on blood samples collected from sentinel cattle yielded a number of isolates of BTV-1 and also another new serotype, BTV-21 (27). Epidemiological studies conducted in several states defined the limits of BTVs in Australia (9, 13, 20, 28). Intensive investigations continued at the Coastal Plains site in far Northern Australia, where the first BTV had Veterinaria Italiana, 40 (3),

18 Global situation been identified. Over a period of four years, another five serotypes of BTV, namely BTV-3, BTV-9, BTV- 15, BTV-16 and a further new serotype, BTV-23, were isolated (14, 15, 16). No new serotype of BTV has been isolated in Australia since Experimental studies of the virulence of BTVs in sheep have shown that there is a spectrum, ranging from apparently non-pathogenic to moderately virulent (12, 17, 18, 19, 26, 29). Among the pathogenic isolates, Australian viruses appear to be considerably less virulent than South African strains (17). Under natural conditions, disease has not been recorded in commercial sheep flocks. This is due to a combination of factors sheep are rarely raised in areas where BTVs are transmitted, the incidence of infection of sheep in these areas is low and the viruses that are present in regions in proximity to major sheep populations are considered to be nonvirulent. Two disease incidents have been observed when small groups of sheep have been moved into the tropical north of the Northern Territory. Clinical signs were observed in six animals from the two incidents. Disease associated with BTV infection has never been observed in commercial sheep flocks, cattle or goats in Australia. Entomological research has shown that there are five species of Culicoides in Australia that are potentially capable of transmitting BTVs, namely C. actoni, C. brevitarsis, C. fulvus, C. brevipalpis and C. wadai (1, 23, 24, 30, 31, 32); C. brevitarsis is considered to be the major BTV vector in Australia and is the most widely distributed. The other midge species have a less expansive distribution and are all confined to areas that lie within the range of C. brevitarsis. Data obtained from intensive studies of the ecology of C. brevitarsis has allowed a detailed understanding of the factors that affect the survival, multiplication and dispersal of this principal BTV vector (2, 3, 4, 5, 6, 7). These data have been utilised to develop models that predict the rates of movement and survival of C. brevitarsis (3, 8). In the last decade, surveillance and research efforts have focussed on reducing the impact of BTVs on the export of livestock, semen and embryos. In 1993, a co-ordinated Australia-wide surveillance and monitoring programme, the National Arbovirus Monitoring Program (NAMP), was formally established as a co-operative initiative between the livestock industries, national and state governments. Funding of the programme is shared between these groups. This programme has operated continuously in Australia since its establishment in The main emphasis of NAMP has been on the definition of the distribution of BTVs and their vectors, together with monitoring annual fluctuations of viruses and vectors. These objectives have been largely achieved by the sampling of sentinel cattle at key locations around Australia and by light-trap collection of insects at these sentinel sites. Groups of young cattle are sampled at sentinel herd locations. Animals are preferably less than 12 months old and free of maternal antibodies at the time of first sampling. In areas where BTV transmission may occur, animals are sampled at monthly intervals from summer through to the start of winter. At the tropical Coastal Plains site in the Northern Territory where the first isolate of BTV was obtained, animals are sampled at weekly intervals for 5-7 months each year. In other areas, sampling is less frequent, depending on the likelihood of BTV infection occurring. In proven free regions, animals may only be sampled at the commencement of summer and at the end of the vector season. Periodically, structured serological surveys are used to complement the sampling of sentinel animals. Data obtained from these surveys is used to refine the location of sentinel sites. Serum samples from sentinel cattle are first tested for serogroup reactive antibodies by competitive ELISA (c-elisa). Positive sera are tested for serotype-specific antibodies using the VN test. Blood cells are retained for future virus isolation attempts if an animal seroconverts. BTVs that are isolated are subjected to genetic analyses (topotyping) to identify the geographical origin of the viruses. While there have been no new serotypes of BTV found in Australia since 1986, the results of genetic analyses would suggest that there is an intermittent introduction of new strains of BTV into Australia from time to time (21, 22). Results of the NAMP have shown that only two serotypes (1 and 21) are widely distributed through northern Australia and along the eastern coast as far south as central New South Wales. The other six serotypes, including the viruses known to be virulent, remain confined to a small area in the far north of the Northern Territory and northern part of Western Australia. A network of insect collection sites is spread throughout the areas of possible BTV transmission and in the adjoining BTV-free areas. Biting midges (Culicoides species) are collected in light traps that are operated for varying periods depending on the local climate and the likelihood of vector activity. Insect collections are sorted to species and quantified. Data collected from sentinel cattle monitoring and insect collections are submitted to a central database through an internet-based data management system called NAMP Info (10). This database holds details of all monitoring sites as well as summaries of all virology and entomology results. The data are used to generate annual maps of BTV distribution 48 Veterinaria Italiana, 40 (3), 2004

19 Global situation throughout Australia and subsequently maps of BTV-free zones, zones of possible BTV transmission, and surveillance zones in accordance with OIE guidelines. There is unrestricted access to the current BTV zone map through the NAMP website ( These maps are available in a range of downloadable electronic formats as well as an online interactive format. The NAMP is reviewed at an annual meeting of representatives from the cattle, sheep and livestock export industries, national government and state governments through key scientists (entomologists, epidemiologists and virologists). At the annual review meeting, monitoring results are reviewed and the location of monitoring sites and frequency of monitoring are adjusted to take into account any variations of virus and vector distribution during the previous year. In addition to supporting the activities of the NAMP, Australian scientists continue to be engaged on a range of BT-related research projects. Current projects are directed towards detailed studies of vector ecology, modelling of virus and vector distribution, vector reduction strategies, improved vector collection techniques and refinements to diagnostic procedures. In order to minimise the disruption to trade and to manage the risks associated with the presence of BTVs, Australia has established a BTV-free zone and a zone of possible transmission. When the entomology and virology data are collated and integrated with national geographical data, there is a strong foundation for the definition of the limits of BTVs in Australia. Australia has a number of unique features that contrast with other countries and continents. It is a large, geologically stable and ancient landmass. The landform is relatively uniform, flat and continuous. Consequently, there is a continuous distribution of vectors and viruses across northern and eastern Australia, progressing towards a southerly limit imposed by the climate. There are no major mountains to disrupt the north-south movement of viruses and vectors. Similarly there is a continuous distribution of livestock, with populations of cattle in the northern and eastern regions of higher rainfall and temperature and very large flocks of sheep in the cooler drier areas. Collectively, these factors, when combined with virus and vector monitoring data that have been gathered for over more than 25 years, have allowed the delineation of BTV-free zones and zones of possible BTV transmission in accordance with OIE guidelines. These zones are now being promoted to trading partners to assist the export of live ruminants. A number of papers and posters presented during this symposium will provide more detailed information on aspects of recent BTV research and management in Australia. References 1. Bellis G.A., Gibson D.S., Polkinghorne I.G., Johnson S.J. & Flanagan M. (1994). Infection of Culicoides brevitarsis and C. wadai (Diptera: Ceratopogonidae) with four Australian serotypes of bluetongue virus. J. Med. Entomol., 31 (3), Bishop A.L. & McKenzie H.J. (1994). Overwintering of Culicoides spp. (Diptera: Ceratopogonidae) in the Hunter Valley, New South Wales. J. Aust. Entomol. Soc., 33, Bishop A.L. Barchia I.M. & Harris A.M. (1995). Last occurrence and survival during winter of the arbovirus vector Culicoides brevitarsis at the southern limits of its distribution. Aust. Vet. J., 72, Bishop A.L., Kirkland P.D., McKenzie H.J., Spohr L.J., Barchia I.M. & Muller M.J. (1995). Distribution and seasonal movement of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) at the southern limits of its distribution in New South Wales and their correlation with arboviruses affecting livestock. J. Aust. Entomol. Soc., 34, Bishop A.L., Kirkland P.D., McKenzie H.J. & Barchia I.M. (1996). The dispersal of Culicoides brevitarsis in eastern New South Wales and associations with arbovirus infections in cattle. Aust. Vet. J., 73, Bishop A.L., McKenzie H.J., Barchia I.M. & Harris A.M. (1996). Effect of temperature regimes on the development, survival and emergence of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in bovine dung. Aust. J. Entomol., 35, Bishop A.L., McKenzie H.J., Barchia I.M. & Harris A.M. (1998). Occurrence and effect of temperature regimes on four species of fly (Diptera) found with Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in bovine dung. Gen. Appl. Entomol., 28, Bishop A. L, Barchia I.M. & Spohr L.J. (2000). Models for the dispersal in Australia of the arbovirus vector Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae). Prev. Vet. Med., 47, Burton R.W. & Littlejohns I.R. (1988). The occurrence of antibody to bluetongue virus in New South Wales. I: Statewide surveys of cattle and sheep. Aust. J. Biol. Sci., 41, Cameron A.R. (2000). Development of an internetbased information system for monitoring veterinary arboviruses and their vectors. In Proc IXth Symposium of the International Society of Veterinary Epidemiology Economics (ISVEE), Breckenridge, August. 11. Della-Porta A.J., Sellers R.F., Herniman K.A.J., Littlejohns I.R., Cybinski D.H., St George T.D., McPhee D.A., Snowdon W.A., Campbell J., Veterinaria Italiana, 40 (3),

20 Global situation Cargill C., Corbould A., Chung Y.S. & Smith V.W. (1983). Serological studies of Australian and Papua New Guinean cattle and Australian sheep for the presence of antibodies against bluetongue group viruses. Vet. Microbiol., 8, Forman A.J., Hooper P.T. & Le Blanc Smith P.M. (1989). Pathogenicity for sheep of recent Australian bluetongue virus isolates. Aust. Vet. J., 66, Gard G.P., Shorthose J.E., Cybinski D.H. & St George T.D. (1985). Epidemiology of orbiviruses in the Northern Territory of Australia. In Veterinary viral diseases: their significance in South-East Asia and the Western Pacific (A.J. Della- Porta, ed.). Academic Press, Sydney, Gard G.P., Shorthose J.E., Cybinski D.H. & Zakrzewski H. (1985). The isolation from cattle of 2 bluetongue viruses new to Australia. Aust. Vet. J., 62, Gard G.P., Shorthose J.E., Weir R.P. & Erasmus B.J. (1987). The isolation of a bluetongue serotype new to Australia. Aust. Vet. J., 64, Gard G.P., Weir R.P., Melville L.F. & Lunt R.A. (1988). The isolation of bluetongue types 3 and 16 from northern Australia. Aust. Vet. J., 64, Hooper P.T., Lunt R.A. & Stanislawek W.L. (1996). A trial comparing the virulence of some South African and Australian bluetongue viruses. Aust. Vet. J., 73 (1), Johnson S.J., Hoffman D., Flanagan M., Polkinghorne I.G. & Bellis G.A. (1990). Pathology, signs and symptoms of bluetongue in Australia. Aust. Adv. Vet. Sci., 22, Johnson S.J., Hoffman D., Flanagan M., Polkinghorne I.G. & Bellis G.A. (1992). Clinicopathology of Australian bluetongue viruses for sheep In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Littlejohns I.R. & Burton R.W. (1988). The occurrence of antibody to bluetongue virus in New South Wales. II: Coastal region and age distribution surveys. Aust. J. Biol. Sci., 41, McColl K.A., Gould A.R., Pritchard L.I., Melville L. & Bellis G. (1994). Phylogenetic characterisation of bluetongue viruses from naturally-infected insects, cattle and sheep in Australia. Aust. Vet. J., 71, Melville L.F., Pritchard L.I., Hunt N.T., Daniels P.W. & Eaton B. (1997). Genotypic evidence of incursions of new strains of bluetongue viruses in the Northern Territory. In Proc. 7th Symposium on arbovirus research (B.H. Kay, ed). QIMR, Brisbane, Muller M.J. (1985). Experimental infection of Culicoides brevitarsis from south-east Queensland with three serotypes of bluetongue virus. Aust. J. Biol. Sci., 38, Muller M.J. (1987). Transmission and in vitro excretion of bluetongue virus serotype 1 by inoculated Culicoides brevitarsis (Diptera: Ceratopogonidae). J. Med. Entomol., 24, St George T.D., Standfast H.A., Cybinski D.H., Dyce A.L., Muller M.J., Doherty R.L., Carley J.G., Filippich C. & Frazier C.L. (1978). The isolation of a bluetongue virus from Culicoides collected in the Northern Territory of Australia. Aust. Vet. J., 54, St George T.D. & McCaughan C.J. (1979). The transmission of the CSIRO 19 strain of bluetongue virus type 20 to sheep and cattle. Aust. Vet. J., 55, St George T.D., Cybinski D.H., Della-Porta A.J., McPhee D.A., Wark M.C. & Bainbridge M.H. (1980). The isolation of two bluetongue viruses from healthy cattle in Australia. Aust. Vet. J., 56, St George T.D., Cybinski D.H. & Standfast H.A. (1982). The continued search for bluetongue related viruses in Australia. In Arbovirus research in Australia (T.D. St George & B.H. Kay, eds). CSIRO/QIMR, Brisbane, Squire K.R.E., Uren M.F. & St George T.D. (1981). The transmission of two new Australian serotypes of bluetongue virus to sheep. Aust. Vet. J., 57, Standfast H.A., St George T.D., Cybinski D.H., Dyce A.L. & McCaughan C.J. (1978). Experimental infection of Culicoides with a bluetongue virus isolated in Australia. Aust. Vet. J., 54, Standfast H.A., Dyce A.L. & Muller M.J. (1985). Vectors of bluetongue virus in Australia. Progr. Clin. Biol. Res., 178, Standfast H.A., Muller M.J. & Dyce A.L. (1992). An overview of bluetongue virus vector biology and ecology in the Oriental and Australasian Regions of the Western Pacific. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June 1991.CRC Press, Boca Raton, Veterinaria Italiana, 40 (3), 2004

21 Vet. Ital., 40 (3), Global situation Studies on bluetongue disease in the People s Republic of China N. Zhang, Z. Li, F. Zhang & J. Zhu Tropical and Subtropical Animal Virology Laboratory, Ministry of Agriculture, Jindian, Kunming , People s Republic of China Summary Bluetongue (BT) is an important infectious, non-contagious, insect-borne viral disease of ruminants, and is classified as a List A disease in the OIE Terrestrial animal health code. Since the first discovery and diagnosis of this disease in the Shizong County of Yunnan Province in 1979, the authors have developed systematic studies of the epidemiology, experimental epidemiology, aetiology, pathology, viral molecular characteristics (nucleic acid), diagnostic techniques, virus identification (grouping and typing) methods, vaccines and immunisation methods of BT in the People s Republic of China. Viruses have been isolated from seven provinces and the major serotypes have been identified. Epidemiological surveys have been conducted in sixteen provinces and a distinct type of BT virus (BTV) epidemiology has been found in China. Sentinel herds were established to monitor BTV regularly from 1995 to Experimental epidemiological studies have revealed the regional distribution of BTV serotypes and the dynamic spread of BTV in different habitats of distinct natural conditions, ecological environment and climate in China. A technical system of diagnosis of BTV infection also has been developed, including the agar gel immunodiffusion (AGID), indirect ELISA, competitive-elisa (c-elisa) and virus neutralisation (VN) tests for detection of antibodies to BTV, and the serum neutralisation (SN), immunofluorescence (IF) and immunoperoxidase (IP), antigen capture-elisa (ac-elisa), virus inoculation methods for virus detection, and nucleic acid electrophoresis analysis, nucleic acid probe techniques, and polymerase chain reaction (PCR) methods for detection of BTV nucleic acid and/or proteins. Attenuated and killed vaccines of BTV serotypes 1 and 16 have been developed and new immunisation methods have been developed. The S7, S10 and portions of the L2 gene segment of Chinese prototype strains of BTV serotypes 1, 2, 3, 4, 12, 15 and 16 have been sequenced and compared to the same strains of prototype and field strains of BTV from the United States of America and Australia. Phylogenetic analyses indicate that the genetic relationships between these viruses correspond to their geographic origins or their serotype. The National Standards for Bluetongue Diagnosis, People s Republic of China has also been enacted. Remarkable social and economic benefits have been obtained through the application and dissemination of these achievements in China. This paper reviews the progress of studies and control of BT in China since Keywords Bluetongue virus Diagnosis Epidemiology People s Republic of China Vaccines. Epidemiology of bluetongue in China Epidemiological survey Bluetongue (BT) normally occurs only in sheep, and fine-wooled breeds of sheep (like the Merino) are more sensitive. Cattle, and possibly buffalo, are the main vertebrate reservoirs of the virus (2). BT was first recognised in China in 1979 in Yunnan Province, and both the disease and antibody have since been recorded in other provinces in China (7, 12, 22, 23). A total of animals of nine different species from 29 provinces were surveyed to investigate the epidemiology of BT in China. Four species of ruminants (goat, sheep, buffalo, cattle) were found to be susceptible (20), and the presence of four vector species of Culicoides, including C. actoni, was confirmed (10). Camels were not affected. The Veterinaria Italiana, 40 (3),

22 Global situation epidemic period of BT coincides with the peak activity of Culicoides (from June to October), and there is an obvious seasonal cycle. Epidemics of clinical BT in sheep occurred in the region south of 35 N to 37 N, whereas only seroconversion occurred in ruminants in the regions north of 35 N to 37 N, such as Inner Mongolia where goats showed higher BT virus (BTV) group antibody prevalences than cattle, in contrast to the findings in most other regions. Another interesting result in Inner Mongolia is that some BTV infections continue in the winter months, suggesting an additional vector system; this finding contrasts with those in subtropical regions and other temperate areas of the world (22, 23). Epidemiological survey of serotype-specific antibodies For the serological investigation of BT in China, a total of serum samples were collected from goats, sheep, buffalo and cattle in seven provinces (Yunnan, Shanxu, Sichuan, Hubei, Inner Mongolia, Xinjiang and Tibet) from 1994 to A total of serum samples that were positive by AGID or c-elisa were tested in micro-sn tests using international reference strains of BTV serotypes 1, 2, 3, 4, 5, 9, 15, 16, 17, 20, 21 and 23, the serotypes that are known to be present in Asia and the Pacific. Antibodies to seven serotypes, namely 1, 2, 4, 9, 15, 16 and 23 were found. In China, the dominant antibodies are to serotypes 1, 2, 4 and 16. The results further suggested that BTV-4 might pose the greatest threat in China. Antibodies were also detected against two or three serotypes of BTV in single serum samples, indicating that cross-reactions occur amongst the serotypes of BTV that are present in China. The results were similar to the results of virus isolation studies, and the survey for BTV antibodies will assist in the epidemiological and pathological investigations of BT in China (9, 21). Epidemiological study of bluetongue virus in China Six sentinel herds of cattle, buffalo or goats have been established in different regions of Yunnan Province that have different climates, natural ecosystems and environment, to study the epidemiology of BTV infection (11). To determine the geographical distribution of BTV and to ascertain the range of serotypes that may be present in China, serum and whole (unclotted) blood samples were regularly collected from sentinel animals to monitor virus activity and to isolate the virus. A total of 110 isolates of BTV were obtained from sentinel cattle, and these were identified by virus and serum neutralisation tests (VN and SN). Serotypes isolated included 1, 2, 3, 4, 12, 15 and 16. Except for BTV serotypes 1 and 16 that have been isolated from sheep with clinical disease or naturally infected goats, the other serotypes (2, 3, 4, 12 and 15) had not been isolated previously in China. BTV serotypes 3, 4, 5 were obtained in Shiaong (which had outbreaks of BT in 1979) and BTV serotypes 1, 3, 4, 15 and 16 in Eshan in the subtropical area of Yunnan Province. The highest level of transmission took place in July and August, which coincided with BT outbreaks in Yunnan. BTV serotypes 1, 2, 12 and 16 were obtained in Xishuangbannai (Jinhong) and BTV serotypes 4, 12 and 15 in Dehong in the tropical region of Yunnan Province, although the date of the peak of transmission is not clear (8, 26). The presence of the vector in this region is extended because of the climate and environment, thus seroconversion of animals occurs throughout the year in Xishuangbannai (Jinhong) and Dehong. In contrast, neither seroconversion of animals nor virus isolation was detected in the 15 goats and 2 cattle in the Kunming sentinel herd. The seropositive cattle as well as those from which BTV was isolated had no obvious clinical signs of BT during the monitoring period, suggesting that cattle provide a virus reservoir in the epidemiology of BTV infection (2). Together, these data show that sentinel herds can provide valuable data on the distribution of BTV serotypes and transmission of the virus. This information will ultimately be useful for the prevention and control of BT (8, 26). Sequence comparison of the L2, S7 and S10 genes of bluetongue viruses from the People s Republic of China and other countries The S10 and a portion of the L2 gene segments of Chinese prototype strains of BTV serotypes 1, 2, 3, 4, 12, 15 and 16 were sequenced and compared to the same genes of prototype and field strains of BTV from the United States of America (USA). Phylogenetic analysis of the S10 gene segregated the Chinese viruses into a monophyletic group distinct from the USA viruses, whereas similar analysis of the L2 gene segregated strains of BTV according to serotype, regardless of geographic origin. Phylogenetic analyses of these viruses also indicate that they are more closely related to one another, and to an Australian strain of serotype 1, than they are to prototype strains of bluetongue virus serotypes 2, 10, 11, 13 and 17 from the USA. The S7 gene segments of these viruses were sequenced and compared to the same genes of prototype strains of BTV from the USA, Australia and South Africa. The S7 genes and predicted VP7 proteins of the Chinese viruses were relatively conserved, with the notable exception of 52 Veterinaria Italiana, 40 (3), 2004

23 Global situation serotype 15. Furthermore, phylogenetic analysis of the S7 genes did not predict geographic origin of the various strains of BTV (3, 4, 27). Study and establishment of laboratory systems for diagnosis and monitoring of bluetongue in China The diagnosis and monitoring of BT have been studied and established using the agar gel immunodiffusion (AGID) test, indirect ELISA, c-elisa and VN tests for detection of antibodies to BTV in ruminants, and the SN, immunofluorescence (IF) and immunoperoxidase (IP) methods, antigen capture-elisa (ac-elisa), polymerase chain reaction (PCR) and virus inoculation methods for the isolation and identification of BTV. Techniques for monitoring the presence of bluetongue virus A system was developed for the isolation of BTV, namely: whole blood from ruminants is inoculated into chicken embryos that are then tested by ac- ELISA, blind passage in C6/36 cells, adaptation of isolated viruses to BHK-21 and Vero cells and, finally, virus identification. Red blood cells from whole (anti-coagulated) blood samples are disrupted and diluted prior to intravenous inoculation into chicken embryos. After incubation, liver samples are collected, homogenised, centrifuged, the supernatant collected and tested by ac-elisa. Only positive samples to the ac-elisa are passaged in cell culture to isolate viruses (firstly with one blind passage in C6/36 cells, followed by up to three passages in BHK-21 cells). By using this method, all serotypes of BTV were recovered from blood samples collected from experimentally inoculated sheep. A total of 110 isolates of BTV were obtained from blood samples collected from sentinel cattle. The results show that this is a rapid, accurate and successful system for the isolation of BTV, and that it is especially useful in identifying subclinical BTV infection of cattle (8). Bluetongue antigen capture enzyme linked immunosorbent assay The BTV ac-elisa was used and refined with polyclonal and monoclonal antibodies to detect BTV serogroup antigens. The assay was used to test samples that had been collected from inoculated sheep and the results were compared with traditional virus isolation methods. Identical results were obtained. Tests were conducted on chicken embryo samples (inoculated with 495 clinical blood samples) and revealed 450 embryo samples (inoculated with 96 blood samples) to be positive to the ac-elisa. From 96 blood samples, 83 isolates of BTV were obtained, confirming that 86.5% of results were identical between the two assays. The results indicated that the ac-elisa is an optimal method for detection or identification of BTV serogroup antigen in chicken embryos and cell culture fluid, and that use of ac-elisa can shorten the period and improve efficiency of virus isolation and identification (6, 8). Bluetongue immunoperoxidase staining test The BTV IP staining test was used with serogroupspecific monoclonal antibodies (MAb) for detection and identification of BTV serogroup antigens. The 24 international reference strains of BTV and related viruses (epizootic haemorrhagic disease virus, etc.) were used to evaluate the specificity of the IP test. The results showed this method can detect all 24 serotypes of BTV, without cross-reactions with related viruses. All 110 BTV isolates from sentinel cattle in China were identified using the IP test, and the results were identical to those obtained with the traditional BTV IF antibody test. This simple method provides a sensitive and specific assay for detection of BTV serogroup antigens (8). Detection of bluetongue virus by polymerase chain reaction The PCR assay was used to confirm the presence of BTV in tissue culture, chicken embryos, and clinical blood samples. BTV VP3-specific oligonucleotide primers were used in PCR-based diagnostic tests. The VP3 oligonucleotide primers that successfully amplified gene sequences from the BTV isolates from China previously were successfully tested on BTV isolates from many temperate regions around the world. A set of oligonucleotide primers with sequences derived from the Australian BTV serotype 1 VP2 sequence was tested for its ability to amplify VP2 sequences, irrespective of viral serotype. From an analysis of these VP2 PCR reactions, a minimum of three separate serotypes were present in the isolates tested. A PCR analysis of blood samples from Shanxi, performed in 36 hours, demonstrated the speed and efficiency with which a positive diagnosis could be made, and compared very favourably with the traditional diagnostic and tissue culture methods for identifying BTV in ruminant blood (15, 16). Bluetongue competitive enzyme-linked immunosorbent assay The AGID and c-elisa (which incorporates BTV serogroup-specific MAbs 8A3B6 and 7D3A2) (1) were used to test 560 serum samples that were collected from goats, sheep and cattle in seven provinces in China. The results obtained with Veterinaria Italiana, 40 (3),

24 Global situation 414 samples were identical. A total of 75 sera that were positive in the c-elisa were negative in the AGID, while seven samples positive in the AGID were negative in the c-elisa. The results suggest that the c-elisa may be more sensitive and specific than AGID in the detection of antibodies to BTV. In tests of serum samples from ruminants, samples were positive using c-elisa (% inhibition >40%). The results showed that the c- ELISA is an optimal method for serological surveys of BTV infection. Identification of local isolates of bluetongue virus in China Storage and preparation of bluetongue reference viruses and antiserum International reference strains of BTV of all 24 serotypes (originally obtained from the World Reference Centre, Onderstepoort, South Africa) and Chinese reference strains of BTV (isolated and identified by the Tropical and Subtropical Animal Virology Laboratory in China) were passaged and multiplied in BHK-21 cells, then adapted to Vero cells and titres determined. Reference antiserum was produced for each virus by inoculation of susceptible animals (healthy sheep). These studies showed that: a) antibody and virus were detected at different intervals after infection of animals with the various viruses, and that the neutralising antibody titres of individual antisera differ b) using virus-infected BHK-21 cell supernatant fluid as the inoculum, antisera may exhibit toxicity for BHK-21 cells but not for Vero cells, indicating that the cell lines used in the preparation of antiserum and in the neutralisation test must be different and that strains of BTVs used in the neutralisation test should be adapted to Vero cells c) the neutralising antibody titre of antiserum produced was higher than the international reference antiserum from South Africa. The storage system for reference viruses is a model for the storage of other viruses and bacteria. Identification of local strains of bluetongue virus Since the first outbreak of BT in Yunnan in 1979, BTVs have been isolated from sheep in Yunnan, Hubei, Sichuan, Anhui, Shandong and Shanxi Provinces, from goats in Xinjian and Inner Mongolia, and from cattle in Gansu. These virus strains were identified as BTV by AGID, IF and agar gel electrophoresis. Eight of the Chinese strains were serotyped by micro-sn tests. These showed that strains S 1 from Yuncheng County in Shanxi, S 2 from Jiansu County in Shanxi, Y863P12 from Yunnan, 27 from Xinjian and Yc from Culicoides trapped in the animal laboratory in Kunming were all BTV serotype 1. Both WP7 from Hubei and SWP7 from Sichuan Province were BTV serotype 16. Early results indicate that NMP11 from Inner Mongolia might be BTV-17 but further confirmatory tests are required. In 1998, 110 strains of BTV were isolated from subclinically infected cattle in four sentinel herds in the Yunnan Province. Using the international reference antiserum, the BTV strains were identified to serotypes using the microneutralisation test. The results showed: a) the serotypes isolated included 1, 2, 3, 4, 12, 15 and 16 b) the predominant serotypes were 1, 4, 15 and 16 c) during short periods of just one or two months, two or three different BTV serotypes were isolated from the same cattle. This suggests that animals were either infected with more than one BTV serotype, or that they can be infected sequentially with multiple serotypes d) these results have provided considerable insight into the distribution and seasonal transmission of BTV in China (8, 25, 26, 27). Detection and identification of bluetongue serogroups and serotypes by reverse transcriptase-polymerase chain reaction Based on the sequence of BTV RNA segments L2 and L3, sets of oligodeoxynucleotide primers were designed and synthesised to establish the BTV serogroup- and serotype-specific reverse transcriptase-polymerase chain reaction (RT-PCR) assays. The serogroup-specific RT-PCR assay can detect all Chinese BTV serotypes, without crossreactions with other related viruses. The serotypespecific RT-PCR assays for BTV serotypes 1 and 16 detect only the appropriate self-serotype and no cross-reaction occurs with other serotypes. This RT- PCR technique was applied to identify clinical isolates and detect BTV in clinical samples (red blood cells and tissue samples) from inoculated sheep. The results were identical to the traditional methods. The results also indicated it is a simple, highly sensitive and specific method for the detection and identification of BTV. Bluetongue vaccines Development of bluetongue vaccines Chicken embryo-adapted attenuated and inactivated vaccines of BTV serotypes 1 and 16 were successfully developed and produced. The protection 54 Veterinaria Italiana, 40 (3), 2004

25 Global situation rate of attenuated vaccines and inactivated vaccines were greater than 90% and 75% respectively, and gave immunity that lasted one year and more than six months, respectively. The technical procedure of viral nucleic acid inactivation has been implemented. A total of animals from Yunnan, Sichuan, Hubei, Shanxi and Jiangsu Provinces were vaccinated, which has controlled the incidence of clinical BT disease in China. The effective vaccination schemes were implemented using single or combined vaccination in different epidemic areas (5, 12, 13, 17, 18, 19, 24, 28). Immunomorphology of vaccinated sheep No viral antigen was detected in the target cells of vaccinated sheep by immunohistochemistry, histology, cellular chemistry and electron microscopy. The immune response to BTV infection was both humoral and cellular. In the immune system, T and B cells and peripheral blood peaked at the same time. No viral antigen was detected in the target cells of vaccinated sheep after challenge (24). References 1. Ben J., Li Z., Li H., Melville L.F., Hunt N.T., Li X., Zhang F. & Zhang N. (1996). Comparison of competitive ELISA and agar gel immunodiffusion tests to detect bluetongue antibodies in ruminants in China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Bi Y., Li C., Li S., Qing B., Zhong N., Hu J. & Yang R. (1996). An epidemiological survey of bluetongue in Yunnan Province, China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Bonneau K.R., Zhang N., Zhu J., Zhang F., Li Z., Zhang K., Xiao L., Xiang W. & MacLachlan N.J. (1999). Sequence comparison of the L2 and S10 genes of bluetongue viruses from the United States and the People s Republic of China. Virus Res., 61, Bonneau K.R., Zhang N.Z., Wilson W.C., Zhu J.B., Zhang J.B., Zhang F.Q., Li Z.H., Zhang K.L., Xiao L., Xiang W.B. & MacLachlan N.J. (2000). Phylogenetic analysis of the S7 gene does not segregate Chinese strains of bluetongue virus into a single topotype. Arch. Virol., 145 (6), Feng J. (1996). Development and advances in the prevention of bluetongue disease. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Hawkes R.A., Kirkland P.D., Sanders D.A., Zhang F., Li Z., Davis R.J. & Zhang N. (2000). Laboratory and field studies of an antigen capture ELISA for bluetongue virus. J. Virol. Methods, 85, Hu Y. & Peng Kegao (1996). Analysis of local isolates of bluetongue viruses in China by SDS- PAGE. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Kirkland P.D., Zhang N., Hawkes R.A., Li Z., Zhang F., Davis, R.J., Sanders D.A., Li H., Zhang K., Ba J., He G.F., Hornitzky C.L. & Hunt N.T. (2002). Studies on the epidemiology of bluetongue virus in China. Epidemiol. Infec., 128 (2), Li G., Peng Kegao, Zhang F. & Li X. (1996). Enhancement of growth of bluetongue viruses in cell culture through trypsin treatment. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Li H., Li Z., Zhang K., Liu G., Zhou F. & Shun Y. (1996). Bluetongue vector surveillance in Yunnan and Sichuan Provinces of China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Li H., Li Z., Zhou F., Ben J., Zhang K., Liu G., Li C., Zhang Y., Shi W., Zhao J., Wang J., Du J., Kong D., Yang G., Niu B. & Niu Y. (1996). Establishment of sentinel herds to monitor bluetongue in China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Li Z., Peng K., Zhang K., Li G., Hu Y., Zhou F. & Liu G. (1996). Development of inactivated vaccine for bluetongue in China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Li Z., Zhang K. & Li G. (1996). Cross-protection studies of bluetongue viruses in sheep. In Bluetongue Veterinaria Italiana, 40 (3),

26 Global situation disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Li Z., Zhang K., Li G., Hu Y., Peng K. & Wu D. (1996) Research on the basic properties of Chinese bluetongue viruses. In Bluetongue disease in South- East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Peng K., Gould A.R., Li H., Ben J., Kattenbelt J.A., Li X. & Zhang F. (1996). Detection of bluetongue in China by polymerase chain reaction (PCR). In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Peng K., Li Z. & Li G. (1996). Studies of a nonradioactive gene probe for bluetongue viruses. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Peng K., Li Z. & Wu D. (1996). Analysis of RNA of inactivated bluetongue vaccine by PAGE. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Shun Y., Zhang Y., Hu Z., Yuan Y., Wu Z. & Liu G. (1996). Control of bluetongue disease with attenuated vaccine. In Bluetongue disease in South- East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Wang Q. (1996). The prevention and control of bluetongue disease in China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Zhang K. (1996). The pathogenicity of Chinese bluetongue viruses in sheep, deer and goats. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Zhang F., Li G., Peng K., Li X. & Hunt N.T. (1996). Differential epidemiology of bluetongue antibodies in ruminants in China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Zhang N., Zhang K., Li Z., Hu Y., Li G., Peng K., Wu D. & Li H. (1993). Progress of bluetongue study in China. In Arbovirus research in Australia. Proc. Sixth Symposium (M.F. Uren & B.H. Kay, eds). CSIRO/QIMR, Brisbane, Zhang N., Li Z., Zhang K., Hu Y., Li G., Peng K., Li H., Zhang F., Ben J., Li X,. Zhou F. & Liu G. (1996). Bluetongue history, serology and virus isolation in China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Zhang N., Li Z., Zhang K., Hu Y., Li G., Peng K., Zhou F., Li H., Zhao K. & Liu G. (1996). Development of bluetongue attenuated vaccine in China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Zhang N., Peng K., Li Z., Zhang K., Li H., Hu Y., Li G., Zhang F., Li X. & Ben J. (1996). Identification of local Chinese strains of bluetongue virus to serotype. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Zhang N. & Kirkland P.D. (1998). Bluetongue research in China. Vet. Rec., 143, Zhang N., MacLachlan N.J., Bonneau K.R., Zhu J., Li Z., Zhang K., Zhang F., Xiao L. & Xiang W. (1999). Identification of seven serotypes of bluetongue virus from the People s Republic of China. Vet. Rec., 145, Zhou Y., Cao Z., Cheng Y., Li Z., Peng K., Zhang K., Duan J., Zhang X., Wang F. & Wu J. (1996). Using inactivated vaccine to prevent bluetongue in Hubei Province, China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR) Proceedings No. 66, Canberra, Veterinaria Italiana, 40 (3), 2004

27 Vet. Ital., 40 (3), Global situation Brief overview of the bluetongue situation in Mediterranean Europe, C. Gómez-Tejedor Laboratorio Central de Veterinaria, Ctra. de Algete, Km 8, Algete, Madrid Summary Until recently, the distribution of bluetongue (BT) virus (BTV) in relation to vector distribution, has been between latitudes 40º North and 35º South. Within these limits the disease occurs in parts of America, Asia, Africa and Australia. Although Europe has suffered several BT epizootics, the disease had not become endemic to the region. The situation is changing. BT has recently emerged in some Mediterranean countries of Europe where it had never previously been reported, in particular in Italy, France, as well as in countries in which only sporadic occurrence of the virus had previously been reported, i.e. Spain (late 1950s), Greece and Turkey (late 1970s). It is relevant to underline that some of the recently affected areas in Europe are not situated within the classical latitudes for BT. Furthermore, although they coincided in time, the recent incursions of BT had two separate origins, coming from beyond the eastern and southern boundaries of Europe. The outbreaks of BT of eastern origin commenced at the end of 1998; they were reported in the Greek islands, and then in the summer of 1999, in Turkey and Bulgaria. In 2001, the disease advanced westwards and northwards, reaching central and north-west mainland Greece, and neighbouring Balkan states; serotypes 4, 9 and 16 were incriminated in this epizootic. The outbreaks of southern origin commenced towards the end of BTV was confirmed in Tunisia and spread to north-eastern Algeria. In the summer of 2000, the virus reached the Italian island of Sardinia, spreading also to Sicily and Calabria (the Italian mainland area closest to Sicily). In October 2000, BT was reported on the French island of Corsica and in the Spanish Balearic island of Menorca from where it spread to another Balearic island, Mallorca. In 2001, BT spread across south-west mainland Italy. Originally only serotype 2 was isolated in the epizootics of North African origin; however, in , BTV-4 appeared in Morocco, Spain and Portugal. Keywords Bluetongue Culicoides Europe France Italy Mediterranean Spain Virus. Prior to 1998, only a few epizootics of bluetongue (BT) occurred in Mediterranean Europe and, after a long period of freedom from BT, however, the situation changed significantly. Epizootics of BT virus (BTV) were introduced in the region from two different origins, BTV-2 was introduced from the south and BTV-4, BTV-9 and BTV-16 from the east (Fig. 1). BTV-2 has now been isolated in Italy, Spain (Balearic islands) and France (Corsica). Figure 1 The two origins of the bluetongue outbreaks in Mediterranean Europe Veterinaria Italiana, 40 (3),

28 Global situation BTV serotypes 4, 9 and 16 have been reported from various countries from Italy eastwards. More recently (2004), BTV-4 has appeared in Morocco, Spain and Portugal. Outbreaks of southern origin BTV-2 occurred in Tunisia in 1999 (5) and spread to the Italian island of Sardinia in From Sardinia it spread to two of the Balearic islands of Spain and to Corsica in France (Fig. 2). Surveillance Surveillance in the Balearic islands was based on sentinel bovines, the capture of Culicoides and movement controls of bovines. Sentinel cattle were tested using the enzyme-linked immunosorbent assay (ELISA) and polymerase chain reaction (PCR) (Table I). The possible presence of BTV in the Culicoides captured was tested by PCR. Table I Sentinel bovine surveillance in the Balearic islands, 2001 France 2000 Spain 2000 Algeria 2000 Italy 2000 Tunisia 1999 Farms/animals Menorca Mallorca Ibiza Farms Animals As part of the movement control programme, bovines in Mallorca and in Menorca, were tested using PCR. On mainland Spain, surveillance was performed along the eastern and southern coasts, covering an area that extended 100 km inland from the coastline. Figure 2 Bluetongue outbreaks of southern origin Balearic islands (Spain) Two islands were affected, namely Menorca and Mallorca. Outbreaks The first outbreak was diagnosed on 10 October 2000 and the last on 27 November. In Menorca, a total of 114 outbreaks were reported, while in Mallorca the total reached 191. Sheep vaccination Vaccination of sheep with attenuated BTV-2 vaccine, commenced as early as 24 October and was completed by 10 November in Menorca and by 10 December in Mallorca (7). In Menorca, the number of vaccinated animals, belonging to 205 farms totalled Vaccinated animals on farms in Mallorca totalled In the spring of 2001, sheep over six months of age were vaccinated in both Menorca and Mallorca. On the nearby island of Ibiza, where no outbreak had been reported, all sheep, irrespective of age, were vaccinated at that time. Corsica (France) Outbreaks In Corsica, the first outbreak of BT was diagnosed on 6 October Outbreaks lasted until December. After a period of absence of new outbreaks, additional outbreaks occurred between July and November The virus isolated in both years was the same, BTV-2. A total of 49 outbreaks were reported in 2000: 17 in the north of the island and 32 in the south. In 2001, 335 outbreaks occurred: 211 in the north and 124 in the south (2) (Table II). Table II Number of animals that showed clinical signs, died or were slaughtered in Corsica, 2000 and 2001 Animals Clinical signs Mortalities Slaughtered Sheep vaccination Between 14 November 2000 and 30 April 2001, a total of sheep were vaccinated with attenuated BTV-2 vaccine. 58 Veterinaria Italiana, 40 (3), 2004

29 Global situation Surveillance In Corsica, surveillance was based on sentinel bovines and the capture of Culicoides. Species of Culicoides found were C. imicola, C. obsoletus, C. pulicaris and C. newsteadi (6). Vaccination Ruminants in Italy were vaccinated both in 2002 and 2003 (4). In south-east mainland France, Culicoides captured were C. obsoletus, C. pulicaris and C. newsteadi. No C. imicola was found (6). Italy Outbreaks In 2000, BTV-2 probably reached the island of Sardinia from Tunisia, and then spread to Sicily and to the region of Calabria in the south-west mainland Italy. Italy is therefore discussed in the Eastern origin section. Outbreaks of eastern origin BT serotypes 4, 9 and 16 arrived from the east, most probably from Greece (Fig. 3) (1, 5). BTV serotypes 9, 4, 16 Figure 3 Bluetongue outbreaks of eastern origin Italy 1998 Greece 2001 Turkey Outbreaks In Italy, abundant outbreaks of BT were reported between 2001 and the spring of 2003 (Fig. 4). BTV-2 was the only serotype detected in Sardinia and in the regions of Lazio and Tuscany (3, 4). In Sicily and in the south of the Italian mainland, outbreaks were not only caused by BTV-2, but also by BTV-9. In mid-august 2003, outbreaks of BTV-4 were reported in Sardinia (11); the origin of these outbreaks is still unknown. This was followed by the appearance of BTV-16 towards the end of Infection Figure 4 Bluetongue in Italy in 2003 Source: Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise, Teramo, Italy Monovalent attenuated BTV-2 vaccine was used in Sardinia and the regions of Lazio and Tuscany. Bivalent attenuated BTV-2 and BTV-9 vaccine was used in Sicily and in the southern areas of the mainland. Corsica (France) Outbreaks of BTV-4 were reported in Corsica in October 2003 (10). In September 2004 outbreaks of BTV-16 were reported for first time in Corsica as well as one outbreak of BTV-4 (12). The presence of these two new serotypes in Corsica has modified the vaccination programme to include vaccination against BTV-4 and BTV-16 in addition to BTV-2. Menorca (Spain) Outbreaks Seroconversion Outbreaks of BTV-4 were reported in Menorca in October 2003 (10). A total of 16 outbreaks have been diagnosed in Menorca. BTV-4 did not affect Mallorca. Vaccination against only BTV-4 has been performed in Menorca. New areas affected by bluetongue in 2004 In the late 1950s an epizootic of BTV-10 affected Morocco, mainland Spain and Portugal. Since then, this area was free until 2004, when BTV-4 was reported in the area. Veterinaria Italiana, 40 (3),

30 Global situation Morocco In September 2004, Morocco reported several outbreaks (13) in four provinces of the north-west on the Atlantic coast. The disease spread to other regions of the country. The virus isolated was BTV-4 but the origin has not yet been established. Until 22 November, the date of the last reported case, 380 outbreaks have been diagnosed (M. El Harrak, personal communication) (Table III). Table III Number of outbreaks and morbidity and mortality rates of bluetongue in Morocco, 2004 Outbreaks Morbidity (%) Mortality (%) Spain In October 2004, BTV-4 was detected in sentinel bovines in the south of mainland Spain, in the Cadiz province of Andalusia. Since then, the disease spread to other municipalities of the south and south-west of the mainland affecting four provinces in Andalusia and two in Extremadura. Outside the mainland outbreaks, the disease was detected in Ceuta, a Spanish province on the border of Morocco. Until the last outbreak reported in Spain on 14 December, a total of 328 outbreaks were reported (8) (Table IV). Table IV Number of outbreaks of bluetongue on mainland Spain, 2004 Location No. of outbreaks Andalusia 274 Extremadura 50 Ceuta 4 Total 328 The virological, serological and entomological surveillance that was in place in Spain has been modified and improved in accordance with the new epidemiological situation in the country. Vaccination A total of sheep were vaccinated on mainland Spain against BTV-4, in Extremadura and in Andalusia (8). Portugal Four outbreaks were reported in November These were located in two districts, one in the southeast of the country and the other in the east; both cases occurred near the Spanish border (14). Until the last case which occurred on 14 December, a total of nine outbreaks have been reported in Portugal (9). Serological and entomological surveillance is in progress. As the outbreaks are located near the border with Spain, the level of northern limit of the restricted zone established in Portugal is the same as that of the restricted zone in Spain. References 1. Anon. (2003). OIE Handistatus II database. OIE, Paris. 2. Grégory M., Zientara S. & Hendrikx P. (2002). La fièvre catarrhale du mouton en Corse en 2000 et Bull. Épidémiol. AFSSA, 4, Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise (IZSA&M) (2001). Bluetongue in Italy IZSA&M, Teramo. 4. Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise (IZSA&M) (2003). Bluetongue in Italy IZSA&M, Teramo. 5. Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean Basin Vet. J., 164, Ministère de l Agriculture, de l Alimentation, de la Pêche et des Affaires rurales (2002). Bulletin de surveillance de la fièvre catarrhale du mouton, juillet Bulletin No. 8, Ministerio de Agricultura, Pesca y Alimentación (2002). Informe medidas de lucha contra la epizootia de lengua azul en las Islas Baleares del año Ministerio de Agricultura, Pesca y Alimentación, Madrid, Ministerio de Agricultura, Pesca y Alimentación (2005). Informe de la situación de la lengua azul en España. Ministerio de Agricultura, Pesca y Alimentación, Madrid, Ministério de Agricultura, Pescas e Florestas (2005). Bluetongue in Portugal. Update of information. Ministério de Agricultura, Pescas e Florestas, Lisbon, Office International des Épizooties (OIE) (2003). Bluetongue in France: in the island of Corsica. Dis. Info., 16 (44), Office International des Épizooties (OIE) (2003). Bluetongue in Italy: circulation of virus serotype 4 in Sardinia. Dis. Info., 16 (39), Office International des Épizooties (OIE) (2004). Bluetongue in France: in the island of Corsica. Dis. Info., 17 (38), Office International des Épizooties (OIE) (2004). Bluetongue in Morocco. Dis. Info., 17 (39), Office International des Épizooties (OIE) (2004). Bluetongue in Portugal. Dis. Info., 17 (48), Veterinaria Italiana, 40 (3), 2004

31 Vet. Ital., 40 (3), Global situation Regional overview of bluetongue viruses, vectors, surveillance and unique features in Eastern Europe between 1998 and 2003 D.E. Panagiotatos Ministry of Agriculture, Athens Centre of Veterinary Institutes, Department of Epidemiology and Bio-Statistics, 25 Neapoleos Street, Athens, Greece Summary Between 1998 and 2002, successive epidemic waves of bluetongue (BT) virus infection were recorded in the Balkans giving rise to clinical outbreaks of BT that caused severe direct losses of livestock in several countries, namely: Greece, Bulgaria, Yugoslavia, Kosovo, the Former Yugoslav Republic of Macedonia, Bosnia-Herzegovina and Albania and probably Turkey and Croatia. Affected countries resorted to different control, safeguard, prevention and epidemiological/ surveillance measures against BT but comprehensive and reliable data are by and large lacking. This review attempts an analysis and extrapolation of the local epidemiological profiles and patterns documented in some countries in south-eastern Europe and assuming that the evolution of BT in these countries reflects the situation of BT in the wider region considers some relevant and timely questions of epidemiological significance. Keywords Bluetongue Culicoides Eastern Europe Epidemiology Surveillance Virus. After a prolonged period of historical freedom, presumed freedom or, at worst, minor, sporadic and geographically confined incidents of seroconversion to bluetongue (BT) virus (BTV) in livestock, massive and multiple epidemics of the disease were recorded in south-eastern Europe starting in 1998 and continuing over subsequent years. Between the autumn of 1998 and winter of 2002, successive waves of BT epidemics were recorded in the Balkans, giving rise to a number of clinical outbreaks, causing severe direct losses in several countries, namely: Greece, Bulgaria, Yugoslavia, Kosovo, the Former Yugoslav Republic of Macedonia, Bosnia-Herzegovina and Albania and probably Turkey and Croatia. Affected countries used different control, safeguard, prevention and epidemiological surveillance measures against BT. With the exception of Greece and Bulgaria, comprehensive and reliable epidemiological data are by and large lacking. However, through an analysis and extrapolation of the epidemiological profiles and patterns observed in Greece and Bulgaria, and assuming that the evolution of BT in these two countries reflects closely the BT situation throughout the Balkans, certain general comments can be made and some questions of epidemiological significance emerge which shed new light on the conventional perceptions and clearly call for a new risk assessment and prevention and control strategy against BT. Outstanding questions arising from the study of BT in the Balkans include the following: a) the geographical occurrence and abundance of efficient vector(s) b) the potential involvement of other, more common and widely spread vectors c) the occurrence and distribution of BTV serotypes, in particular those perceived as exotic d) the most appropriate prevention, control and safeguard measures. Temporal and spatial occurrence of bluetongue in Eastern Europe Epidemiological conditions and perceptions preceding primary incursion In the wake of the 1979 epidemic of BT which affected the Greek Island of Lesbos in the eastern Veterinaria Italiana, 40 (3),

32 Global situation Aegean Sea (12, 22), the Greek veterinary authorities were acutely aware of the risk of re-incursion and, consequently, routinely applied active serological monitoring on all Greek islands opposite and along the western Turkish coastline. During 1997 and up until late September 1998 (1), approximately serum samples collected in the Dodecanese islands alone were tested, with negative results. Consequently, the estimated date of primary incursion of BTV into Eastern Europe can be determined with some accuracy. These encouraging findings, however, were shadowed by sporadic reports of BT outbreaks in Turkey throughout the 1990s as well as by the inherent risk factor posed by the documented presence of efficient BTV vectors (Culicoides imicola) on most Greek islands of the eastern Aegean Sea (9, 10). It is worth mentioning that prior to 1998, no other country in eastern Europe, or the rest of Europe, considered BT a relevant and potentially emerging disease. History of recent outbreaks reported in Eastern Europe Against this background, BT was confirmed on four Greek islands in October 1998, namely Rhodes, Kos, Samos and Leros (Fig. 1) adjacent to the western coast of Turkey. The causative virus was identified as BTV serotype 9, which had only been reported previously in in western and southern Turkey. The vector involved was definitely C. imicola. The means of introduction was presumed to be airborne infective vectors carried by the prevailing easterly winds. The source of infection was not identified (1). Between mid-october and late December 1998, 84 outbreaks (flocks) were recorded in the four Greek islands and a total of Clinical outbreaks Serological evidence Figure 1 Regions of Eastern Europe clinically affected by bluetongue in 1998 approximately animals (exclusively sheep) died or were culled due to BT (1). This was the first incursion of BT into Europe since 1979 and caused no real surprise since it involved a known risk area located well within the vector zone between 35 N and 40 N. In late June 1999, BT was reported for the first time in the region of Burgas in south-eastern Bulgaria. By the end of December, the disease had spread in a south-south westerly direction affecting four Bulgarian regions, namely: Burgas, Yambol, Haskovo and Kardjali (Fig. 2). Clinical outbreaks Serological evidence Figure 2 Regions of Eastern Europe clinically affected by bluetongue in 1999 BTV serotype 9 was identified. Vector surveillance failed to confirm the presence of vectors, but revealed an abundance of C. pulicaris and C. obsoletus which have long been suspected as potential vectors of BTV (2, 13). Both the means of introduction and the source of infection remain unclear. In total, 85 outbreaks (or villages) were reported in Bulgaria in 1999 and 667 animals (sheep) died or were culled due to BT (2). This second incursion of BT into Europe does not appear to be linked to the 1998 epidemic of BT in the Greek islands some 600 km to the south, as areas affected were as far north as N and supported the hypothesis that other vectors besides C. imicola may have been involved in some cases (13). In July 1999, Turkey reported the presence of BT, supposedly having originated in Bulgaria, in two provinces bordering Bulgaria and Greece, namely Kirklareli and Edirne (Fig. 2) and responded by vaccinating some sheep with a locally produced live virus vaccine against BTV serotype 4. However, despite the implied success of the vaccination campaign, the field virus isolated in 62 Veterinaria Italiana, 40 (3), 2004

33 Global situation 1999 in the European part of Turkey was later identified as being serotype 9 (13). No systematic entomological studies had been undertaken until this time in the European part of Turkey to determine the presence, geographical distribution and seasonal variation of BTV vectors. Subsequent evolution of BT in Turkey during 1999 remains unclear and the only comment that can be made is the contradiction between official reports, or absence of such reports, and unofficial personal communications. In August 1999, BT was predictably reported in the prefecture of Evros, north-eastern Greece, adjacent to the borders to Bulgaria and Turkey (3). By December, the disease had spread in a south-southwesterly direction along the prevailing wind patterns and involved nine prefectures in mainland Greece. In addition, serological evidence of BTV infection was found in four additional prefectures (4). Incursion and spread was traced along the valley of the Ardas River that flows from Burgas (Bulgaria) to Evros. As in Bulgaria, BTV serotype 9 was involved and C. imicola was not identified (at the time) in northern Greece, thus lending support to the hypothesis that other species of Culicoides may be involved in transmission (13). In the same epidemic, however, outbreaks of BT were reported on the south-eastern tip of Evros, which is far from the predicted direction of spread (Fig. 2). Intensive epidemiological inquires ruled out any link with the known sources of infection and laboratory tests identified BTV serotype 4 in this sub-cluster of outbreaks and subsequently in other outbreaks in mainland Greece (3). This was a novel and unexpected occurrence and, since no official information was available suggesting the recent presence of serotype 4 in the region, the original source of BTV-4 was designated as unknown. It was evident, however, that an incursion of multiple BTV serotypes was in progress and from that point onwards the isolation and typing of as many field strains as possible became a necessity. The conclusion that multiple serotypes of BTV had entered the region was confirmed in September 1999 when, shortly after an official but flimsy report of BT in the vicinity of Smyrna, Turkey, a massive epidemic of BT swept the islands of Lesbos and the Dodecanese in the eastern Aegean Sea, adjacent to the Turkish coast. Again, intensive epidemiological inquiries ruled out any link with the known sources of infection on mainland Greece and laboratory tests identified BTV serotypes 4 and, astonishingly, 16. By December 1999, two additional prefectures on the islands were clinically affected, specifically Chios and Lesbos, while serological evidence of BTV infection was found in another prefecture, namely Samos (Fig. 2). In total, outbreaks were recorded in Greece in 1999 causing the death or culling of sheep (3). The following year, 2000, was relatively quiet, with no clinical evidence of BT anywhere in Eastern Europe, except in the prefecture of Arta in centralwestern Greece, where a cluster of ten BT outbreaks caused by BTV serotype 4 involving 50 animals was linked to a known internal source of infection through illegal movement of viraemic bovines (5) (Fig. 3). Prompt identification and efficient application of targeted vector control measures prevented any spread and the epidemic burned itself out. Areas free of bluetongue Seroprevalence : 2.6% Seroprevalence: 7.9% Clinical symptoms Lake Figure 3 Regions of Greece that were clinically affected by bluetongue in 2000 In late September 2001, BT was first reported in north-western Greece, adjacent to the borders with the Former Yugoslav Republic of Macedonia and Albania. The epidemic gradually expanded to the south, eventually involving 11 prefectures, most of which were not affected during the 1999 epidemic (6) (Fig. 4). C. imicola was found in the eastern and coastal prefectures and C. obsoletus in the northern and central mountainous areas. Remarkably, however, BTV serotype 1 was identified in both the primary and northern-most outbreaks. This particular BTV serotype has never been reported anywhere near the Mediterranean Basin or Middle East and the epidemiological picture is further obscured by the failure to identify this serotype anywhere else in Eastern Europe in the course of the 2001 epidemic. In total, 174 outbreaks were recorded in Greece in 2001 accounting for the death or culling of sheep. Veterinaria Italiana, 40 (3),

34 Global situation Clinical outbreaks Clusters Figure 4 Regions of Eastern Europe clinically affected by bluetongue in 2001 Following the alert from Greece, reports came in from various countries in the region within a matter of days, retrospectively announcing the presence of BT in their territories in 2001, as follows: a) In early October 2001, Bulgaria announced the widespread presence of BTV along the entire length of its western border with the Former Yugoslav Republic of Macedonia and Serbia, and in particular in the Kiustendil Province, extending as far north as N on the Bulgarian- Romanian border (Fig. 4). The BTV serotype involved has not been identified and, again, no C. imicola was detected in the region. In total, 75 outbreaks (or villages) were affected by BT in Bulgaria in 2001 and 23 severely affected sheep died or were culled (4). b) In October and November 2001, the Former Yugoslav Republic of Macedonia reported a total of 36 outbreaks, starting from the district of Kriva Palanka on the borders with Bulgaria and Serbia and eventually spreading along the entire length of the northern, western and eastern borders with Albania, Kosovo, Yugoslavia and Bulgaria (16) (Fig. 4). A total of 178 sheep died or were culled and the BTV serotype involved was not identified. c) In October 2001, Kosovo reported six outbreaks of BT in as many villages in the provinces of Strpce, Podujevo, Glogovac and Vitina (14) (Fig. 4). The BTV serotype was not identified. d) In late October 2001, Yugoslavia confirmed the presence of BT (suspected since late August 2001) in 37 outbreaks extending along the southern borders with Bulgaria and Kosovo (15) (Fig. 4). BTV serotype 9 was identified. e) In mid-december 2001, Croatia announced the suspicion of BT in three outbreaks in the region of Dubrovnik (17) (Fig. 4). The suspicion was later confirmed but the BTV serotype was not identified. 64 Veterinaria Italiana, 40 (3), 2004

35 Global situation In the following year, 2002, evidence of BTV circulation was reported from several countries, as follows: a) In early September 2002, Bulgaria reported seroconversion in three sentinel animals in the Smolian region, near the border with Greece (18) (Fig. 5). The BTV serotype involved was not identified. b) At the same time (early September 2002), Bosnia- Herzegovina reported 19 outbreaks for the first time, involving 169 animals in 11 villages (18) (Fig. 5). BTV serotype 9 was identified. c) In September and October 2002, Yugoslavia reported 9 outbreaks involving 25 animals in the regions of Sebac (44 40 N) and Kraljevo (43 43 N) (19) (Fig. 5). The BTV serotype involved was not identified. d) In December 2002, Albania reported one outbreak in the region of Librazhd, near the border with the Former Yugoslav Republic of Macedonia (8) (Fig. 5), in which apparently no animals died or were culled. The BTV serotype involved was not identified. Clinical outbreaks Figure 5 Regions of Eastern Europe clinically affected by bluetongue in 2002 Since then, there have been no reports of BTV circulation anywhere in Eastern Europe (until the time of writing at the end of September 2003). Serotypes isolated in Eastern Europe Considering the history of BT in Eastern Europe from 1998 to 2002, one of the most striking and unexpected findings is the multitude of different BTV serotypes that were identified in the region. Indeed, at the beginning of the primary incursion, the only historical data was that BTV serotype 4 was identified on the island of Lesbos in 1979 (9, 22) and that BTV serotypes 2, 4, 6, 9, 10, 13 and 16 had been reported over a number of years in Anatolian Turkey, Syria, Jordan and Israel (13). However, although the westward movement of some of these serotypes is documented, there was no hint that they may already be at the threshold of Europe. BTV serotype 1, in particular, has never previously been reported anywhere near Europe or the Middle East. The unravelling of the mystery was triggered in Greece in 1999 as a result of the failure to explain some outbreaks on epidemiological grounds. Once it was understood that more than one BTV serotype may occur in the same country or region, the Greek veterinary authorities undertook to isolate and type as many field strains of the virus as possible. Unfortunately, due to inadequate resources and laboratory capabilities, this policy was not adopted by other countries in the region and, consequently, only a very limited number of BTV field strains have been isolated and typed from Eastern Europe. In summary, the BTV serotypes identified in Eastern Europe, except Greece, between 1999 and 2002, are as follows (Fig. 6): Bulgaria: serotype 9 (1999: 1 typing) Turkey: serotype 9 (1999: 1 typing) serotype 16 (2000: 1 typing) Serbia: serotype 9 (2001: 1 typing) Bosnia-Herzegovina: serotype 9 (2002: 4 typings). The BTV serotypes identified in Greece between 1998 and 2001 are summarised as follows (Fig. 6): 1998 : serotype 9 (5 typings) 1999 : serotype 9 (10 typings) : serotype 4 (12 typings) : serotype 16 (4 typings) 2000 : serotype 4 (1 typing) 2001 : serotype 1 (4 typings) : serotype 4 (3 typings) : serotype 9 (3 typings). Furthermore, isolation and typing of >100 frozen field samples collected between 1999 and 2001 is in progress. Vectors identified in Eastern Europe The distribution and vectorial capacity of efficient and potential BTV vectors, as well as the impact of climatic changes on these factors has been comprehensively reviewed (13). Prior to 1998, however, knowledge on the abundance, geographical distribution and seasonal variation of BTV vectors in Eastern Europe was fragmented and limited to certain Greek islands of the eastern Aegean Sea (9, Veterinaria Italiana, 40 (3),

36 Global situation BTV-9 BTV-4 BTV-16 BTV-1 Figure 6 Bluetongue virus (BTV) serotypes identified in Eastern Europe, ). Furthermore, previous random catches had failed to identify C. imicola in mainland Greece. In the context of the BT epidemics that occurred in Eastern Europe in the period from 1998 to 2001, systematic entomological surveys were undertaken in Greece and Bulgaria and they are still ongoing. All available findings in Bulgaria, as well as preliminary findings in Greece in 1999 and 2000, failed to identify C. imicola anywhere north of 40 N while they revealed an abundance of C. obsoletus and C. pulicaris. This led to the working hypothesis that the latter two species may be potential, though much less efficient, vectors of BTV but that they compensate for their low efficacy by their large populations. If this hypothesis proves correct, then BT becomes very relevant for large parts of western and northern Europe previously considered to be free of the disease due to the absence of efficient vectors. Accumulated results of vector monitoring in Greece from 1999 to 2002 are summarised in Figure 7 (7). The results indicated that: C. imicola occurs regularly in mainland Greece, particularly in the eastern coastal areas C. imicola has been identified in northern Greece, north of 40 N, near the border with Bulgaria. In regard to the trapping protocol, mainland Greece was composed of 59 quadrants of 50 km 50 km 2 (labelled 1-59), which were sampled for Culicoides over two years (Fig. 7). Two farms (at least 10 km apart) were sampled in each quadrant. During the summer (July to October), each farm was sampled for two nights. During the winter (December to March), farms where C. imicola was found in the summer were sampled for a further five to seven nights. Serological monitoring of sentinel animals Serological monitoring (surveillance) of sentinel herds was introduced in Bulgaria and Greece and routinely applied after each annual BT epidemic so as to detect residual BTV circulation in the affected 66 Veterinaria Italiana, 40 (3), 2004

37 Global situation Culicoides imicola Culicoides pulicaris Culicoides obsoletus 30 quadrants sampled in quadrants sampled in 2002 Figure 7 Vectors of bluetongue identified in Greece, areas and/or re-incursion of BTV in areas at risk. The principle was the same in both countries but the objectives and the methods were different, as follows: a) In Bulgaria, sentinel herds were widely distributed and comprised 10 cattle and 10 goats each. From 2000 to 2001 some 40 sentinel herds were deployed solely along the Greek-Bulgarian and Turkish-Bulgarian borders in an effort to detect external re-incursion rather than internal residual infection. Following the 2001 epidemic in western Bulgaria, a further 22 sentinel herds were also deployed along the western borders following the same rationale (Fig. 8). b) In Greece, sentinel herds were established in both affected areas and areas at risk and involved naive cattle exclusively. As a rule, five groups of 10 cattle each were placed in each targeted prefecture near vector breeding or outbreak sites and the list of serologically monitored prefectures was added to address the issue of the annual evolution of BTV (Fig. 8). c) In both countries, serological monitoring was performed seasonally (from April to December) and sentinels were sampled every 30 days and tested for antibodies to BTV. In case of seroconversion of sentinel animals in Greece, virus isolation was attempted for typing and the viraemic animal was eliminated. The results of serological monitoring of sentinel herds from 1999 to 2003 are summarised as follows (the location of seroconverting sentinels indicated in Figure 8): a) In Bulgaria, in 2000 and 2001, approximately samples were tested with consistently negative results. In 2002, approximately samples were tested and three seroconversions were detected in late August in the district of Smolyan. The BTV serotype involved was not identified. In the absence of any other evidence of virus circulation, seroconversion was attributed to recurrence from previously infected animals through a carrier state mechanism involving the γδ T-cells (21). In 2003, serological monitoring of sentinel animals continued, with presumably negative results. Veterinaria Italiana, 40 (3),

38 Global situation Seroconversion (colour = year) Figure 8 Serological monitoring of sentinels for bluetongue in Greece, b) In Greece, in 1999, 12 seroconversions were observed from 639 samples collected in two prefectures. In 2000, 23 seroconversions were observed in samples from 17 prefectures. In 2001, 46 seroconversions were detected in samples from 18 prefectures. In 2002, no seroconversion was observed from samples taken in 17 prefectures. In 2003 (until the end of September), no seroconversion was observed in samples from 14 prefectures. Seroprevalence in susceptible livestock Insofar as is known, large-scale serological surveys for antibodies to BTV in the general livestock population have only been undertaken in Greece as part of internal safeguard measures requiring premovement testing (with negative results) of animals. Although more than samples have been serologically tested since 1998, Figure 9 represents only the results of approximately serological tests carried out between April and May This selective presentation is justified because it reflects the accumulated seroprevalence over successive waves of BT epidemics (7). On the basis of results presented in Figure 9, the following comments can be made specifically for Greece: a) Despite multiple incursions and successive BT epidemics, large areas of the country have not been affected by the disease. Arguably, this provides a measure of the success of disease control and safeguard measures. b) Seroprevalence in most affected areas ranges from between 1% and 25%, leaving enough naive animals to sustain a new epidemic should a reincursion or recurrence of BTV infection occur. NA 0 < >50 Figure 9 Seroprevalence in susceptible livestock in Greece, 2002 c) Seroprevalence in areas at greatest risk is 50% or higher (as high as 90% in the islands) and, therefore, the animals have already developed a natural, lasting and effective immune response 68 Veterinaria Italiana, 40 (3), 2004

39 Global situation against the BTV serotype(s) that prevail in the area. Control-safeguard and preventive measures against bluetongue in Eastern Europe Control measures With the notable exception of Kosovo, where clinically affected animals were spared due to financial constraints, all countries of Eastern Europe applied similar control measures when confronted with the epidemic, namely: a) modified stamping-out policy by slaughter and destruction of clinically affected animals b) vector control measures using insecticides and/or insect repellents c) intensive clinical, sometimes augmented by serological, surveillance. Safeguard measures All affected countries in Eastern Europe established protection and surveillance zones, extending over a radius varying from 20 km to 100 km, and introduced movement restrictions of animals and germplasm products from these zones. In some countries, a curfew was imposed on animal movements from dusk to dawn and gatherings of animals (e.g. trade fairs, exhibitions etc.) were suspended. Preventive measures and vaccination As mentioned above, in the autumn of 1999, Turkey vaccinated some sheep along the borders with Greece and Bulgaria. The vaccine used was a nationally produced live virus vaccine containing serotype 4 but information is lacking concerning the application and results of vaccination. It should be remembered, however, that the BTV serotype circulating in the area at the time was later identified as serotype 9. Bulgaria resorted to vaccination as a means to prevent recurrence of BT in the areas affected in The vaccination campaign was conducted in early 2000, after the lambing season, and involved some lambs. A commercially available pentavalent live-attenuated vaccine containing serotypes 3, 8, 9, 10 and 11 was used. With the exception of seroconversions observed in sentinel animals in 2002, subsequent evolution of BT in the vaccination area was no different from that in the adjacent prefectures of northern Greece where no vaccination was practised and, therefore, the effects of vaccination are uncertain. It is noted, however, that Bulgaria refrained from vaccination during the 2001 BT epidemic. Kosovo announced its intention to apply mass vaccination in 2002, but no follow-up information is available. The rest of the affected counties in Eastern Europe did not resort to vaccination. Discussion of the main epidemiological features of bluetongue in Eastern Europe Although clinical manifestation of BT is a far from safe and accurate criterion of BTV circulation, understandably in Eastern Europe it was the one most commonly relied upon to signify presence and delineate spread of BTV infection. This partly explains the lack of uniform, comprehensive and consistent epidemiological data, further aggravated by a likely under-detection and/or under-reporting of the disease. A further complicating factor was the different definition of an outbreak used by different countries in the region to describe the spatial distribution of BT, with some countries attributing the term to individual flocks and other countries encompassing entire villages. Apparent differences in the virulence of different BTV serotypes and strains may also have complicated reporting of BT in Eastern Europe. Clinical observations made in Greece suggest the following: a) BTV serotype 9 is consistently highly virulent, causing severe clinical symptoms and high morbidity and mortality (on average 25% and 10%, respectively). BTV-9 is the only serotype identified so far in Eastern Europe, with the exception of Greece b) BTV serotypes 4 and 16 are generally less virulent, causing mild and transitory clinical symptoms, low morbidity (<10%) and almost no mortality. As an example, it is noted that during the 2000 epidemic in Greece (due to serotype 4), BT was obscured by concurrent Orf infection and the combined morbidity/mortality rates were 10% and 0%, respectively BTV serotype 1 displays varying virulence depending on the species and abundance of local vectors, breed of affected animals (sheep) and local climatic conditions and terrain. As an example, it is noted that during the 2001 epidemic in the north-western part of Greece, morbidity and mortality attributed to BTV-1 were 4.5% and 0.7%, respectively, while in the epidemic that occurred at the same time due to the same serotype on the island of Lesbos, morbidity and mortality were approximately 30% and 15%, respectively. Veterinaria Italiana, 40 (3),

40 Global situation Available serotyping data from strains of BTV in Greece indicated that it is not uncommon for two or more BTV serotypes to be identified in the same country, region or flock. It is difficult to accept that the multitude of BTV serotypes identified so far in Greece (i.e. serotypes 1, 4, 9 and 16) simply appeared out of nowhere and that these are self-restricted inside national boundaries. On the contrary, epidemiological reasoning strongly suggests that the situation of BT in Greece reflects the evolution and spread of the disease in the wider region and appears to be more complicated than elsewhere because of more stringent epidemiological surveillance. If this working hypothesis is valid, complete and accurate mapping of BTV serotypes circulating in Eastern Europe is an essential prerequisite to the design and implementation of efficient disease control/prevention policies and, therefore, should become one of the priorities of surveillance. Another provisional finding of epidemiological significance emerging from the epidemics of BT in Eastern Europe between 1998 and 2002 is the potential implication of less efficient, but more abundant and widely occurring BTV vectors, notably C. obsoletus and C. pulicaris. Recent publications support this possibility (11, 20). Arguably, the recent identification of C. imicola in northern Greece, near the border with Bulgaria casts doubt on the exact vector species implicated in the Balkans and clearly warrants further vector monitoring in the region. In connection with vector monitoring, it is understood that BTV vectors are certainly a risk factor and merit close monitoring but do not signify the presence of BTV per se, as long as they are not infective. In this respect, the consistent presence of efficient BTV vectors in a disease-free/seroconversion-free region where sufficient numbers of naive susceptible animals are present, might be construed to imply absence of BTV circulation in this region. Serological monitoring of sentinel animals is an indispensable component of epidemiological surveillance in areas at risk for either re-introduction or recurrence of BT, although the scheme must be correctly deployed and properly managed. Seroconversion in sentinels not only provides early warning of BTV circulation, but also may be used as an indicator for initiating early and targeted control/prevention measures, e.g. vector control in the vicinity. The success of such early measures depends largely on local conditions, such as species and abundance of vectors, climate and terrain but, under certain circumstances, this approach may provide the basis for an alternative prevention policy. This was probably the case in northern mainland Greece in 2001 and in southern Bulgaria in 2002, where seroconversions in sentinel animals were not followed by an epidemic, and serological evidence indicated that BTV did not persist either in the sentinel animals or in contiguous herds. Random serological surveys in the livestock population are considered impractical and of little relevance when faced with an active epidemic but they may provide valuable information in the aftermath or between epidemics of BT. The value of such surveys includes the following: a) helping to accurately delineate affected areas b) evaluating the success of safeguard and control measures c) providing insight to the probable evolution of BTV in areas at risk and to applicable surveillance and prevention measures. To qualify this last statement, where seroprevalence is low, clinical surveillance is a meaningful component of overall epidemiological surveillance, while where seroprevalence is high, the animals have already developed a natural, lasting and effective immune response to the particular BTV serotype(s) circulating in the region and, consequently, vaccination is superfluous. Control measures must reflect the complex and elusive epidemiology of a vector-borne disease, such as BT, thus conventional control measures, such as culling of clinically affected animals, are of psychological and financial rather than of epidemiological significance. However, field experience gained in Greece suggests that vector control measures may have a beneficial effect in reducing BTV circulation and proliferation if applied in a timely, targeted and multi-level manner, i.e. at breeding sites, inside and around animal holdings and on individual animals. Safeguard measures, on the other hand, aim to prevent or reduce the spread of infection through movements of viraemic animals, their semen and embryos, and may be effective as long as they are strictly observed. Finally, with vaccines which are commercially available today, vaccination remains a controversial tool to control or eradicate the disease and its apparent value is limited to reducing direct losses and culling due to severe clinical symptoms. Conclusions and recommendations With the exception of the Greek islands of the eastern Aegean Sea, which are constantly at risk due to either the re-introduction or year-round persistence of C. imicola, and without prejudice to the situation in the northern Balkans and Turkey, it would appear that BTV has not become endemic and may even be diminishing in south-eastern Europe. The recession of the epidemic, however, is 70 Veterinaria Italiana, 40 (3), 2004

41 Global situation no reason for complacency but should be a stimulus to enrich collective and shared knowledge and understanding of the complicated factors that influence the disease. Furthermore, a co-ordinated regional approach that will strengthen multilateral co-operation and ensure prompt dissemination of reliable information is an essential component for a meaningful disease control/prevention strategy. To this end, the European Union is sponsoring three timely and relevant research programmes aiming, respectively, to achieve the following: a) develop predictive models allowing identification of regions at risk and mapping of BTV vectors therein. b) assess the safety and efficacy of existing live virus vaccines and develop inactivated whole-virus or sub-unit vaccines. c) establish a database of BTV genome segments to allow tracing back of BTV incursions and to enable detection of live vaccine strain reversion to virulence. Additional efforts must be made, however, to attract a broader participation in and a firmer commitment to these programmes. Acknowledgements Grateful thanks are extended to the National Veterinary Service of Bulgaria for freely and willingly sharing information, and to colleagues and friends Kiki Nomikou, Olga Mangana-Vougiouka and Michalis Patakakis for verifying the facts and debating the views expressed. Much of the field and laboratory work on BT was performed in Greece with the assistance of grants from the European Union (Contracts No. QLK2-CT and QLK2-CT ). References 1. Anon. (1998). Report on the incursion and evolution of bluetongue in Greece in Ministry of Agriculture, Directorate General of Veterinary Services, Department of Infectious Diseases, Athens. 2. Anon. (2002). 2nd Annual Report: Bluetongue and other Culicoides-borne diseases threatening the EU. Ministry of Agriculture, State Veterinary Service, Sofia. 3. Anon. (1999). Final report on the incursion and evolution of bluetongue in Greece in Ministry of Agriculture, Directorate General of Veterinary Services, Department of Infectious Diseases, Athens. 4. Anon. (1999). Preliminary report on the incursion and evolution of bluetongue in Greece in Ministry of Agriculture, Directorate General of Veterinary Services, Department of Infectious Diseases, Athens. 5. Anon. (2000). Report on the recurrence of bluetongue in Greece in Ministry of Agriculture, Directorate General of Veterinary Services, Department of Infectious Diseases, Athens. 6. Anon. (2001). Report on the incursion and evolution of bluetongue in Greece in Ministry of Agriculture, Directorate General of Veterinary Services, Department of Infectious Diseases, Athens. 7. Anon. (2002). 2nd Annual Report: Bluetongue and other Culicoides-borne diseases threatening the EU. Ministry of Agriculture, Directorate General of Veterinary Services, Athens. 8. Anon. (2003). OIE Handistatus II database. OIE, Paris. 9. Boorman J.P.T. (1986). Presence of bluetongue virus vectors on Rhodes. Vet. Rec., 118, Boorman J.P.T. & Wilkinson P.J. (1983). Potential vectors of bluetongue in Lesbos, Greece. Vet. Rec., 113, Caracappa S., Torina A., Guercio A., Vitale F., Calabro A., Purpati G., Ferantelli V., Vitale M. & Mellor P.S. (2003). Identification of a novel bluetongue virus vector species of Culicoides in Sicily. Vet. Rec., 153, Mastroyanni M., Axiotis I. & Stoforos E. (1981). Study of the first outbreak of bluetongue disease in sheep in Greece. Bull. Hell. Vet. Med. Assoc., 32 (2), Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean Basin Vet. Rec., 164 (1), ProMED-mail (2001). Bluetongue Kosovo: OIE Report. Archive No , 12 October (promedmail.org/pls/askus/f?p=2400:1001: ::::f2400_p1001_back_page,f2400_p1 001_archive_number,f2400_p1001_use_archive:1202, ,y accessed on 23 June 2004). 15. ProMED-mail (2001). Bluetongue Yugoslavia. Archive No , 6 November (promedmail.org/pls/askus/f?p=2400:1202: ::no::f2400_p1202_check_display,f24 00_p1202_pub_mail_id:x,16508 accessed on 23 June 2004). 16. ProMED-mail (2001). Bluetongue Bulgaria, Greece, Macedonia: OIE Report. Archive No , 14 November (promedmail.org/pls/ askus/f?p=2400:1202: ::no::f2 400_p1202_check_display,f2400_p1202_pub_mail_id :x,16556 accessed on 23 June 2004). 17. ProMED-mail (2001). Bluetongue, sheep Croatia (Dubrovnik): suspected. Archive No , 23 December (promedmail.org/pls/askus/f?p=2400: 1202: ::no::f2400_p1202_chec k_display,f2400_p1202_pub_mail_id:x,17104 accessed on 23 June 2004). 18. ProMED-mail (2002). Bluetongue, ruminants Bosnia & Herzegovina, Bulgaria. Archive No. Veterinaria Italiana, 40 (3),

42 Global situation , 6 September (promedmail.org/pls/ askus/f?p=2400:1202: z438030::no::f 2400_p1202_check_display,f2400_p1202_pub_mail_i d:x,19245 accessed on 23 June 2004). 19. ProMED-mail (2003). Bluetongue Yugoslavia: OIE. Archive No , 3 February (promedmail.org/pls/askus/f?p=2400:1202: ::no::f2400_p1202_check_display,f24 00_p1202_pub_mail_id:x,20590 accessed on 23 June 2004). 20. Savini G., Goffredo M., Monaco F., de Santis P. & Meiswinkel R. (2003). Transmission of bluetongue virus in Italy. Vet. Rec., 152, Takamatsu H., Mellor P.S., Mertens P.P.C., Kirkham P.A., Burroughs J.N. & Parkhouse R.M.E. (2003). A possible overwintering mechanism for bluetongue virus in the absence of insect vectors. J. Gen. Virol., 84, Vassalos M. (1980). Cas de fièvre catarrhale du mouton dans l Ile de Lesbos (Grèce). XLVIIIth General Session, Report No Bull. Off. Int. Épiz., (7-8), Veterinaria Italiana, 40 (3), 2004

43 Vet. Ital., 40 (3), Global situation Overview of bluetongue disease, viruses, vectors, surveillance and unique features: the Indian sub-continent and adjacent regions D. Sreenivasulu (1), M.V. Subba Rao (2), Y.N. Reddy (3) & G.P. Gard (4) (1) Department of Microbiology, College of Veterinary Science, Tirupati AP, India (2) Faculty of Veterinary Science, Acharya N.G. Ranga Agricultural University, Rajendra Nagar, Hyderabad AP, India (3) Department of Microbiology, College of Veterinary Science, Rajendra Nagar, Hyderabad AP, India (4) 8 Casey Crescent, Mystery Bay, NSW 2546, Australia Summary The occurrence of bluetongue (BT) disease in India was initially confined to exotic breeds of sheep and subsequently became endemic in native breeds. BT virus (BTV) antibodies are common in cattle, buffaloes and goats although clinical disease has not been reported. Exotic breeds of sheep and their cross-breeds are more susceptible to disease than native breeds. Overall, morbidity, mortality and case fatality rates of 9.3%, 2.7% and 28.8%, respectively, have been reported in rural flocks; these rates are higher than in organised farms. The disease is mostly cyclical in occurrence. Outbreaks usually occur between June and December during the monsoon period when livestock biting midges greatly increase. BTVs have been isolated from native sheep, and sentinel herds have been used to demonstrate virus activity. A total of 21 serotypes of BTV have now been reported in the country. Major impediments to control the disease include the presence of multiple virus serotypes, the broad vertebrate host range of the virus and a lack of detailed knowledge of vectors. Inactivated vaccines prepared from local isolates are currently under field trials. BTV occurs in regions adjacent to India. An antibody prevalence of 48.4% has been reported in Pakistan with serotypes 3, 9, 15, 16 and 18 identified. BTV antibody, but not disease, has been reported in Bangladesh and Sri Lanka. Keywords Bluetongue Bluetongue virus serotypes Epidemiology India Serological surveillance Vectors. Bluetongue (BT) has become one of the important sheep diseases of the Indian sub-continent. The disease was first reported in Pakistan in 1959 and in India in 1964 (19). The disease has since become established in India, a geographically vast and climatically diverse country. The Indian subcontinent, a peninsula, lies between 8.4 N and 37.6 N and 68.7 E and E. India is divided into seven climatic regions, namely: the northern mountains, the northern plains, the Rajastan Desert, the Deccan plateau, the west coast, the south-east coast lands and Assam in the extreme north-east. The Indian climate is dominated by the great wind system called the Asiatic monsoon and which reverses direction at certain times of the year. From June to October, India is influenced by the moist rain-bearing monsoon from the south-west. The coolest, driest period over most of India is from December to February. From March to May the climate becomes very hot and the drought continues. Usually, the monsoon enters the south during late May or early June, reaching the north about six weeks later. In some years, the rains are torrential, but in other years they will be only light. India has significant populations of domestic and wild ruminants, which are known to be susceptible to BTV infection. Several exotic breeds of sheep were introduced into the country between 1960 and 1970 for the genetic improvement of the national flock by crossbreeding with native breeds. This increase in the national susceptible population, along with favourable climatic conditions, appears to have led to the establishment of BT in the country. Veterinaria Italiana, 40 (3),

44 Global situation Epidemiology First decade of bluetongue ( ) After the initial report of BT in Maharashtra, the disease was reported in exotic sheep, namely Southdown, Rambouillet, Russian Merino and Corriedale, between 1967 and Severe BT was also reported in the Dorset breed in Andhra Pradesh in However, the native sheep maintained in close proximity did not present any symptoms. However, the disease was subsequently recorded in native sheep and disease outbreaks have been reported annually since Endemic phase During 1981, BT was widely spread in southern India. Initially, the disease was detected in Karnataka and in the adjoining regions of Maharashtra and Andhra Pradesh, with mortality rates ranging from 2% to 50%. Morbidity was as high as 80%. Later, in 1983, BT outbreaks were reported all over Andhra Pradesh with a case fatality rate of 21.9%. From 1985 onwards, outbreaks were recorded regularly in Andhra Pradesh with case fatality rates ranging from 2.37% to 38.14% (Table I). A cyclical pattern of the disease was observed with variations in severity of infection. Table I Outbreaks of bluetongue in native sheep in Andhra Pradesh, Year Outbreaks Cases Deaths Case fatality (%) Source: Animal Disease Surveillance Reports, Department of Animal Husbandry, Government of Andra Pradesh, India The outbreaks of the disease in Maharashtra were characterised with morbidity and mortality rates of 7.66% and 1.11%, respectively. The case fatality rate was 11.82% (5). Later, an increase in the severity of infection was reported by Kulkarni et al. (10) with overall morbidity of 32%, mortality of 8% and a case fatality rate of 25% in rural areas. The disease was recorded regularly in Tamil Nadu where a total of 258 outbreaks were reported between 1986 and Saravanabava (20) reported morbidity ranging from 3.3% to 22.8% and mortality from 0% to 6.1%. The pattern of disease was studied in the organised farms and rural flocks of Andhra Pradesh. The study revealed that the pattern of the disease in organised farms and rural flocks is quite different. Morbidity, mortality and case fatality rates of rural and organised farms were 9.34%, 2.69%, 28.84% and 6.22%, 0.47%, 7.63%, respectively. Higher morbidity and mortality in rural areas may be because of stress factors, such as poor nutrition, parasitic burden, fatigue due to long walks and absence of veterinary aid. Investigations in Andhra Pradesh revealed that sheep aged 6 to 12 months were more susceptible than adults. The disease has not been reported in lambs. Similar observations were also reported from Maharashtra and Haryana (5, 25). The occurrence of BT varies between parts of India depending on time of rainfall. Maximum numbers of outbreaks were recorded during the north-east monsoon period (October to December) followed by the south-west monsoon period (June to September) in Andhra Pradesh (Table II). Similarly, in Tamil Nadu the outbreaks were more frequent during the north-east monsoon period (20). In Rajasthan, the outbreaks occurred mostly in September and November (12). Clinical disease During the initial outbreaks in the country, all exotic sheep breeds imported into India (Merino, Rambouillet, Corriedale, Dorset and Suffolk) exhibited classical signs of BT. Similar clinical observations of less intensity were noticed in crossbred sheep. However, clinical disease was slightly different in native sheep, the major difference being that swelling of the lips and face was less conspicuous. Mucocutaneous borders appeared to be very sensitive to touch and to bleed easily upon handling. The classical signs of cyanosis of the tongue and reddening of the coronary band are not a common feature of the disease in native sheep. Clinical disease has not been reported in cattle, buffalo and goats in spite of high seroprevalence. Clinical disease is known to occur in Pakistan. 74 Veterinaria Italiana, 40 (3), 2004

45 Global situation Table II Seasonal occurrence and number of outbreaks of bluetongue in Andhra Pradesh, Year South-west monsoons (June-Sep) Rainfall in mm* No. of outbreaks ** North-east Winter Hot weather South-west North-east Winter period monsoons period period monsoons monsoons (Jan-Feb) (Oct-Dec) (Jan-Feb) (Mar-May) (June-Sep) (Oct-Dec) Hot weather period (Mar-May) Total * Source: Statistical Abstracts, Directorate of Economics and Statistics, Government of Andra Pradesh, India ** Source: Animal Disease Surveillance Reports, Department of Animal Husbandry, Government of Andra Pradesh, India However, the disease has not been reported from Bangladesh, Myanmar and Sri Lanka (OIE, Annual Report, 2003). Serosurveillance Extensive serological surveys have been undertaken in different parts of the country. Studies conducted in Andhra Pradesh during 1991 revealed a higher prevalence of BT virus (BTV) antibodies in sheep and goats (45.71% and 43.56%, respectively) than in cattle (33.4%) and buffalo (20%). This higher prevalence in small ruminants may reflect their involvement in the basic ecology of the virus. Similar observations were made by Sharma et al. (21) and Prasad et al. (17) in Rajasthan and Haryana. Harbola et al. (5) reported BTV antibodies in 37.5% of sheep serum samples collected from Maharashtra State. Others have reported the seroprevalence of BTV antibodies from Gujarat, Maharashtra, Madhya Pradesh, West Bengal and Tamil Nadu (9, 14). Sodhi et al. (22) reported a seroprevalence of 6.64% in Punjab State with a higher prevalence in exotic breeds than in indigenous sheep. Bandopadhyay and Mullick (2) made similar observations. A seroprevalence of 13.76% and 7.10% was recorded in sheep and goats in Kerala State, though clinical disease was not noted (18). The investigations of the authors have demonstrated BTV antibodies in 23% of native cattle and 71.9% of exotic cattle in Andhra Pradesh. Oberoi et al. (16) reported BTV antibodies in 37.5% of buffalo and 70% of cattle sera in Punjab. In Gujarat, 13.4% of buffalo and 15.6% cattle sera were positive for BTV antibodies (24). Jain et al. (8) noted the incidence of BTV antibodies as higher in buffalo (10.6%) than in cattle (4.2%). Mehrotra and Shukla (14) tested cattle sera from Andhra Pradesh, Karnataka, Gujarat, Punjab, Orissa, Himachal Pradesh and West Bengal and reported 18% BTV antibody positive. Prasad et al. (17) performed a serological survey in Rajasthan and Haryana and reported that 33.33% goat sera were positive for BTV antibodies. These reports established the fact that BTV infection is present in cattle, buffalo and goats in India. Vectors Culicoides insects are the vectors of BTV. Of over species present worldwide, at least 39 have been reported to occur in India. Very few species of Culicoides have been demonstrated to be vectors for BTV, with the principal vectors varying geographically. Midges collected from Harayana, Punjab, Rajasthan and Himachal Pradesh were identified as C. oxystoma (3). Jain et al. (7) isolated BTV from midges, but vector speciation was not performed. C. imicola and C. oxystoma were found to be prevalent in Tamil Nadu. Details of vector species responsible for transmission of BTV in India are lacking. Virus-vector relationships also need to be analysed critically. Veterinaria Italiana, 40 (3),

46 Global situation Virus isolation Of the 24 serotypes of bluetongue viruses recognised internationally, 21 have been reported from India. Eleven of these serotypes were identified after virus isolation while 10 serotypes were presumed present, based on serology. BTV serotypes 3, 9, 15, 16 and 18 have been reported from sheep flocks of Pakistan (1). Reports on isolation of BTV from India commenced with Kulkarni and Kulkarni (11) in 1984 who isolated BTV serotypes 9 and 18. Jain et al. (6) later reported type 1, employing chicken embryos and BHK-21 cell cultures for virus isolation. Mehrotra et al. (13, 15) recovered BTV serotypes 3, 9, 16, 18 and 23 from sheep from Madhya Pradesh, Maharashtra, Tamil Nadu, Uttar Pradesh and Jammu and Kashmir. BTV-2 was reported by Sreenivasulu et al. (23) from sheep outbreaks in Andhra Pradesh. Deshmukh and Gujar (4) isolated BTV serotype 1 from Maharashtra. Table III summarises the detection of BTV serotypes in different Indian states. Table III Distribution of bluetongue serotypes in India State Species Virus isolation Serotype antibodies Tamil Nadu Sheep 3, 16, 23 1, 4-7, 11, 12, 14-17, 19, 20 Andhra Pradesh Sheep 2 4, 12-14, Cattle 6, 12 Karnataka Sheep 23 1, 2, 12, 16, 17, 20 Cattle 1, 14, 16 Maharashtra Sheep 1, 2-4, 8, 9, 16, 18 Gujarat Buffalo 1, 15, 17 Cattle 2, 12, 20 Madhya Pradesh Sheep 18 Uttar Pradesh Sheep 9, 18, 23 Haryana Sheep 1, 4 14 Cattle 1, 2, 8, 12, 16 Himachal Pradesh Sheep 3, 9, 16, 17 4 Jammu and Kashmir Sentinel studies Sheep 18 In 1993, sentinels were used to follow the circulation of BTV serotype 12 in Andhra Pradesh. Seroconversion was recorded in September and was associated with rainfall and increased Culicoides populations. Similarly, active circulation of BTV was detected in Himachal Pradesh, Punjab and Rajasthan between June and November (17). Vaccines It is evident that multiple BTV serotypes are circulating in this region and virulence characteristics need to be studied to identify the pathogenic serotypes. Most BTV serotypes have been reported from Maharashtra, Gujarat, Andhra Pradesh, Tamil Nadu and Harayana. However, data is incomplete because systematic studies have not been undertaken to elucidate the prevalence of serotypes in different states. In view of this, the Indian Council of Agricultural Research has considered it necessary to map the BTV serotypes circulating in different Indian states with a long-term objective of the production of suitable vaccines. Research was initiated in seven states where BT is prevalent, whilst work on BTV vaccines incorporating endemic serotypes has commenced. Accordingly, a hydroxyl amine inactivated BTV-2 vaccine has been developed and is presently under field evaluation. References 1. Akhtar S., Howe R.R., Jadoon J.K. & Naqvi M.A. (1995). Prevalence of five serotypes of bluetongue virus in a Rambouillet sheep flock in Pakistan. Vet. Rec., 136, Bandopadhyay S.K. & Mullick B.B. (1983). Serological prevalence of bluetongue antibodies in India. Ind. J. Anim. Sci., 53, Bhatnagar P., Prasad G., Kakkar N.K., Das Gupta S.K., Rajpurohit B.S. & Srivastava R.N. (1997). A potential vector of bluetongue virus in northwestern India. Ind. J. Anim. Sci., 67, Deshmukh V.V. & Gujar M.B. (1999). Isolation and adaptation of bluetongue virus in cell culture systems. Ind. J. Comp. Microbiol. Immunol. Infec. Dis., 20, Harbola P.C., Chaudhary P.G., Lal Krishna, Siriguppi B.S. & Kole R.S. (1982). Incidence of bluetongue in sheep in Maharashtra. Ind. J. Comp. Microbiol. Immunol. Infec. Dis., 3, Jain N.C., Sharma R. & Prasad G. (1986). Isolation of bluetongue virus from sheep in India. Vet. Rec., 119, Jain N.C., Prasad G., Gupta Y. & Mahajan N.K. (1988). Isolation of bluetongue virus from Culicoides species in India. Rev. Sci. Tech. Off. Int. Épiz., 7, Jain N.C., Gupta Y. & Prasad G. (1992). Bluetongue virus antibodies in buffaloes and cattle in Haryana State of India. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International 76 Veterinaria Italiana, 40 (3), 2004

47 Global situation Symposium, Paris, June CRC Press, Boca Raton, Janakiraman D., Venugopalan A.T., Ramaswamy V. & Venkatesan R.A. (1991). Serodiagnostic evidence of prevalence of bluetongue virus serotypes among sheep and goats in Tamil Nadu. Ind. J. Anim. Sci., 61, Kulkarni D. D., Bannalikar A.S., Karpe A.G., Gujar M.B. & Kulkarni M.N. (1992). Epidemiological observations on bluetongue in sheep in Marathawada region of Maharashtra State in India. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Kulkarni D.D. & Kulkarni M.N. (1984). Isolation of bluetongue virus from sheep. Ind. J. Comp. Microbiol. Immunol. Infec. Dis., 5, Mahajan N.K., Prasad G., Jain N.C., Dhanoea J.S. & Gupta Y. (1991). Epizootiological studies on bluetongue at an organised sheep breeding farm near Hissar, Haryana. Ind. J. Anim. Sci., 61, Mehrotra M.L., Shukla D.C. & Kataria (1989). Isolation and identification of viral agent resembling bluetongue virus in an outbreak in sheep. Ind. J. Anim. Sci., 59, Mehrotra M.L.& Shukla D.C. (1990). Seroprevalence, diagnosis and differential diagnosis of bluetongue virus disease in India. Ind. J. Virol., 6, Mehrotra M.L., Shukla D.C. & Khanna P.N. (1996). Studies on bluetongue disease in India isolation and serotyping of field isolates. Ind. J. Comp. Microbiol. Immunol. Infec. Dis., 17, Oberoi M.S., Singh G. & Kwatra M.S. (1988). Serological evidence of bluetongue virus activity in cattle and buffalo populations. Ind. J. Virol., 4, Prasad G., Jain N.C., Mahajan N.K. & Vasudevan B. (1987). Prevalence of bluetongue precipitating antibodies in different domestic animals. Ind. J. Anim. Sci., 57, Ravi Sankar (2003). Seroprevalence of bluetongue in sheep and goats in Kerala. MVSc thesis submitted to Kerala Agricultural University. 19. Sapre S.N. (1964). An outbreak of bluetongue in goats and sheep. Ind. Vet. Rev., 15, Saravanabava K. (1992). Studies on bluetongue virus in sheep. PhD thesis, Tamil Nadu Veterinary and Animal Sciences University, Madras, India. 21. Sharma M.M., Lonkar P.S., Sreevastava C.P., Dubey S.C., Maru A. & Kalra D.B. (1985). Epidemiology of bluetongue in sheep at an organised farm in semi-arid part of Rajasthan, India. Ind. J. Comp. Microbiol. Immunol. Infec. Dis., 6, Sodhi S.S., Oberoi M.S., Sharma S.N. & Baxi K.K. (1981). Prevalence of bluetongue precipitating antibodies in sheep and goats in Punjab, India. Zentbl. Vet. Med., 28, Sreenivasulu D., Subba Rao M.V. & Gard G.P. (1999). Isolation of bluetongue virus serotype 2 from native sheep in India. Vet. Rec., 144, Tongaonkar S.S., Ayangar S.K., Singh B.K. & Kant R. (1983). Seroprevalence of bluetongue virus in Indian buffalo (Bubalus bubalis). Vet. Rec., 112, Uppal P.K. & Vasudevan B. (1980). Occurrence of bluetongue in India. Ind. J. Comp. Microbiol. Immunol. Infec. Dis., 1, Veterinaria Italiana, 40 (3),

48 Global situation Vet. Ital., 40 (3), Epidemiological observations on bluetongue in sheep and cattle in Japan Y. Goto (1), O. Yamaguchi (2) & M. Kubo (1) (1) Department of Infectious Diseases, National Institute of Animal Health, 1-5 Kannondai 3-chome, Tsukuba-shi, Ibarakiken, 305 Japan (2) Tochigi Kenhoku Livestock Hygiene Service Center, Nasu-gun, Nishinasuno-cho, Midori 12-14, Japan Summary Bluetongue (BT) first occurred in Japan between late August and October 1994 in 23 cattle in three prefectures of the northern Kanto region, and between the end of October and mid-november in 23 Suffolk sheep in the same region. The affected cattle had fever, deglutitive difficulty, hypersalivation, facial oedema, scabbing of the corner of the mouth and dysphagia. The BT virus (BTV) was isolated from blood cells of the affected sheep. Surveillance for BTV antibody conducted by prefectures in the affected region has detected seroconversion to BTV in some prefectures every year thereafter. In the autumn of 2001, again in the northern Kanto region, 45 sheep developed BT, and BTV was isolated. It is considered important that Japan has imported numerous cattle from Australia, the United States of America (USA), and New Zealand every year. In particular, BTV was isolated from cattle imported from the USA during quarantine although some of the serotypes isolated are not present in the USA. Furthermore, BTV is not present in New Zealand. The third RNA segment encoding the serogroup-specific VP3 protein of Japanese BTV isolates and reverse transcriptase-polymerase chain reaction (RT-PCR) positive blood cells was amplified by RT-PCR. Molecular phylogenetic analysis of the third RNA segment based on the sequence homology of the PCR products led to the classification of Japanese BTV isolates into two major groups. Keywords Bluetongue Cattle Epidemiology Japan Molecular epidemiology Sequencing Sheep. Introduction Seroepidemiological surveillance conducted in 1974 by Miura et al. (4) showed that domestic cattle in the Kyushu and Okinawa regions of Japan had antibodies to BTV serotypes 1, 12 and 20. In July and August 1979, cattle seroconverting to these serotypes were also confirmed on Miyakojima Island in the Okinawa Prefecture. Since the later national arbovirus surveillance conducted in Japan detected seroconversion in sentinel domestic cattle in all prefectures in the Kyushu and Okinawa Prefectures, subclinical infection with BTV appears to be prevalent in these regions in the absence of obvious bluetongue (BT) disease. Materials and methods Collection of blood from cattle and sheep in epidemic areas Heparinised blood and serum for antibody detection were collected from 170 sheep including those that developed BT. Blood was centrifuged at rpm at 4 C for 10 min, and separated into plasma and blood cells to isolate viruses. Blood cells were washed with 4 C phosphate-buffered saline (PBS) three times. Viruses were isolated using HmLu-1 cells derived from the baby hamster lung (5). In addition, blood cell suspensions were inoculated into the veins of 11-day-old embryonated chicken eggs, and chicken embryos and embryonic organs of eggs that died were inoculated onto Vero cells for virus isolation. 78 Veterinaria Italiana, 40 (3), 2004

49 Global situation Pathology Affected cattle and sheep that died were autopsied, and organs were examined histologically using routine techniques. Seroepidemiological survey Antibody surveys in areas of BT outbreaks, and a nationwide antibody survey in sentinel cattle, were performed by the neutralisation test in microplates as previously reported (4). The agar gel immunodiffusion (AGID) test for group-reactive antibodies was performed in accordance with the OIE Manual. Polymerase chain reaction and phylogenetic analyses Reverse transcriptase-polymerase chain reaction (RT-PCR) was performed as described by McColl et al. (2, 3). The third RNA segment (RNA3) encoding the serogroup-specific VP3 protein of Japanese BTV isolates and RT-PCR-positive blood cells was amplified by RT-PCR. Based on the sequence homology of the PCR products, a molecular phylogenetic tree of RNA 3 was constructed by the unweighted pair-group method using arithmetic means (7). Results Increase in imported cattle and isolation of BTV In recent years, Japan has imported numerous cattle from the USA and Australia, and the number of cattle imported is on the increase (Table I). Thus, BTV has been isolated frequently from the blood of imported cattle during quarantine. In particular, from 121 Holstein cattle imported from the USA in September 1990, 33 strains of BTV were isolated; from 89 cattle imported in September 1995, 13 strains of BTV and 18 strains of epizootic haemorrhagic disease virus (EHDV); and from 123 cattle imported in November 1995, five strains of BTV and three strains of EHDV. The BTV isolates were of serotypes 4, 11, 13 and 20. Table I Inspected cattle and articles, by exporting country Exporting country United States of America Australia Canada New Zealand Other Importantly, however, BTV serotypes 4 and 20 do not occur in the USA. The EHDV isolates were of serotypes 1 and 2. In contrast, no BTV has been directly isolated from the cattle imported from Australia. However, 7 strains of BTV were isolated from Culicoides brevitarsis collected in the quarantine house for imported cattle at the Okinawa Branch of the Animal Quarantine Station. Isolation of BTV from domestic cattle and sheep in Japan Although BTV has been isolated from the blood of domestic cattle and C. brevitarsis since 1985, no outbreak of BT has been confirmed. Symptoms and histopathological findings in BT outbreaks in Japan and isolation of BTV Cattle Between August and October in 1994, a total of 23 Japanese Black beef cattle in three prefectures of the northern Kanto region developed BT with symptoms of fever, loss of appetite, hyper-salivation, facial oedema, scabbing of the corner of the mouth, and dysphagia. Histopathological examination of autopsy material from diseased cattle showed that the lesions were localised in the oesophagus, and consisted of hyaline degeneration and rupture of oesophageal striated muscle fibres, lymphocytic infiltration and regeneration of muscle fibres. No BTV was isolated from the blood of 199 cattle including the diseased cattle and other cattle on the same farm. Sheep A total of 23 sheep on a Suffolk sheep farm in the northern Kanto region developed BT between the end of October and mid-november Symptoms included fever, dysphagia, tongue ulceration and laminitis, which were similar to but slightly more severe than those in cattle. BTV was isolated from three of the diseased sheep and identified as serotype 21. Some 45 sheep on the same farm where sheep had developed BT in 1994 developed BT between late September and late November Breeding ewes and lambs were most affected and developed symptoms of fever, dysphagia, auricular oedema and laminitis. Five sheep died. Although the symptoms of the diseased sheep were similar to those in 1994, ulceration of the tongue and scabbing of the nasal arch were less severe. Serotype 21 was isolated from the blood of three diseased sheep. Veterinaria Italiana, 40 (3),

50 Global situation Seroepidemiology of BTV infection in Japan: yearly seroconversion rates in sentinel cattle in Japan Seroconversions in sentinel cattle throughout Japan between 1999 and 2002 are shown in Fig. 1. In western Japan in 1999, the seroconversion rate was approximately 30% in the Shimane and Hiroshima Prefectures, 10%-13% in the Okayama and Yamaguchi Prefectures, and as high as 71% and 36.4% in the Kumamoto (in Kyushu) and Okinawa Prefectures, respectively. The seroconversion rates in the Tochigi Prefecture (where an outbreak of BT in sheep was confirmed in 2001) and in the neighbouring Ibaraki and Fukushima Prefectures were 10.8% and 5%-6%, respectively. In 2001, BTV seroconversions occurred in numerous cattle in western Japan: the conversion rates in Okayama, Ehime and Kochi were 44.8%, 43.8%, and 36%, respectively, confirming an epidemic to have occurred in some regions. In contrast, approximately 30% of cattle have seroconverted every year between 1999 and 2002 in the Okinawa Prefecture. Japanese BTV isolates: PCR-based detection and molecular phylogeny RT-PCR-positive cattle were detected in Fukushima in 1994, in Saga and Miyazaki in 1995, in Kagoshima and Okinawa in 1996, and in Tottori in 1997 during annual arboviral surveillance. Based on nucleotide sequence homology, the PCR products from Japanese BTV isolates and RT-PCR positive blood cells of cattle were broadly divisible into the 1994 Percentage rates Percentage rates Percentage rates Percentage rates Figure 1 Seroconversion percentage rates of sentinel cattle in Japan, Veterinaria Italiana, 40 (3), 2004

51 Global situation epidemic group and another group (Fig. 2). Many BTVs isolated to date were identified as belonging to serotype 21. Discussion Although the epidemiological features of BT in Japan have not been well characterised, no cases of BT have occurred in Okinawa or other western Japan prefectures where BTV has been isolated (8). However, the results of seroepidemiological surveillance suggest that subclinical or mild BTV infection of ruminants in Japan is prevalent almost every year, and that there are repeated and recurrent infections among domestic cattle and sheep in the affected regions. The serotype most frequently isolated to date is BTV-21, but its pathogenicity is not clear. Figure 2 Phylogenetic tree based on sequence analysis of the amplified cdna fragment of segment 3 of the bluetongue virus strains Veterinaria Italiana, 40 (3),

52 Global situation In Japan, a BT outbreak first occurred in 1994 in the northern Kanto region (1) that had been regarded as a non-epidemic region. It is not clear why BT suddenly became prevalent in the mountainous area of the northern Kanto Region. Since BT developed in aged Japanese Black cattle, decreased immunological function and genetic factors cannot be excluded. The development of BT in sheep in the same region has attracted little attention because of the very small number of domestic sheep in Japan. The development of BT in cattle in a region, followed by development in domestic sheep in the same region, suggests that the virus spread from infected cattle to adjacent sheep. The second outbreak of BT among sheep on the same farm in 2001 might suggest that there had been persistent infection among sheep since the original outbreak in 1994, although an endemic cycle of infection or repeated incursions of BTV are also potential explanations. Although BTV has been isolated from imported cattle, the serotypes isolated have not always been endemic in the countries of export, suggesting infection after arrival in Japan; however, negotiations are needed with exporting countries over hygienic conditions including the quarantine system for imported cattle and sheep. The Ibaraki virus, which is a member of the EHDV group of the Orbivirus genus, incurs into western Japan every 5 to 10 years, and infected cattle develop a BT-like disease (6) with dysphagia due to paralysis of the oesophagus and pharynx; therefore, the differentiation of BT and similar diseases will become increasingly necessary in the future. Moreover, antigenically distinct strains of Ibaraki virus have caused many abortions in cattle, and their relationship with cattle dystocia has been suspected (9). References 1. Goto Y., Yamakawa M. & Miura Y. (1996). An outbreak of bluetongue in cattle in Japan. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, McColl K.A. & Gould A.R. (1991). Detection and characterization of bluetongue virus using the polymerase chain reaction. Virus Res., 21, McColl K.A. & Gould A.R. (1994). Bluetongue virus infection in sheep: haemorrhagic changes and detection by polymerase chain reaction. Aust. Vet. J., 71, Miura Y., Inaba Y., Tsuda T., Tokuhisa S., Sato K. & Akashi H. (1982). Seroepizootiological survey on bluetongue virus infection in cattle in Japan. Natl Inst. Anim. Hlth Q. (Jpn), 22, Okumura H. (1968). Spontaneous malignant transformation of hamster lung cells in tissue culture. In Cancer cells in culture (H. Katsuta, ed.). University of Tokyo Press, Tokyo, Omori T. (1970). Ibaraki disease: a bovine epizootic disease resembling bluetongue. Natl Inst. Anim. Hlth Q. (Jpn), 10, Saitou N. & Nel M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molec. Biol. Evol., 4, Takayoshi K., Hamakawa M., Higa H., Tohma S. & Kokuba T.(1990). A bluetongue virus isolated in Okinawa in Annual report of the Okinawa Prefectual Institute of Animal Health, 26, Uchinuno Y., Ito T., Goto Y., Miura Y., Ishibashi K., Itou T. & Sakai T. (2003). Differences in Ibaraki virus RNA segment three sequences from 3 epidemics. J. Vet. Med. Sci., 65 (11), Acknowledgements The authors wish to acknowledge the excellent technical assistance of Mrs Hiromi Kato. Dr Ito was supported by the National Institute of Animal Health, which also partly funded the molecular analyses of bluetongue viruses. 82 Veterinaria Italiana, 40 (3), 2004

53 Vet. Ital., 40 (3), Global situation Distribution of bluetongue in the United States of America, E.N. Ostlund (1), K.M. Moser (1), D.J. Johnson (1), J.E. Pearson (2) & B.J. Schmitt (1) (1) Diagnostic Virology Laboratory, National Veterinary Services Laboratories, Veterinary Services, APHIS, PO Box 844, Ames, IA 50010, United States of America (2) 4016 Phoenix, Ames, IA 50014, United States of America Summary Bluetongue virus (BTV) distribution in the United States of America (USA) is limited by the range of the vector Culicoides spp. Regional differences exist with the north-eastern states being free of BTV, while the central and north-western states are seasonally free of virus. Activity of the virus can be observed throughout the year in the southern USA. Serological evidence defining the distribution of BTV in selected regions of the USA is gathered regularly through serological surveys conducted on samples from slaughter cattle. From 1991 to 2002, ten serological surveys were completed. Results from Alaska, Hawaii, Michigan, Minnesota, New York, Wisconsin and New England consistently demonstrated a seropositive rate of less than 2%, confirming BTV-free status. Antibody against BTV was sporadically detected in cattle originating from states contiguous to the BTV-free regions. Additional information on BTV distribution in the USA is obtained through identification of BTV or BTV RNA in diagnostic, surveillance and export specimens submitted to the National Veterinary Services Laboratories. Results confirm that BTV serotypes 2, 10, 11, 13 and 17 are present in the USA. Keywords Bluetongue virus Culicoides Epizootic haemorrhagic disease virus Isolation Serotype Survey United States of America. Serological surveys To assess the distribution of bluetongue (BT) virus (BTV) in the United States of America (USA), a nationwide survey was conducted in the winter of 1977/1978 (4). In this comprehensive survey, market cattle samples from eighteen northern and northeastern states including Alaska and Hawaii had 1.0% seropositive samples in the BTV complement fixation test. Subsequently, 19 regional surveys incorporating subsets of the 50 states have been completed. Results of nine regional surveys conducted prior to 1991 were reported previously (6). Since 1991, ten BTV serological surveys utilising market cattle samples have been completed. The agar gel immunodiffusion (AGID) test supplanted the complement fixation test as the primary BTV antibody detection method for regional surveys completed between 1979 and 1992 and the BTV competitive-enzyme-linked immunosorbent assay (c-elisa) was used to test serum collected in surveys from 1993 to Both c-elisa and AGID are current OIE-prescribed BT serological methods for international trade (1). For each survey, samples were collected in late autumn and early winter in order to optimise detection of current year exposure to BTV. In this report, the surveys are described by calendar year of sample collection. For each survey, the majority of serum samples were collected in the specified calendar year; in some surveys a few samples were collected as late as mid- January of the subsequent year. At least 600 samples per region were tested for each survey. If all 600 samples tested negative, a 95% confidence level indicated that the true incidence rate for the region fell between 0.0% and 0.5%. Historically, a seropositive rate of less than 2% per region has been required to mitigate United States cattle export requirements for movement into Canada. Traditional states survey results ( ) Surveys of cattle from 18 traditionally BTV-free states comprising 11 regions were conducted annually from 1991 to Biennial surveys were conducted in 1998, 2000 and States included in all regional surveys were as follows: Connecticut Delaware Veterinaria Italiana, 40 (3),

54 Global situation Indiana Maine Maryland Massachusetts Michigan Minnesota New Hampshire New Jersey New York North Dakota Ohio Pennsylvania Rhode Island Vermont West Virginia Wisconsin. Regions consisted of a single state with the exceptions of New England (Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island and Vermont) and the combinations of Maryland/Delaware and Pennsylvania/New Jersey. Ohio and West Virginia were sampled as separate regions for surveys from 1991 to 1996 but were combined into a single region for surveys from 1998 to An abbreviated survey was conducted in 1997 and included only six traditionally surveyed states comprising four regions namely: Indiana, Ohio, Maryland/Delaware and Pennsylvania/New Jersey. For five regions, i.e. New England, Michigan, Minnesota, New Jersey and Wisconsin, BTV seropositive samples did not exceed 2.0% in any survey. The numbers of BTV antibody-positive samples for all surveys from 1991 to 2002 were as follows: New England 28/8 808 (0.3%) Michigan 16/5 999 (0.3%) Minnesota 40/6 298 (0.6%) New Jersey 15/6 912 (0.2%) Wisconsin 24/6 028 (0.4%). For surveys ( ) in which West Virginia was evaluated as a region, the seropositive rate was less than 2.0% each year with an overall average of 0.7%. The New Jersey/Pennsylvania region samples averaged 1.2% seropositive in surveys conducted from 1991 to 2002 and fell below 2.0% positive samples in all surveys, except in In 1996, this region revealed 2.2% seropositive samples. For seven of nine surveys that included North Dakota (1991, , 1998 and 2000), the seropositive rate fell below 2.0%. The North Dakota region samples exceeded 2.0% positive only in 1992 (2.7%) and 2002 (5.7%). The region of Delaware/Maryland had less than 2% positive samples in and , but exceeded 2.0% positive samples in 1996 (2.5%). In 2000 and 2002, insufficient numbers of samples were collected from Delaware/Maryland for inclusion in the survey data. For surveys ( ) where Ohio was considered a region, Ohio samples exceeded 2.0% seropositive twice with 2.2% in 1992 and 2.6% in When Ohio/West Virginia was examined as a region, the seropositive rate fell below 2.0% in 1998 and 2002 but was 3.6% in The BTV seropositive rate for samples from Indiana was less than 2% for seven of ten surveys but exceeded 2.0% in 1992, 1993 and 1996, with 4.2%, 2.3% and 2.5% seropositive rates, respectively. Table I illustrates the seropositive rates per survey for the traditional regions. Samples from non-traditional states ( ) In addition to the traditional regions surveyed, samples from additional states were included in some of the serological surveys of market cattle. Samples from Alaska and Hawaii were examined in annual surveys from 1991 to 1996 and BTV antibody testing revealed negligible seropositive rates ( %) for these regions. Virginia was included in surveys from 1991 to 1993, Illinois in 1992 and 2002, eastern and western Washington in 1996, South Dakota in 2000 and 2002, Oregon in 2000 and Idaho, Iowa, Montana and Wyoming in Samples from western Washington had a 1.3% BTV seropositive rate in the 1996 survey. However, samples from Iowa, Idaho, Illinois, Montana, Oregon, South Dakota, Virginia, eastern Washington and Wyoming exceeded 2.0% seropositive in all surveys in which they were included. Figure 1 illustrates the composite results of BTV antibody detection in regional surveys conducted from 1991 to Bluetongue virus solation and bluetongue virus RNA detection in animals in the United States of America ( ) Disease due to BTV infection is observed only sporadically in domestic animals in the USA. The OIE-prescribed BTV isolation method of intravenous inoculation of embryonating chicken eggs followed by cell culture passage (1) is lengthy, requiring 4 to 5 weeks, and is limited to facilities that have appropriate resources and personnel with expertise in the procedures. Consequently, few diagnostic laboratories in the USA attempt BTV isolation. The nested polymerase chain reaction 84 Veterinaria Italiana, 40 (3), 2004

55 Global situation Table I Percentage of bluetongue virus antibody positive market cattle samples from selected regions of the United States of America, Region Delaware/Maryland * * Indiana Michigan Minnesota New England North Dakota New Jersey/Pennsylvania New York Ohio West Virginia Ohio/West Virginia Wisconsin not done quantity not sufficient Bluetongue virus antibody seroprevalence rate <2% Bluetongue virus antibody seroprevalence rate <2% in some years and >2% in other years Bluetongue virus antibody seroprevalence rate >2% in each survey in which samples were tested Not included in any serological survey between 1991 and 2002 Figure 1 Results of regional market cattle surveys conducted for bluetongue between 1991 and 2002 Not all regions were included in each survey (PCR) was first listed as an OIE prescribed test in Over the last decade, requirements for testing to certify animals or animal products free of BTV for the purposes of interstate and international movement have shifted somewhat from virus isolation towards more expedient and sensitive PCR methods. The United States Department of Agriculture National Veterinary Services Laboratories (NVSL) performs diagnostic, surveillance and export testing to detect BTV. Results of NVSL testing from 1991 to 2002 are reported here. Submissions originated from domestic ruminants, native wild ruminants and cervids, and zoo species. OIE-prescribed virus Veterinaria Italiana, 40 (3),

56 Global situation isolation and PCR methods were employed. BTV isolates were subjected to further testing by PCR (2) or virus neutralisation to determine the serotype. When PCR was the only method identifying BTV nucleic acid in a sample, serotype identification was not performed. In all years from 1991 to 2002, BTV was isolated from blood or tissues submitted to the NVSL. Species from which BTV isolates were obtained include bighorn sheep, bison, cow, elk, goat, domestic sheep, Persian gazelle and white-tailed deer. Additional species in which BTV was detected only by PCR were gerenuk, mule deer and pampas deer. BTV serotypes 17 and 11 were isolated most frequently during this period followed by BTV serotypes 13 and 10. Although there had not been evidence of BTV serotype 2 activity since 1986 (6), it was isolated from sheep residing in Florida in Figure 2 illustrates the states from which BTV was detected by PCR and/or virus isolation and shows the distribution of BTV serotypes observed in submissions to the NVSL for the years Epizootic haemorrhagic disease Owing to similarities in transmission and clinical presentation, epizootic haemorrhagic disease virus (EHDV) infection is frequently considered among differential diagnoses with BTV. Virus isolation in cell culture and PCR were used to identify EHDV in specimens submitted to the NVSL (5, 7). Figure 3 illustrates EHDV isolate distribution data for NVSL submissions from 1991 to EHDV serotypes 1 and 2 exist in the USA. From 1991 to 2002, isolation of EHDV serotype 2 was more common than isolation of EHDV serotype 1. EHDV serotype 2 was isolated from bighorn sheep, cattle, mule deer and white-tailed deer during this period. EHDV serotype 1 was isolated from deer (species not specified) in California and white-tailed deer in New Jersey. EHDV was also detected by PCR in bovine samples from Illinois and Florida but the serotype was not determined. Although EHDV infections are often subclinical in domestic species, EHDV contributed to morbidity in cattle from the midwestern portion of the USA during the late 1990s (3). BTV isolate/pcr-positive samples submitted to the NVSL between 1991 and 2002 BTV serotype identified Figure 2 Bluetongue virus (BTV) isolation and bluetongue virus RNA identification between 1991 and 2002 States from which BTV isolates and PCR positive samples were identified from samples submitted to the NVSL between 1991 and 2002 Source: National Veterinary Services Laboratory, Ames, Iowa 86 Veterinaria Italiana, 40 (3), 2004

57 Global situation EHDV-1 EHDV-2 EHDV PCR-positive (serotype not identified) Figure 3 Epizootic haemorrhagic disease virus (EHDV) isolation and RNA identification between 1991 and 2002 States from which EHDV isolates and PCR-positive samples were identified from samples submitted to the NVSL between 1991 and 2002 Source: National Veterinary Services Laboratory, Ames, Iowa Conclusions Taken together, the results of regional serosurveys and virus isolation/identification help define the distribution of BTV in the USA. Based on serosurveys conducted between 1977 and 1997, the states of Alaska and Hawaii are BTV-free. In all 19 market cattle surveys conducted since 1977, cattle from the geographic region of New England and the states of Michigan, Minnesota, New York and Wisconsin have had low (less than 2%) seroprevalence for BTV antibodies. BTV isolation has not been reported from animals in these regions. These northern and north-eastern states comprise a BTV-free zone of the continental USA. Animals from regions bordering the BTV-free zone are customarily seronegative for BTV antibodies. The regions of Delaware/Maryland, Ohio/West Virginia, Pennsylvania/New Jersey, Indiana and North Dakota have had less than 2% BTV antibodypositive samples in the majority of market cattle surveys. No BTV isolation from animals in these states was obtained at the NVSL in the past decade although BTV RNA was detected in one sheep from Maryland in Agent identification and serological survey results for remaining states indicate presence of BTV, at least seasonally. Since 1991, all five known USA serotypes of BTV (types 2, 10, 11, 13 and 17) and both EHDV serotypes 1 and 2 have been isolated, documenting that these viruses are present in the USA. The distribution of EHDV is similar to that of BTV. No additional serotypes of either virus have been identified in animals in the USA. References 1. Eaton B. (2000). Bluetongue. In Manual of standards for diagnostic tests and vaccines, 4th Ed. Office International des Épizooties, Paris, Johnson D.J., Wilson W.C. & Paul P.S. (2000). Validation of a reverse transcriptase multiplex PCR test for the serotype determination of U.S. isolates of bluetongue virus. Vet. Microbiol., 76, Loiacono C.M., Turnquist S.E., Ostlund E.N., Johnson G.E., Pace L.W., Turk J.R., Miller M.A., Kreeger J.M., Ramos J.A. & Gosser H.S. (1999). An outbreak of epizootic hemorrhagic disease in Missouri cattle. Proc. Am. Assoc. Vet. Lab. Diagn., 42, Metcalf H.E., Pearson J.E. & Kingsporn A.L. (1981). Bluetongue in cattle: a serologic survey of slaughter cattle in the United States. Am. J. Vet. Res., 42, Veterinaria Italiana, 40 (3),

58 Global situation 5. Pearson J.E., Gustafson G.A., Shafer A.L. & Alstad A.D. (1992). Diagnosis of bluetongue and epizootic hemorrhagic disease. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Pearson J.E., Gustafson G.A., Shafer A.L., & Alstad A.D. (1992). Distribution of bluetongue in the United States. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Wilson W.C. (1994). Development of a nested-pcr test based on sequence analysis of epizootic haemorrhagic disease viruses non-structural protein 1 (NS1). Virus Res., 31, Veterinaria Italiana, 40 (3), 2004

59 Vet. Ital., 40 (3), Global situation Bluetongue virus in South America: overview of viruses, vectors, surveillance and unique features I.A. Lager Instituto de Virología, INTA-Castelar, CP 1712, Hurlingham, Buenos Aires, Argentina Summary Since the first published report of bluetongue (BT) virus (BTV) infection in South America from Brazil in 1978, serological surveys have determined that the infection is widespread in sheep, cattle, goats and water buffalo but generally without clinical signs. Only four outbreaks of BT disease have been reported so far in Brazil. Brazil and Argentina are the only countries in South America where BTV serotypes 12 and 4 have been isolated, respectively. By serology, serotypes 4, 6, 14, 17, 19 and 20 were detected in Brazil, 12, 14 and 17 in Colombia, 14 and 17 in Guyana and 6, 14 and 17 in Suriname. Culicoides insignis is the predominant vector in the area, but C. pusillus could also be a BTV vector. The virus has not yet been isolated from the vector in the region. Keywords Bluetongue Culicoides Serotypes South America Virus. Distribution of bluetongue virus infection in South America The global distribution of bluetongue (BT) virus (BTV) historically has been shown to be between latitudes of approximately 40 N and 35 S (25). Eleven of the thirteen South American countries have their territory entirely in that area. As in other parts of the world, the livestock population in all these countries lying in the tropics and subtropics are considered BTV-infected (10). Parts of Chile and Argentina are south of that area. Chile detected seropositive bovines and ovines in 1985 (30), but recent information indicates that BTV infection is not present (33). In Argentina two serological surveys were conducted to determine the distribution of the BTV infection. Climatic factors, such as temperature and precipitation, were considered for the sampling design. In the first survey, bovine serum samples from 602 farms corresponding to six north-east provinces (Chaco, Formosa, Santa Fé, Entre Ríos, Corrientes and Misiones) were analysed between 1995 and The results indicated that Formosa, Santa Fé and Entre Ríos were seronegative while Misiones had 539 positive samples from a total of and the Departments of Santo Tomé and Ituzaingó of Corrientes revealed that 11 bovines possessed antibodies from a total of In 1998, a second serological survey was performed, examining bovine and 746 ovine samples from Chaco, Formosa, Santa Fé, Entre Rios, Corrientes and Misiones. Again, only the Misiones Province had a prevalence of antibodies of 125/248 in bovines and 19/20 in ovines, and the same Departments of Corrientes Province had 8/295 in bovines and 13/405 in ovines. Similar data was obtained in Brazil where it was found that in the Rio Grande do Sul State there were few positive animals compared with Paraná, Santa Catarina and Sao Pablo States (6). Uruguay, which is close to the seronegative area of Argentina, also considers its territory free of BTV infection. These factors suggest that the southern extent of BTV infection in South America could be further north than the parallel 35 S or that the distribution of the infection has a particular pattern like that in Australia. For this reason, it is important to improve the surveillance and monitoring work in the region. Susceptible species infected and seroprevalence The first published report of BTV infection in South America was from Brazil in 1978 (5). Serological evidence of infection was detected in livestock in the Veterinaria Italiana, 40 (3),

60 Global situation States of Sao Paulo and Rio de Janeiro (21). Since then, several serological surveys have determined that the infection is widespread in South America but generally without overt disease (1, 2, 5, 6, 7, 8, 9, 10, 11, 15, 16, 17, 19, 21, 22, 23, 29, 30). In most of these serological surveys the agar gel immunodiffusion (AGID) technique was used. It is very likely that some BTV-seropositive animals were actually infected with related orbiviruses, such as epizootic haemorrhagic disease virus, since the BTV antigen is group-specific and cross-reactions are common with AGID. The domestic species involved in these surveys were bovines (2, 5, 6, 7, 10, 11, 13, 15, 16, 17, 21, 22, 30, 32), ovines (5, 8, 11, 15, 29), water buffalo (19, 32) and caprines (1, 5, 8, 9, 11). The percentages of antibody prevalence of these surveys showed a wide range even in the same country (Table I). Table I Bluetongue virus serological surveys in countries of South America Country Location No. of samples/species/ technique Seroprevalence (%) Ref. Argentina Misiones / bov/ AGID 40.7 * 248 / bov / AGID ** 20 / ov / AGID 95 ** Corrientes 1528 / bov / AGID 0.7 * 295 / bov / AGID 2.7 ** 415 / ov / AGID 3.13 ** 93 / bov / c-elisa (13) Chaco, Formosa / bov / AGID 0 * Entre Ríos 956 / bov / AGID 0 ** Santa Fé 311 / ov / AGID 0 ** Brazil Paraiba 137 / bov /AGID 4.82 (22) Sergipe 97/ bov / AGID (23) Mina Gerais 410 / bov / AGID 76.3 (2) 340 / cap / AGID 5.9 (9) 329 / buf / AGID 54.4 (19) Rio de Janeiro 553 / bov / AGID (7) 626 / cap / AGID (8) 66 / ov / AGID (8) Sao Pablo 214 / bov / AGID (6) Paraná 106 / bov / AGID (6) Sta Catarina 174 / bov / AGID (6) Rio Grande do Sul 409 / bov / AGID 1.22 (6) Chile X Region 1752 / bov / AGID 19.6 (30) 434 / bov / NS 0 (35) NS / bov / NS 0 (35) / ov / NS 0 (35) Colombia Antioquia, Cordoba 635 / bov / AGID 51.8 (16) Valle de Aburra 86 / bov / AGID 56 (17) Ecuador El Oro 87/ bov / AGID 10 (21) Guyana Diverse areas 719 / bov / AGID 56 (11) 387 / bov / AGID 50 (11) 255 / cap / AGID 40 (11) Rupununi 50 / bov / AGID 8 (11) 25 / ov / AGID 0 (11) 25 / cap / AGID 4 (11) Peru North 8 / ov / AGID 87.5 (29) Central 17/ ov / AGID 41 (29) South 9 / ov / AGID 55.5 (29) Suriname Diverse areas 451/ bov/ AGID 82 (11) 77/ ov / AGID 88 (11) 68 / cap / AGID 91 (11) Venezuela Aragua 151 / bov / AGID 74.8 (10) 151/ bov / c-elisa 94.7 (10) * survey (data not published) ** 1998 survey (data not published) NS not specified 90 Veterinaria Italiana, 40 (3), 2004

61 Global situation This could be due to differences in climatic and environmental factors that affect the distribution of the vector/s and/or the susceptible host (6, 11, 12). The prevalence of antibodies is not consistently high in any one species or country. Most of these data were obtained in the 1980s, and in some countries they are the only available data so they must be only considered as indicative. There have been changes in climatic factors and in land use affecting the geographical distribution of ruminants that could have modified those values. Other species have been analysed for the presence of antibodies as possible hosts or reservoirs. In Argentina, free-ranging llamas, guanacos, vicuñas and Pampean deer were negative (18, 20, 26, 35) but Peru found that alpaca could be infected (27). In 1991, cervids in the Rio de Janeiro Zoo were affected by a haemorrhagic disease and, in January-February 1992, the disease was described in a herd of brown brocket (Mazama gouazoubira) on the campus of the University of São Paulo State (UNESP), Brazil. One of four brockets died. Again in this institution, six brown brockets died in Serological studies by AGID indicated antibodies against BTV or a related orbivirus (4). In July 1992, the disease was documented in one specimen from a marsh deer (Blastocerus dichotomus) in the Ilha Solteira Zoo in Brazil (4). Clinical disease Only four outbreaks of BT have been reported to date in South America, all of which occurred in the last three years in Párana State, Brazil. The first was in April 2001, on a mixed farm in the area of Curitiba (Table II). The affected animals were eight sheep and one goat that had severe and acute disease. The clinical signs in the sheep included temperature, depression, hyperaemia of the oral cavity, facial oedema especially on the lips, tongue, muzzle and submandibular space. Table II One outbreak of bluetongue 04/2001, Paraná, Brazil Species Number of animals Susceptible Cases Deaths Bovine Caprine Ovine Source: OIE, Handistatus II Necrosis of the epithelium of the nose and tongue, hyperaemia and pettechial haemorrhages of the pharynx, oesophagus, ruminal and omasal mucosa were also present (3). In February 2002, the second outbreak in goats took place and the last two occurred in March 2002 affecting sheep and goats (Tables III and IV). These recent outbreaks might suggest changes in the virulence of the local serotypes/strains of BTV, introduction of new serotypes or strains, movement of susceptible animals due to livestock trade, or factors that favour proliferation of the vector/s. Table III One outbreak of bluetongue 02/2002, Paraná, Brazil Number of animals Species Susceptible Cases Deaths Caprine Source: OIE, Handistatus II Table IV Two outbreaks of bluetongue 03/2002, Paraná, Brazil Species Number of animals Susceptible Cases Deaths Caprine Ovine Source: OIE, Handistatus II Bluetongue virus isolation The first BTV isolation from naturally infected animals in South America was serotype 4 from Zebu cattle that were imported from Brazil to the USA (14). Brazil and Argentina are the only countries in South America where BTV has been isolated. Isolation of the virus in Brazil was made by Panaftosa from blood and tissue samples of clinically affected sheep and goats during the April 2001 outbreak, and was confirmed by reverse transcriptase-polymerase chain reaction (RT-PCR) and typed as serotype 12 by virus neutralisation (VN) (3). In Argentina, BTV was isolated at the National Institute of Agricultural Technology (INTA)- Castelar from the blood of sentinel cattle without clinical signs and serotyping was confirmed by the Institute for Animal Health in Pirbright, as type 4 by VN and RT-PCR with primers corresponding to segment 2 of the BTV genome (13, 34). Detection of serotypes by serology Using serological techniques, the serotypes that may be present in South America are: 4, 6, 14, 17, 19 and 20 in Brazil (5, 14); 12, 14 and 17 in Colombia (17); 14 and 17 in Guyana and 6, 14 and 17 in Suriname (15). These results should be considered as preliminary because of the serological cross-reactions among BTV serotypes. No other reports are available from the other countries in the region, indicating that information on the serotypes present in South America is very limited and not recent. Veterinaria Italiana, 40 (3),

62 Global situation Vectors Very little information is available about the vector/s involved in the transmission of BTV in South America. C. insignis is possibly the predominant vector in the area (6, 13, 17, 28). BTV has been isolated from C. insignis in Central America and the Caribbean (24) and also has been shown capable of the transmission of BTV in southern Florida (31). However, as BTV was isolated from C. pusillus in Central America and the Caribbean and this species is also present in South America, it is possible that C. pusillus could be a BTV vector in the area (24, 28). The information available in South America is insufficient to exclude the possibility that additional species of Culicoides may also transmit BTV. The virus has not yet been isolated from the potential vectors in the region. Present and future studies Argentina is conducting an epidemiological surveillance programme for BTV. This programme is conducted by a BTV working team of professionals from SENASA (National Animal Health Service) and the BTV Argentina working team SENASA-INTA, namely: S. Duffy, J. Miquet, A. Vagnozzi, C. Gorchs, G. Draghi, B. Cetrá, C. Soni, V. Ramirez, N. Pacienza and M. Ronderos. A sentinel animal monitoring project was conducted from 1999 to 2001 in the Santo Tomé and Ituzaingó Departments of the Corrientes Province. The conclusions were that though clinical disease has never been reported in Argentina, viral activity was present and the BTV strain isolated was serotype 4. In addition, a marked variability in the cumulative incidence of BTV infection among herds and between years was detected, with absence of BTV activity from May to September. C. insignis was the predominant potential vector species detected. A new project has started and, as was the case previously, it is supported by SECYT (Science and Technology Secretary). This investigation will provide information on the incidence of BTV infection in sheep and cattle, the seasonal incidence of the virus and vector/s, the serotypes of the BTV isolated and the Culicoides species that can potentially serve as vectors in BTV-infected regions of Argentina. The introduction of positive animals in the areas free of infection will be monitored to determine if they can be a source of infection for the native animals and for how long they remain seropositive. Therefore, although there is work to be done, awareness of the problem is increasing and measures are being taken to improve the knowledge of the epidemiological situation of BTV infection in South America. Acknowlegements The author is grateful to Zelia Lobato, Jorge Lopez and Natalia Pacienza for providing references and to Laura Weber, Mariano Ramos and Jorge Miquet for their editorial comments and review of this manuscript. References 1. Brown C.C., Olander H.J., Castro A.E. & Behymer D.E. (1989). Prevalence of antibodies in goats in north-eastern Brazil to selected viral and bacterial agents. Trop. Anim. Hlth Prod., 21, Castro R.S., Leite R.C., Abreu J.J., Lage A.P., Ferraz I.B., Lobato Z.I.P. & Balsamao S.L.E. (1992). Prevalence of antibodies to selected viruses in bovine embryo donors and recipients from Brazil, and its implications in international embryo trade. Trop. Anim. Hlth Prod., 24, Clavijo A., Sepulveda L., Riva J., Pessoa-Silva M., Tailor-Ruthes A. & Lopez J.W. (2002). Isolation of bluetongue virus serotype 12 from an outbreak of the disease in South America. Vet. Rec., 151 (10), Cubas Z.S. (1996). Special challenges of maintaining wild animals in captivity in South America. Rev. Sci. Tech. Off. Int. Épiz., 15 (1), Cunha R.G. (1990). Neutralizing antibodies for different serotypes of bluetongue virus in sera of domestic ruminants from Brazil. Rev. Bras. Med. Vet., 12, Cunha R.G., de Souza D.M. & Texeira A.C. (1982). Anticorpos precipitantes para o virus da lingua azul em soros de bovinos do estado do Rio de Janeiro. Biológico, São Paulo, 48 (4), Cunha R.G., de Souza D.M. & da Silva Passos W. (1987). Antibodies to bluetongue virus in sera from cattle of São Paulo s state and Brazil s south region. Rev. Bras. Med. Vet., 9 (6), Cunha R.G., de Souza D.M. & Teixeira A.C. (1988). Incidencia de anticorpos para o virus da lingua azul em soros de caprinos e ovinos do estado do Rio de Janeiro. Arq. Flum. Med. Vet., 3 (2), Da Silva J., Modena C., Moreira E., Machado T., Viana F. & Abreu V. (1988). Frequency of foot and mouth disease, bluetongue and enzootic bovine leucosis in goats in the state of Minas Gerais. Brazil. Arq. Bras. Med. Vet. Zoot., 40 (6), Gonzalez M.C., Perez N. & Siger J. (2000). Serologic evidence of bluetongue virus in bovines from Aragua State, Venezuela. Rev. Fac. Cs Vets UCV, 41 (1-3), Gibbs J., Greiner E., Alexander F., King H. & Roach C. (1984). Serological survey of ruminant livestock in some countries of the Caribbean region and South America for antibody to bluetongue virus. Vet. Rec., 114 (26), Veterinaria Italiana, 40 (3), 2004

63 Global situation 12. Gibbs J. & Greiner E. (1994). The epidemiology of bluetongue. Comp. Immun. Microbiol. Infec. Dis., 17, Gorsch C., Vagnozzi A., Duffy S., Miquet J., Pacheco J., Bolondi A., Draghi G., Cetra B., Soni C., Ronderos M., Russo S., Ramirez V. & Lager I. (2002). Bluetongue: isolation and characterization of the virus and vector identification in the northeast of Argentina. Rev. Arg. Microb., 34, Groocock C.M. & Campbell C.H. (1982). Isolation of an exotic serotype of bluetongue virus from imported cattle in quarantine. Can. J. Comp. Med., 46, Gumm I.D., Taylor W.P., Roach C.J., Alexander F.C.M., Greiner E.C. & Gibbs E.P.J. (1984). Serological survey of ruminants in some Caribbean and South American countries form type-specific antibody to bluetongue and epizootic haemorrhagic disease viruses. Vet. Rec., 30, Homan E.J., Lorbacher de Ruiz H., Donato A.P., Taylor W.P. & Yuill T.M. (1985). A preliminary survey of the epidemiology of bluetongue in Costa Rica and Northern Colombia. J. Hyg., Camb., 94, Homan E.J., Taylor W.P., Lorbacher de Ruiz H. & Yuill T.M. (1985). Bluetongue virus and epizootic haemorrhagic disease of deer virus serotypes in northern Colombian cattle. J. Hyg., Camb., 95, Karesh B.W., Uhart M., Dierenfeld E.S., Braselton W.E., Torres A., House C., Puche H. & Cook R.A. (1998). Health evaluation of free-ranging guanaco (Lama guanicoe). J. Zoo Wildlife Med., 29 (2), Lage A.P., Castro R.S., Melo M.I.V., Aguiar P.H.P., Barreto Filho J.B. & Leite R.C. (1996). Prevalence of antibodies to bluetongue, bovine herpesvirus 1 and bovine viral diarrhea/mucosal disease viruses in water buffaloes in Mina Gerais State, Brazil. Rev. Elev. Med. Vet. Pays trop., 49 (3), Leoni L., Cheetham S., Lager I., Parreño V., Fondevila N., Rutter B., Martínez Vivot M., Fernández F. & Schudel A. (2001). Prevalencia de anticuerpos contra enfermedades virales del ganado, en llama (Lama glama), guanaco (Lama guanicoe) y vicuña (Vicugna vicugna) en Argentina. II Congreso Latinoamericano de Especialidad en Pequeños Rumiantes y Camelidos Sudamericanos y XI Congreso Nacional de Ovinocultura, Mérida, Yucatán, Mexico, Proceedings CD Lopez W.A., Nicoletti P. & Gibbs E.P.J. (1985). Antibody to bluetongue virus in cattle in Ecuador. Trop. Anim. Hlth Prod., 17, Melo C.B., Oliveira A.M., Azevedo E.O., Lobato Z.I.P. & Leite R.C. (2000). Antibodies to bluetongue virus in bovines of Paraiba State, Brazil. Arq. Bras. Med. Vet. Zootec., 52 (1), Melo C.B., Oliveira A.M., Castro R.S., Lobato Z.I.P. & Leite R.C. (1999). Precipitating antibodies against the bluetongue virus in bovines from Sergipe, Brazil. Cienc. Vet. Trop. Recife, 2 (2), Mo C.L., Thompson L.H., Homan E.J., Oviedo M.T., Greiner E.C., Gonzalez J., Saenz M. & Interamerican Bluetongue Team (1994). Bluetongue virus isolation from vectors and ruminants in Central America and the Caribbean. Am. J. Vet. Res., 55 (2), Office International des Épizooties (OIE) (2003). Terrestrial animal health code, Chapter OIE, Paris. 26. Puntel M., Fondevila N., Blanco Viera J., Marcovecchio J., Carrillo B. & Schudel A. (1999). Serological survey of viral antibodies in llamas (Lama glama) in Argentina. J. Vet. Med., B46, Rivera H., Madewell B. & Ameghino E. (1987). Serologic survey of viral antibodies in the Peruvian alpaca (Lama pacos). Am. J. Vet. Res., 48 (2), Ronderos M., Greco N. & Spinelli G. (2003). Diversity of biting midges of the genus Culicoides Latreille (Diptera: Ceratopogonidae) in the area of the Yacyreta dam lake between Argentina and Paraguay. Mem. Inst. Oswaldo Cruz, 98 (1), Rosadio R.H., Evermann J.F. & DeMartini J.C. (1984). A preliminary serological survey of viral antibodies in Peruvian sheep. Vet. Microbiol., 10 (1), Tamayo R., Schoebitz R., Alonso O. & Wenzel J. (1983). First report of bluetongue antibody in Chile. Progr. Clin. Biol. Res., 178, Tanya V., Greiner E. & Gibbs E. (1992). Evaluation of Culicoides insignis (Diptera: Ceratopogonidae) as a vector of bluetongue virus. Vet. Microbiol., 32, Viegas de Abreu V.L. (1983). Prevalence of reactions to the immunodiffusion test for bluetongue antibodies among cattle and buffaloes in northern Brazil. Abstract of thesis. Arq. Bras. Med. Vet. Zootec., 35 (5), European Commission Health & Consumer Protection Directorate-General (2000). Final report of a mission carried out in Chile from 24 to 28 January 2000 for the purpose of evaluating its Veterinary Services and animal health situation. Report DG(SANCO)/1021/2000-MR-Final. European Commission, Brussels, 21 pp (europa.eu. int/comm/food/fs/inspections/vi/reports/chile/vi_ rep_chil_ _en.pdf accessed on 9 July 2004). 34. Institute for Animal Health (IAH) (2004). The RNAs and proteins of dsrna viruses (P.P.C. Mertens & D.H.Bamford, eds). IAH, Compton (iah.bbsrc.ac.uk/dsrna_virus_proteins/orbivirusphylogenetic-trees.htm accessed on 9 July 2004). 35. Fundación Vida Silvestre Argentina (FVSA) (2004). Operativo de Captura de venados de las pampas en campos del Tuyú. FVSA, Buenos Aires (vidasilvestre.org.ar/pastizales/2_6.asp accessed on 9 July 2004). Veterinaria Italiana, 40 (3),

64 Global situation Vet. Ital., 40 (3), Regional overview of bluetongue viruses in South-East Asia: viruses, vectors and surveillance P.W. Daniels (1), I. Sendow (2), L.I. Pritchard (1), Sukarsih (2) & B.T. Eaton (1) (1) CSIRO Australian Animal Health Laboratory, P.O. Bag 24, Geelong 3220, Australia (2) Research Institute for Veterinary Science, P.O. Box 52, Bogor, Indonesia Summary Structured epidemiological studies based on sentinel herds in Indonesia and Malaysia have provided much information regarding the bluetongue (BT) viruses (BTV) and their likely vectors in South- East Asia. Serotypes 1, 2, 3, 7, 9, 12, 16, 21 and 23 have been isolated. Molecular analyses show all group within the Australasian topotype, with four genotypic sub-groupings identified to date. There are relationships to isolates from both India and Australia. Strains of BTV in South-East Asia do not appear to be highly virulent, since BT disease is not seen in local sheep. Known vector species identified include Culicoides fulvus, C. actoni, C. wadai and C. brevitarsis. C. imicola has not been identified in Malaysian or Indonesian studies. Molecular analyses indicate movement of South-East Asian strains of BTV into northern Australia, and the gradation in observations between India and eastern Australia regarding serotype, genotype, virulence and vector species suggests movement along a conceptual gradient through South-East Asia. Keywords Bluetongue virus Molecular epidemiology Sentinel herd Serotype Topotype Transect Vector Virulence. Since sheep are not a major livestock species in many countries of South-East Asia, studies of bluetongue (BT) have not had the priority given to some other diseases. However, in Malaysia and Indonesia there is interest in the production of sheep meat, and in these countries imports of European sheep breeds have been attempted from time to time. On two occasions this has led to the diagnosis of BT in the imported sheep (1, 33). These outbreaks led to subsequent studies of the epidemiology of bluetongue viruses (BTV) in these countries, from which much of our knowledge of BTV in the region derives. Surveillance programmes For a number of years during the 1990s there was extensive collaboration between Australia and Indonesia (3) and between Australia and Malaysia (29) in BTV epidemiological studies. At the Research Institute for Veterinary Science in Bogor, Indonesia, arbovirology was developed as a discipline, with early work based on serological surveys (17, 18). This progressed to epidemiological studies based on sentinel herd technology (2, 4, 21). The resulting knowledge of arboviruses in Indonesia has been reviewed (3), and the information regarding BTV is summarised again in this paper together with comparisons with data from elsewhere in South-East Asia as well as India to the north-west and Australia to the south-east. Sentinel herd technology was also the basis of BTV studies in peninsular Malaysia (29), where a laboratory capacity was developed at the Veterinary Research Institute in Ipoh. Similar studies have not been reported from neighbouring countries such as Thailand, Vietnam and the Philippines, and little seems to be known of the BTV fauna in these places. It is noted that sentinel herd studies in Southern China, in the Yunnan Province, have also resulted in numerous isolations of BTV over the three-year period 1995 to 1997 in a geographical area in close proximity to South-East Asia (10). In Indonesia, collections of Culicoides were routinely made in conjunction with the sampling of sentinel 94 Veterinaria Italiana, 40 (3), 2004

65 Global situation animals, and extensive data is available regarding the species present in various parts of the country (34, 35). Unfortunately, for those interested in the natural history of BTV in the region, other animal health issues have taken precedence in all countries in the region, and the information from the mid-1990s is in many cases the most recent. Serological surveys Serological surveys for BTV have been conducted in different areas in Indonesia from cattle, buffalo, goats and sheep. Sera were screened with group reactive tests such as the agar gel immunodiffusion (AGID) test and the competitive ELISA (c-elisa) (11). Reactors were evaluated by the serotypespecific serum neutralisation (SN) test (19). Large ruminants had a higher prevalence of exposure than small ruminants. Reactors to BTV serotypes 1, 12, 20 and 21 were widespread on the main islands of Indonesia (17, 19) and, as the studies were broadened, antibodies to BTV serotypes 2, 3, 5, 6, 7, 9, 15, 16 and 23 were also detected. Multiple BTV serotypes were hypothesised to be circulating. Reactors to BTV serotypes 1, 15, 16, 20, 21 and 23 were most prevalent. Elsewhere in the region, in Papua New Guinea to the east, early serological studies detected seroconversions to BTV-1 and other BTVs not identified at that time (6). Subsequently antibodies to BTV-20 and BTV-21 were reported (16). Serological surveillance conducted from 2000 to 2002 identified cattle with antibodies to BTV-16 and BTV-21, and in 2003 antibodies to BTV-3, BTV-20 and BTV-23 were detected (R.A. Lunt, J. Lee, I. Puana and P.W. Daniels, personal communication). Similarly there is serological evidence of BTV infections in East Timor. Isolation of viruses Virus isolation from sentinel cattle in Indonesia has yielded eight serotypes of bluetongue (20, 22, 26). All were recovered in West Java with the exception of serotype 16 which has been isolated to date only in the Province of Papua, formerly known as Irian Jaya, some km km to the east of the West Java site. Serotypes 1, 21 and 23 have also been isolated from Papua (23). These BTV serotypes are widely distributed in Indonesia. It is noted that serotypes 1 and 21 are the two serotypes that are also widely distributed in Australia. Isolation attempts from sentinel cattle in Malaysia have yielded serotypes 1, 2, 3, 9, 16 and 23 (29). Table I compares published information on the BTV serotypes isolated in Malaysia and Indonesia with data from India and northern Australia. BTV 2 was the only serotype found in Malaysia but not Indonesia. Some serotypes were common to all four countries in the comparison (BTV-1, BTV-3, BTV-9 and BTV-16), while BTV-4, BTV-8, BTV-17 and BTV-18 have not been identified further east of India. BTV-2, BTV-7 and BTV-12 were reported in South-East Asia but not Australia, and BTV-20, BTV-21 and BTV-23 have been isolated in South- East Asia and Australia, but not further west in India. Other serotypes may exist there that have not yet been isolated. For instance, the serological survey of Sendow and colleagues (19) showed reactors to serotype 20 were prevalent (23%) in Indonesia, although this virus has not yet been isolated there. Interestingly, the serotypes isolated in southern China a few years later were essentially similar, namely: serotypes 1, 2, 3, 4, 9, 11, 12, 15, 16, 21 and 23 (10). BTV-11 was the only serotype identified in the study conducted in China that had not already been reported in the region (Table I). Genotyping of bluetongue virus isolates from South-East Asia The virus of BT contains 10 double-stranded ribonucleic acid (RNA) segments, or genes. Seven code for structural proteins. RNA3 codes for one of Table I Serotypes of bluetongue viruses isolated in India, South-East Asia and Australia BTV serotype India Malaysia Indonesia Northern Australia Eastern Australia Veterinaria Italiana, 40 (3),

66 Global situation the proteins of the inner core and is relatively conserved among the BTVs. Early molecular epidemiological studies showed that the pattern of nucleotide changes in this gene allow groupings of BTV, irrespective of serotype, that correspond with the geographic origin of the isolates. Hence BTV can be identified to the global region of origin through molecular analyses of RNA3 (7). Within those topotypes, more defined subgroupings can also be recognised, and on this basis, the BTVs of Indonesia and Malaysia have been analysed and compared (15). This latest report further refines previous observations (27). In summary, from the Depok sentinel site in West Java, four genotypes of RNA3 were identified. All were grouped within the Australasian topotype of Gould (7) and, for the purposes of discussion, these subgroupings are now designated as Java A, Java C, Malaysia A and Australia A. Over the 25 years since the first detection of BTV in Australia, a stable genotype has predominated. This is now designated as Australia A in these studies, and some of the isolates from West Java group to this genotype (15). The Java C grouping includes isolates from India. The Malaysia A grouping includes the isolates from Malaysia analysed to date as well as the isolates from the Province of Papua in eastern Indonesia (15). Hence, in Indonesia, genetic groupings of BTV have been identified that overlap with genetic groupings from India to the north-west as well as with those long identified in Australia to the south-east. These relationships are illustrated in Table II. Vector studies Major studies of the distribution of Culicoides spp. have been conducted in Indonesia and extensive lists of collected insects published (34, 35). Four species are proven vectors in Australia, namely C. actoni, C. brevitarsis, C. fulvus and C. wadai; they are also present and widely distributed in Indonesia. There are other closely related Culicoides spp. that could perhaps be vectors and which are not as widely distributed in the region, being exotic to Australia for instance. C orientalis and C. nudipalpis are considered of particular interest. Specific studies of vector competence have not been conducted but there has been a programme of isolation of viruses from insects caught in the wild that has yielded an isolate of BTV serotype 21 from a mixed pool of C. fulvus and C. orientalis (24). An isolate of serotype 21 has also been recovered from a pool of Anopheles mosquitoes (25), but it would be incorrect to consider mosquitoes as vectors without experimental confirmation. Some 42 species of Culicoides have been identified in Malaysia, and it was suggested on the basis of their abundance in insect collections, geographical distribution and host preference, that the possible vector species in that country include C. peregrinus, C. orientalis and C. shortti (8). However, more rigorous studies of vector competence have not been reported. Throughout South-East Asia, there is a need for such objective studies and, at present, the most conservative course of action may be to extrapolate from studies in other regions that have conclusively implicated certain species as vectors. The four species implicated as vectors in Australia occur throughout the region and have been reported in India. It is interesting to note that C. imicola, the dominant vector species in Africa and the Middle East, has not been reported in much of South-East Asia. It is known to occur at least as far east as India and has been reported in Thailand and Vietnam (36). Pathogenicity of bluetongue virus in South-East Asia Bluetongue has been reported in the region only in imported European breeds of sheep (1, 33). There have been no reports of BT in local breeds of sheep, in spite of serological evidence that BTV infections of these breeds has been quite common (8, 19). Experimental infections of Merino sheep with some Indonesian BTV strains maintained by animal Table II Bluetongue virus genotypes identified in South-East Asia showing relationships with isolates from India and eastern Australia (15) Genotype India Malaysia Western Indonesia Eastern Indonesia Northern Australia Eastern Australia Java C Malaysia A Java A Australia A 96 Veterinaria Italiana, 40 (3), 2004

67 Global situation transmission did not elicit clinical disease (28), indicating that these strains were of low virulence. This contrasts with the situation in India where BT is a serious disease of local sheep, resulting in mortalities and pathology, such as myocarditis as well as the usual changes associated with vascular permeability (14, 31). The observations suggest that European breeds of sheep may be more susceptible to BT than Asian breeds, and that the BTV strains in South-East Asia are not as virulent as some strains in India. To the south-east, in northern Australia, the BTV strains are considered to be of only moderate virulence. Certainly they are not as virulent as some South African strains as assessed experimentally (9). Strains in eastern Australia are considered to be essentially non-virulent. This perceived variation in virulence potential of the BTV strains from India through to eastern Australia is illustrated in Table III. Patterns in the ecology of BTV in South- East Asia Variability occurs in the serotypes, genotypes, virulence and vectors of the BTV strains across the South-East Asian region. To give a framework for these observations, it is helpful to consider a conceptual transect from India across South-East Asia, through northern Australia to eastern Australia (Fig. 1). Such transects are used in ecological studies to give rigor to comparative observations of related phenomena in adjacent geographical areas, particularly, where changes may be expected as distance increases from a nominated point of origin (5). Patterns can be described along this transect. At the western end, in India, a greater number of serotypes have been described, including serotypes that do not occur further east. Conversely, serotypes 20, 21 and 23 that are present to the east have not been reported in India. BTVs are considered recent introductions to the viral fauna of northern and eastern Australia, at the eastern end of the transect (30, 32). They are parasites with ruminants as the vertebrate host, and would not have been established in Australia prior to the introduction of ruminants in the last 200 years. The most likely mode of introduction is via infected insects carried on wind movements from infected areas to the north and north-west. The presence of BTV serotypes in northern Australia indicates movement from the west to the east along the transect. In the molecular epidemiological analyses of isolates from South-East Asia, four genotypic subgroupings have been described. The first includes isolates from both India and western Indonesia, the second isolates only from western Indonesia, the third isolates from both western and eastern Indonesia as well as Malaysia, and the fourth isolates from western Indonesia and northern and eastern Australia (Fig. 1). Importantly, genotypes isolated in Indonesia in 1988 and 1990 have appeared in northern Australia in 1992, 1994 and 1995 (5, 13, 15), further evidence of movement of BTV from north-west to south-east along the transect. Most recent molecular analyses have detected more de novo appearances of South- East Asian genotypes in northern Australia, such as the Malaysia A genotype in 2001 (15). At the western end of the transect, highly virulent strains of BTV occur, whereas those at the eastern end are considered essentially non-virulent. Midway along the transect in South-East Asia and northern Australia, some strains occur that show moderate virulence, being moderately pathogenic for European breeds of sheep but non-pathogenic for Asian breeds. Other strains in these areas are also non-virulent. Similarly, it has been noted that species of Culicoides identified as vectors of BTV in northern Australia also occur throughout South-East Asia and in India. Differences between the western and eastern ends of the transect are that C. imicola occurs in India as well as the other vector species, whereas in eastern Australia the main vector species has been reduced to C. brevitarsis, although C. actoni and C. wadai also occur in the more northern areas. Table III Observed variability in virulence of bluetongue viruses from India through South-East Asia to eastern Australia Bluetongue virus India Malaysia Indonesia Northern Australia Eastern Australia Virulent No No No No Mild to moderate virulence? No Non-pathogenic?? Veterinaria Italiana, 40 (3),

68 Global situation Figure 1 Patterns of change in genotype of bluetongue virus isolates along a transect from South Asia through South-East Asia into northern and eastern Australia, based on analyses of RNA 3 Hence, in addition to the observed patterns along the transect that suggest a continuum of BTV activity, there are specific observations for which the most biologically plausible explanation is movement of strains along the transect from west to east. Will future movements from west to east result in the arrival of more virulent strains than currently occur in some areas? C. imicola is recognised as a vector in countries where virulent BTV occurs. Would extension of the range of C. imicola further eastwards be an event of concern for animal health in the region? Its presence where virulent BTV strains are circulating and absence where BTV is less pathogenic may be an association that should be investigated experimentally. To balance these concerns, it should be noted that there may be ecological blocks along the transect. For example, although C. imicola has been reported in the more northerly countries of South-East Asia, it has not moved into Malaysia and Indonesia. Is it simply a matter of time, or are other unidentified factors involved? Similarly, in northern Australia six of the eight identified serotypes that are maintained in a complex vector system involving C. fulvus, C. actoni, C. wadai as well as C. brevitarsis (12) have not moved to the eastern states where serotypes 1 and 21 are maintained in a predominantly C. brevitarsis ecosystem. There is a challenge to devise experimental systems to explore the observed associations. In any dynamic biological system there will be checks and balances. These are not understood for BTV in South-East Asia, where the minimum requirement is for more surveillance and monitoring of the BTV ecosystems to provide basic current data. Acknowledgements The support of the United States Naval Medical Research Unit (NAMRU), the North Australia Quarantine Strategy (NAQS), the Australian Exotic Animal Disease Preparedness Program (EXANDIS) and the Australian National Arbovirus Monitoring Program (NAMP) is gratefully acknowledged. References 1. Chiang B.L. (1989). An outbreak of bluetongue in imported sheep. In Proc. 1st Congress, Veterinary Association of Malaysia, Daniels P.W., Sendow I., Soleha E., Jennings J. & Sukarsih (1991). A veterinary arbovirus monitoring program. In Proc. 6th Congress of the International Society for Veterinary Epidemiology and Economics (ISVEE) (S.W. Martin, ed.). ISVEE, Ottawa, Daniels P.W., Sendow I., Soleha E., Sukarsih, Hunt N.T. & Bahri S. (1995). Australian-Indonesian collaboration in veterinary arbovirology. A review. Vet. Microbiol., 46, Daniels P.W., Sendow I. & Melville L. (1996). Epidemiological considerations in the study of 98 Veterinaria Italiana, 40 (3), 2004

69 Global situation bluetongue viruses. In Bluetongue disease in South- East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Daniels P.W., Melville L., Pritchard I., Sendow I., Sreenivasulu D. & Eaton B. (1997). A review of bluetongue viral variability along a South Asia Eastern Australia transect. Arbovirus Res. Aust., 7, Della-Porta A.J., Sellers R.F., Herniman K.A.J., Littlejohns I.R., Cybinski D.H., St George T.D., McPhee D.A., Snowdon W.A., Campbell J., Cargill C., Corbould A., Chung Y.S. & Smith V.W. (1983). Serological studies of Australian and Papua New Guinea cattle and Australian sheep for the presence of antibodies against bluetongue group viruses. Vet. Microbiol., 8, Gould A.R. (1987). The nucleotide sequence of bluetongue virus serotype 1 RNA 3 and a comparison with other geographic serotypes from Australia, South Africa, the United States of America and other orbivirus isolates. Virus Res., 7, Hassan A. (1992). Epidemiology of bluetongue virus infection in Malaysia. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Hooper P.T., Lunt R.A. & Stanislawek W.L. (1996). A trial comparing the pathogenicity of some South African and Australian bluetongue viruses. Aust. Vet. J., 73, Kirkland P.D., Zhang N., Hawkes R.A., Li Z., Zhang F., Davis R.J., Sanders D.A., Li H., Ben J., He G.F., Hornitzky C.L. & Hunt N.T. (2002). Studies on the epidemiology of bluetongue virus in China. Epidemiol. Infec., 128, Lunt R.A., White J.R. & Blacksell S.D. (1988). Evaluation of a monoclonal antibody blocking ELISA for the detection of group-specific antibodies to bluetongue virus in experimental and field sera. J. Gen. Virol., 69, Melville L., Hunt N.T. & Daniels P.W. (1996). Application of the polymerase chain reaction (PCR) test with insects in studying bluetongue virus activity. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Melville L.F., Pritchard L.I., Hunt N.T., Daniels P.W. & Eaton B. (1997). Genotypic evidence of incursions of new strains of bluetongue viruses in the Northern Territory. Arbovirus Res. Aust., 7, Prasad G., Jain N.C. & Gupta Y. (1992). Bluetongue virus infection in India: a review. Rev. Sci. Tech. Off. Int. Épiz., 11, Pritchard L.I., Daniels P.W., Melville L.F., Kirkland P.D., Johnson S.J., Lunt R. & Eaton B.T. (2004). Genetic diversity of bluetongue viruses in Australasia. In Bluetongue, Part II (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (4), Puana I. (1996). Bluetongue virus status in Papua New Guinea. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Sendow I., Young P. & Ronohardjo P. (1986). Preliminary survey for antibodies to bluetonguegroup virus in Indonesian ruminants. Vet. Rec., 119, Sendow I., Hunt N., Bale A., Bartz C., Tarmudji, Ronohardjo P. & Daniels P.W. (1988). Studies of arboviral infections of sentinel cattle in Indonesia. In Proc. Sixth Congress of the Federation of Asian Veterinary Associations (D. Sastradipradja & S.H. Sigit, eds). Indonesian Veterinary Association, Jakarta, Sendow I., Daniels P.W., Cybinski D.H., Young P.L. & Ronohardjo P. (1991). Antibodies against certain bluetongue and epizootic haemorrhagic disease viral serotypes in Indonesian ruminants. Vet. Microbiol., 28, Sendow I., Daniels P.W., Soleha E., Hunt N. & Ronohardjo P. (1991). Isolation of bluetongue viral serotypes 7 and 9 from healthy sentinel cattle in West Java, Indonesia. Aust. Vet. J., 68, Sendow I., Daniels P.W., Soleha E. & Sukarsih (1992). Epidemiological studies of bluetongue viral infections in Indonesian livestock. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Sendow I., Daniels P.W., Soleha E., Erasmus B., Sukarsih & Ronohardjo P. (1993). Isolation of bluetongue virus serotypes new to Indonesia from sentinel cattle in West Java. Vet. Rec., 133, Sendow I., Soleha E., Daniels P.W., Sebayang D., Achdiyati J., Karma K. & Erasmus B.J. (1993). Isolation of bluetongue virus serotypes 1, 21 and 23 from healthy cattle in Irian Jaya, Indonesia. Aust. Vet. J., 70, Sendow I., Sukarsih, Soleha E., Erasmus B.J. & Daniels P.W. (1993). Isolation of bluetongue virus serotype 21 from Culicoides spp. in Indonesia. Vet. Microbiol., 36, Sendow I., Sukarsih, Soleha E. & Daniels P.W. (1994). Isolation of bluetongue virus serotype 21 Veterinaria Italiana, 40 (3),

70 Global situation from mosquitoes in West Java, Indonesia. Penyakit Hewan, 26 (48), Sendow I., Sukarsih, Soleha E., Pearce M., Bachri S. & Daniels P.W. (1996). Bluetongue virus research in Indonesia. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Sendow I., Pritchard L.I., Eaton B.T. & Daniels P.W. (1997). Genotypic relationships of bluetongue viruses from Indonesia. Arbovirus Res. Aust., 7, Sendow I., Hamid H., Sukarsih, Soleh E., Bahri S., Pearce M. & Daniels P.W. (1997). Clinical and pathological changes associated with the propagation of Indonesian bluetongue viral isolates in Merino sheep. Hemera Zoa, 79, Sharifah S.H., Ali M.A., Gard G.P. & Polkinghorne I.G. (1995). Isolation of multiple serotypes of bluetongue virus from sentinel livestock in Malaysia. Trop. Anim. Hlth Prod., 27, Snowdon W.A. (1979). Bluetongue virus infection in Australia. In Arbovirus research in Australia. Proc. Second Symposium (T.D. St George & E.L. French, eds). CSIRO Division Tropical Animal Science and Queensland Institute for Medical Research, Brisbane, Sreenivasulu D., Subba Rao MV & Gard G.P. (1996). Bluetongue viruses in India: a review. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, St George T.D., Standfast H.A., Cybinski D.H., Dyce A.L., Muller M.J., Doherty R.L., Carley J.G., Filippich C. & Frazier C.L. (1978). The isolation of a bluetongue virus from Culicoides collected in the Northern Territory of Australia Aust. Vet. J., 54, Sudana I.G. & Malole M. (1982). Penyidikan penyakit hewan bluetongue di desa Caringin, Kabupaten Bogor. In Annual report of disease investigation in Indonesia during the period , Direktorat Kesehatan Hewan, Direktorat Jenderal Peternakan, Jakarta, Sukarsih, Daniels P.W., Sendow I. & Soleha E. (1993). Longitudinal studies of Culicoides spp. associated with livestock in Indonesia. In Arbovirus research in Australia. Proc. Sixth Symposium (M.F. Uren & B.H. Kay, eds). CSIRO/QIMR, Brisbane, Sukarsih, Sendow I., Bahri S., Pearce M. & Daniels P.W. (1996). Culicoides surveys in Indonesia. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Wirth W.W. & Hubert A.A. (1989). The Culicoides of South-East Asia (Diptera: Ceratopogonidae). The American Entomological Institute, Gainesville, Florida, 508 pp. 100 Veterinaria Italiana, 40 (3), 2004

71 Vet. Ital., 40 (3), Global situation Serological surveillance of bluetongue virus in cattle, sheep and goats in Albania M. Di Ventura (1), M. Tittarelli (1), G. Semproni (1), B. Bonfini (1), G. Savini (1), A. Conte (1) & A. Lika (2) (1) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy (2) Institute of Veterinary Research Bilal Golemi, Department of Virology, Tirana, Albania Summary The recent spread of the bluetongue (BT) virus (BTV) in the Mediterranean Basin encouraged numerous countries to undertake entomological and serological surveillance programmes to identify affected areas and control the infection. Hitherto, no data on the presence and diffusion of BTV in Albania were available. Between October and November 2002 serum samples from 857 cattle and 870 sheep and goats were collected by the Albanian Veterinary Services in 15 districts, some bordering Yugoslavia, Macedonia and Greece, and others along the Adriatic coast. At the Albanian Veterinary Research Institute the samples were tested for the presence of BTV antibodies using a competitive enzyme-linked immunosorbent assay (c-elisa) (bluetongue antibody test kit, IZSA&M, Teramo); in Italy, the virus neutralisation (VN) test was used to confirm the ELISA results and determine the serotype of BTV circulating. Overall seroprevalence was 18.9% in cattle and 4.4% in sheep and goats; seropositive animals occurred in all districts surveyed. The highest prevalence of BT was observed in the Tirana District, with 61% of the cattle and 20% of the sheep and goat populations BT-positive. The VN test confirmed the c-elisa results revealing the presence of antibodies against BTV serotype 9. Keywords Albania Bluetongue Cattle Competitive enzyme-linked immunosorbent assay Goats Sheep Surveillance Virus Virus neutralisation. Introduction The incursion of bluetongue (BT) virus (BTV) into the Mediterranean Basin had a serious impact on numerous countries especially in those that had never previously experienced the infection. To identify affected areas and better control the spread of infection, some countries undertook entomological and serological surveillance programmes. Whilst four BTV serotypes (BTV-1, BTV-4, BTV-9 and BTV-16) have been reported to occur in the Balkans (Fig. 1) (4), no data on the presence and spread of BT in Albania were hitherto available. Accordingly, and under the aegis of the National Survey on Foot and Mouth Disease and Bluetongue (a joint project between the Albanian Veterinary Services, the Veterinary Research Institute of Tirana and the Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale [IZSA&M]), a population of mixed domestic ruminants was surveyed for the presence of BT antibodies. Figure 1 Bluetongue virus serotypes reported in the Balkans, 2002 Veterinaria Italiana, 40 (3),

72 Global situation Materials and methods Sampling Between October and November 2002, serum samples from 857 cattle and 870 sheep and goats, were collected by the Albanian Veterinary Services. Animals were randomly selected in 15 Albanian districts, some bordering Yugoslavia, Macedonia and Greece, others facing the Adriatic Sea. Serological tests At the Albanian Veterinary Research Institute, all serum samples were tested for the presence of BT antibodies using the competitive enzyme-linked immunosorbent assay (c-elisa) produced by the IZSA&M in Teramo (3). The virus neutralisation (VN) test (2) was used to confirm the positive ELISA results. Positive and negative controls for the VN were kindly provided by the Onderstepoort Veterinary Institute in South Africa. The microtitre neutralisation method was used in this study. Fifty µl of several sera dilutions, from 1:10 to 1:1 280, were added to each test well of flat-bottomed microtitre plates and mixed with an equal volume of Office International des Épizooties standard reference BTV serotypes 2, 4, 9 and 16 (100 TCID 50 ). They were incubated at 37 C in 5% CO 2. After 1 h incubation, approximately 10 4 Vero cells were added per well in a volume of 100 µl of minimum essential medium (MEM) containing antibiotics and, after incubation for 4 to 6 days, the test was read using an inverted microscope. Wells were scored for the degree of cytopathic effect (CPE) observed. A sample was considered positive when it showed a CPE of more than 50% neutralisation at the lowest dilution (1:10). The serum titre represented the highest serum dilution capable of neutralising more than 50% CPE in the tissue culture. Statistical analysis For each district, the prevalence level and 95% confidence intervals were determined for both cattle and sheep. The prevalence level in both species was calculated through Beta (s+1, n-s+1) distribution where s is the total number of positives and n is the total number of animals tested. The probability distribution of the percentage of positive animals shows not only the most probable value of prevalence, but also the level of uncertainty due to sample size. Results Of the serum samples screened, 200 were found positive for BT antibody by c-elisa. Of the 200 ELISA-positive samples, only 71 were found positive by the VN test. All neutralising antibodies were against serotype 9 and the titres ranged from 1/10 to 1/320. Prevalence levels and confidence intervals are given in Tables I, II and III and the most probable values with the relative uncertainties of BT prevalence in Albanian cattle, sheep and goats Table I Prevalence levels and 95% confidence intervals of bluetongue found in Albanian cattle according to district of origin (total: 857) District Animals tested Positive to VN test Positive to c-elisa Prevalence level Lower level Upper level Bulgise % 17.94% 50.83% Devoll % 1.23% 13.46% Dibre % 0.77% 16.22% Durres % 5.70% 23.87% Fier % 4.44% 21.41% Gjirokaster % 21.21% 46.70% Has % 12.28% 42.28% Kolonje % 7.02% 26.26% Korce % 3.84% 10.75% Librazhd % 38.46% 65.25% Permet % 3.26% 18.88% Pogradet % 8.00% 24.38% Shkoder % 10.95% 33.91% Tirane % 47.73% 72.98% Tropoje % 17.49% 41.74% Total/average % 16.42% 21.66% VN virus neutralisation c-elisa competitive enzyme-linked immunosorbent assay 102 Veterinaria Italiana, 40 (3), 2004

73 Global situation Table II Prevalence levels and 95% confidence intervals of bluetongue found in Albanian sheep and goats according to district of origin (total: 870) District Animals tested Positive to VN test Positive to c-elisa Prevalence level Lower level Upper level Bulgise % 0.79% 16.70% Devoll % 2.18% 16.24% Dibre % 2.48% 18.27% Durres % 0.05% 6.72% Fier % 0.04% 5.96% Gjirokaster % 0.03% 4.30% Has % 4.54% 31.22% Kolonje % 0.05% 6.98% Korce % 0.61% 4.79% Librazhd % 10.03% 31.44% Permet % 0.52% 11.29% Pogradet % 0.88% 9.81% Shkoder % 0.07% 9.74% Tirane % 11.29% 33.12% Tropoje % 2.22% 16.55% Total/average % 3.20% 5.94% VN virus neutralisation c-elisa competitive enzyme-linked immunosorbent assay Table III Prevalence levels and 95% confidence intervals of bluetongue found in Albanian sheep, goats and cattle according to district of origin (total: 1 727) District Animals tested Positive to c-elisa Prevalence level Lower level Upper level Bulgise % 9.69% 28.97% Devoll % 2.21% 11.18% Dibre % 2.17% 12.93% Durres % 2.78% 12.25% Fier % 2.03% 10.29% Gjirokaster % 7.63% 18.80% Has % 10.63% 31.43% Kolonje % 3.48% 13.76% Korce % 2.62% 6.73% Librazhd % 26.64% 45.18% Permet % 2.30% 11.62% Pogradet % 5.00% 14.39% Shkoder % 6.08% 20.05% Tirane % 32.34% 50.98% Tropoje % 11.03% 25.82% Total % 10.16% 13.18% c-elisa competitive enzyme-linked immunosorbent assay are shown in Figure 2. The overall prevalence was 18.9% in cattle (animals tested: 857) and 4.4% in sheep (animals tested: 870). The Tirana District was the region in which the highest prevalence was observed, with 61% of the cattle and 20% of the sheep and goat populations positive for BT antibodies (Fig. 3). Veterinaria Italiana, 40 (3),

74 Global situation Probability % 20% 40% 60% 80% 100% Percentage of positive sample estimate Cattle Sheep Cattle and sheep Figure 2 Probability distributions of the percentage of animals with bluetongue antibodies found in Albania (n = 1 727) not determined. The finding that serotype 9 was the only type present would suggest that BTV infection in Albania originated from a neighbouring country. As with the prevalence levels obtained in this study, the infection was not uniformly distributed in the areas and species examined. Librazhd and Tirana were the districts with the highest infection rates and BT was found to be more prevalent in cattle than in sheep. Bluetongue is a non-contagious disease of ruminants transmitted between its vertebrate hosts by Culicoides biting midges. There is evidence that at least some species of Culicoides prefer to feed on cattle rather than on sheep (5). This preference could account for the species prevalence differences observed in this study. As mentioned before, Tirana and Librazhd were the districts with the highest BT prevalence; interestingly, Tirana was also the district in which animals showed the highest neutralising antibody titres. Moreover, Tirana was the region in which the highest number of ELISA-positive samples were confirmed by the VN test (41.9%) and double the ratio found in Librazhd. The c-elisa is able to detect BT antibodies earlier and over a longer period than the VN test. The results of this study would suggest that, at least with regard to these two districts where a large number of animals were tested, BT occurred first in Tirana and then in Librazhd. Based on the low number of positive sera confirmed by the VN test (0.2%) and the low neutralising antibody titres found, the infection seemed to be recent. The BT outbreak reported at the end of 2001 in Librazhd provided further evidence of the presence of recent infection (1). Prevalence (%) >20 Figure 3 Prevalence of bluetongue in sheep and cattle in Albania, according to geographic origin Discussion The most remarkable developments of BT infection have been observed recently in south-eastern Europe. Between 1998 and 1999, BT, predominantly caused by serotype 9, occurred in north-eastern Greece, southern Bulgaria and in western Turkey. In 2001, BT was reported from north-western Greece, western Bulgaria, Kosovo, Macedonia and Montenegro and Serbia. According to the results obtained in this study, BTV was also circulating in Albania. As not all districts were included in the survey, the possible origin of BTV in Albania was References 1. Anon. (2003). Bluetongue in Albania. Dis. Info., 16, Gard G.P. & Kirkland P.D. (1993). Bluetongue virology and serology. In Australian standard diagnostic techniques for animal disease (L.A. Corner & T.J. Bagust, eds). CSIRO Information Services, Melbourne, Lelli R., Portanti O., Langella V., Luciani M., Di Emidio B. & Conte A. (2003). Produzione di un kit ELISA competitiva per la diagnosi sierologica della Bluetongue. Vet. Ital., 47, Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean Basin Vet. J., 164, Nevill E.M. (1979). The use of cattle to protect sheep from bluetongue infection. J. Sth Afr. Vet. Assoc., 49, Veterinaria Italiana, 40 (3), 2004

75 Vet. Ital., 40 (3), Global situation The epizootiological occurrence of bluetongue in the central Balkans B. Djuričić, D. Nedić, D. Lausević & M. Pavlović Katedra za zarazne bolesti, Fakultet veterinarske medicine, Bulevar JNA 18, Belgrade, Serbia Summary In the last five years, bluetongue has been diagnosed in the following areas of the Mediterranean Basin: Algeria, France, Greece, Italy, Spain, Tunisia and Turkey. In the Balkan Peninsula, the disease has been recorded in Bulgaria (since 1999), Macedonia, Serbia and Montenegro, Croatia, Bosnia and Herzegovina and Albania. Bluetongue arrived in the Balkans from Turkey in In Serbia and Montenegro, the disease was first diagnosed in July 2001 in the Zubin Potok region of Kosovo. In August, the neighbouring regions of Novi Pazar, Tutin, Rozaje and Leposavic were also affected. The disease was simultaneously observed in eastern Serbia in the regions bordering Bulgaria (city regions of Bosilegrad, Pirot and Knjazevac). The occurrence of the disease and clinical signs indicate that the disease was also present in western Serbia in During 2002, more extensive serological investigations of animals in some regions of Yugoslavia indicated that the disease had spread towards the north of the country. At the same time, serological evidence also revealed the presence of the disease in the east of Bosnia and Herzegovina. The epizootiological data show that the northernmost point of disease spread in Serbia was the River Sava. Keywords Balkans Bluetongue Bosnia Culicoides Disease spread Epizootiology Herzegovina Mediterranean Montenegro Serbia Virus. Bluetongue (BT) is an acute viral disease of sheep, goats and cattle, to which deer are also susceptible and is transmitted by haematophagous insects of the genus Culicoides. Symptoms include fever, catarrhalnecrotic changes of the nasal and oral cavity mucosa, tongue, digestive organs, coronet, as well as degenerative changes in skeletal muscles. Swelling of the face or tongue, especially in cattle, as well as abortions and stillbirths, can also occur. The worldwide incidence of the disease is linked to the habitat of the infected animals and vector insects. With increased global warming, insect carriers have spread from 40 N to 43 N. In the last five years, in the Mediterranean region, the disease has been diagnosed in Algeria, France, Greece, Italy, Spain, Tunisia and Turkey. In the Balkans, the disease first appeared in Bulgaria in 1999 apparently originating from neighbouring Turkey. Macedonia, Serbia and Montenegro, Croatia, Bosnia and Herzegovina, and Albania were affected subsequently. In Serbia and Montenegro, the disease was first diagnosed in July 2001 in the Zubin Potok region (Kosovo), and during August in neighbouring regions, namely: Novi Pazar, Tutin, Rozaje, Leposavic, etc. (Fig. 1). Seropositive animals without clinical symptoms First cases recorded in Serbia Outbreaks in Kosovo (UNMIK: United Nations interim administration mission in Kosovo) Kosovo Cases recorded after BT reported in Kosovo Figure 1 Bluetongue in Serbia, 2001 Veterinaria Italiana, 40 (3),

76 Global situation Simultaneously, the disease was observed in eastern Serbia, in the border regions with Bulgaria (city regions of Bosilegrad, Pirot and Knjazevac). Based on the epizootiological occurrence of the disease and on clinical symptoms (Fig. 2), it can be concluded that the disease was also present in eastern Serbia in 2001, but low lethality and mild clinical signs meant that it remained undetected. the current study on the occurrence and spread of the disease in this area, BT still remains a major international problem. The disease will demand constant attention from the National Veterinary Services, as well as from the Office International des Épizooties. It is also clear that effective control will require more concerted action between all the regional veterinary services of southern Europe. First cases (June 2001) Bluetongue in Kosovo (2001) (Bosnia and Herzegovina and Croatia) Figure 3 Epizootiological situation of bluetongue in Serbia and Montenegro and Bosnia and Herzegovnia, 2003 Figure 2 Laminitis, clinical symptoms observed in sheep in Serbia, 2001 During 2002, more extensive serological investigations of animals in the Yugoslav regions were performed with the spread of the disease confined to the north of the country. The first serological evidence coincided with reports of the disease in the eastern part of Bosnia and Herzegovina (Fig. 3). From an epizootiological point of view, it can be said that the most northern point of disease spread in Serbia is the River Sava. Diagnostic studies currently being performed in Serbia and Montenegro, as well as in Bosnia and Herzegovina, will hopefully clarify further the true epizootiological picture of the disease in the central Balkans. It is necessary to also investigate which species of Culicoides are involved in the transmission of BT in the central Balkans. With the occurrence of BT in the Balkans in mind, it can be concluded that, irrespective of the results of Additional reading 1. Anon. (1992). Conclusions of the Research Subcommittee of the International Embryo Transfer Society (IETS) Import/Export Committee. Denver, 14 January. Rev. Sci. Tech. Off. Int. Épiz., 11 (3), Djuričić B. & Radojičić S. (2001). Prva pojava Bluetongue u Jugoslaviji Zbornik radova 7 [First occurrence of bluetongue in Yugoslavia, Collection No. 7]. Savetovanje veterinara, R. Srpske, Terslić, Banja Vrućica, Campbell C.H., Breese S.S. Jr & McKercher P.D. (1975). Antigenic and morphologic comparisons of Ibaraki and bluetongue viruses. Can. J. Microbiol., 21 (12), Doherty R.L. (1977). Arthropod-borne viruses in Australia, Aust. J. Exp. Biol. Med. Sci., 55 (2), Luedke A.J., Jones R.H. & Walton T.E. (1977). Overwintering mechanism for bluetongue virus: biological recovery of latent virus from a bovine by bites of Culicoides variipennis. Am. J. Trop. Med. Hyg., 26 (2), Veterinaria Italiana, 40 (3), 2004

77 Global situation 6. Office International des Épizooties (OIE) (2000). Manual of standards for diagnostic tests and vaccines. OIE, Paris, Osburn J. & Stott G. (1976). Fed. Proc., 38 (3), Sellers R.F., Pedgley D.E. & Tucker M.R. (1978). Possible windborne spread of bluetongue to Portugal, June-July J. Hyg. (Lond.), 81 (2), Sellers R.F., Gibbs E.P., Herniman K.A., Pedgley D.E. & Tucker M.R. (1979). Possible origin of the bluetongue epidemic in Cyprus, August J. Hyg. (Lond.), 83 (3), Sergeev V.A. & Ananeva-Rjaščenko N.P. (1980). Veterinarija, 12. Veterinaria Italiana, 40 (3),

78 Global situation Vet. Ital., 40 (3), Overview of bluetongue in Greece K. Nomikou (1), O. Mangana-Vougiouka (1) & D.E. Panagiotatos (2) (1) Ministry of Agriculture, Athens Center of Veterinary Institutes, Virology Department, 25 Neapoleos Str., Ag. Paraskevi, Athens, Greece (2) Ministry of Agriculture, Athens Center of Veterinary Institutes, Department of Epidemiology and Bio-Statistics, 25 Neapoleos Str., Ag. Paraskevi, Athens, Greece Summary The history and epizootiology of bluetongue (BT) in Greece are described in detail. Three major epidemics of BT occurred in Greece, the first in 1979, due to bluetongue virus (BTV) serotype 4, in due mainly to BTV-4 and BTV-9 and less to BTV-16 and in 2001 due mainly to BTV-1. The evolution of the disease, surveillance and control measures are presented, as well as the distribution of the BTV serotypes isolated. Keywords Bluetongue virus Culicoides Disease control Epidemiology Greece Outbreaks Sentinel Serotypes. Introduction The first incursion of bluetongue (BT) into Greece occurred on the island of Lesbos in October The island is situated in the eastern part of the Aegean Sea, km from the Asian coast. The sheep population of Lesbos (area: km 2 ) numbers head, mainly of a local breed that is characterised by high milk production. The outbreak started in the north-eastern part and extended to the east of the island. From October to the end of December 1979, 17 communities, with a sheep population of , were affected. Sixtyeight flocks with about sheep were infected. The morbidity rate in individual flocks varied from 10-90% with a mortality rate of 29% in infected sheep (11, 24). The cattle and goats showed no clinical signs. No clinical cases were observed in subequent years on the island. Antibodies to BTV have been found in animals on Rhodes (6), but no clinical disease has been reported on the island or on the mainland Greece. Regionalisation of the area was applied and exports of live ruminants from the island were prohibited. In addition, the slaughter of all infected, recovered and seropositive ruminants was undertaken. One year later, an eradication programme was initiated. All cattle on the island were serologically tested each year at the end of the vector season. Between 1980 and 1986, approximately 560 seropositive bovines were slaughtered. Exports of small ruminants were allowed only during winter (season of low vector activity) and only after negative serological results had been received. A sentinel bovine and ovine system was applied. Restrictions continued until Surveillance and control measures were also applied on the islands of the east Aegean Sea from 1980 to The successful application of surveillance and control measures resulted in the official lifting of EU restrictions in From 1991 through to the summer of 1998, the surveillance measures continued on the islands of the east Aegean and clinical disease was not reported. A total of sera, mainly from sheep and goats from the Dodecanese, were examined during the summer of 1998 and gave negative results. In October 1998, clinical outbreaks of BT were reported on three east Aegean islands (Rhodes, Kos and Leros) with serological evidence of BT on another island (Samos), adjacent to the coast of Anatolian Turkey (Table I). From October until late December, 84 outbreaks were recorded. No further outbreaks of BT were reported over the following months until August 1999 when BT outbreaks occurred on mainland Greece, commencing in the north-eastern prefecture of Evros, adjacent to the Turkish and Bulgarian borders. The virus spread across northern and central Greece, affecting 16 prefectures (approximately one third of the country). Also during September 1999, a separate 108 Veterinaria Italiana, 40 (3), 2004

79 Global situation Table I Incidence of outbreaks of bluetongue in Greece, Year of outbreak No. of prefectures Location of infected prefectures North-eastern Greece (island of Lesbos) South-eastern Greece (islands of Rhodes, Kos and Leros) Eastern, north, and mainland Greece No. of prefectures with serological evidence Location of prefectures with serological evidence 1 South-eastern Greece (island of Rhodes) 1 Eastern Greece (island of Samos) No. of outbreaks No. of animals destroyed Presence of Culicoides vectors Serotypes isolated C. imicola BTV C. imicola BTV-9 5 North-eastern Greece C. imicola C. obsoletus BTV-4 BTV-9, BTV Western Greece 10 0 C. obsoletus BTV North-western, western, north-eastern, and mainland Greece 3 Western Greece C. imicola C. obsoletus BTV-1 BTV-4 BTV-9 incursion of BT occurred on the islands of the north-east, east and south-east Aegean Sea. A total of clinical outbreaks were reported and ruminants destroyed. Mild symptoms were reported in some cases in bovines and goats (Table I). The most severely affected area was the island of Lesbos, where 260 clinical cases occurred from September until the end of December In the autumn of 2000, only one prefecture was infected in central-western Greece. Ten clinical outbreaks involved 50 sick but no dead animals. Seroprevalence in this prefecture was 7.9%. In September 2001, BT recurred but this time in north-western Greece on the border with Albania and then subsequently in western and central Greece. Outbreaks were also reported on the island of Lesbos which was heavily infected from October until the end of December During this incursion, 14 prefectures reported 174 clinical outbreaks, with ruminants sick and destroyed (Table I). Measures to safeguard the trade in live animals, i.e. control of Culicoides spp. and other control and epidemio-surveillance measures, have been applied from 1998 to date. Serological monitoring of sentinel bovines in 17 prefectures was established in Serological surveillance of sentinel bovines for virus activity will be conducted in 14 prefectures of east, north, west and mainland Greece from May 2003 until the end of December Materials and methods Safeguard control and surveillance measures Safeguard measures for live animal trade When confronted with the epidemic, three zones were established around outbreaks, on the basis of geographical and administrative criteria, as follows: a) a zone 20 km in radius (protection zone), where a complete standstill of animal movements within the zone was implemented b) areas of affected prefectures outside the radius of 20 km (surveillance zone), where no movements were allowed unless serological tests were performed giving 100% negative results c) non-affected prefectures adjacent to affected areas (buffer zones), where clinical and serological surveillance including sentinels was applied d) prefectures far from affected areas (free zones), where no restrictions were applied to the domestic trade in live susceptible animals e) a blanket export ban on live susceptible animals was applied. Control measures The following control measures are applied at present: 1) when faced with an epidemic(s): a) elimination of clinically affected animals b) large-scale and regular vector control measures. Veterinaria Italiana, 40 (3),

80 Global situation 2) in the absence of evidence of virus circulation: a) serological monitoring of sentinel bovines b) in case of seroconversion, elimination of viraemic bovines and targeted vector control measures. Control of Culicoides spp. Control of BT vectors is applied as follows: 1) during winter and early spring to destroy vectors at the larval stage and reduce the probability of infective vectors over-wintering 2) during spring and summer to destroy vectors at the adult stage and/or prevent vectors from biting naive susceptible animals. Spraying is applied at three levels and insect repellent is provided free of charge by the local State Veterinary Services. The three levels of spraying are as follows: Level 1: spraying, with insect repellent, of individual animals (bovines, sheep and goats) every 4-7 days; applied by farmers. Level 2: spraying, with insecticide, of damp sites inside and around animal holdings; applied by farmers. Level 3: spraying, with insecticide, of breeding sites; applied by local authorities and the State. Epidemiological surveillance Included in epidemiological surveillance are the following: a) Serological monitoring of sentinel bovines b) Isolation and serotyping of BT viruses c) Seroprevalence d) BTV vector monitoring Sentinel animal surveillance To study the incidence of infection in the country, a sentinel bovine system was established. Sentinel herds were selected from areas in each prefecture where vector populations are sufficient to ensure virus transmission. Insects were trapped for species identification and population trends by using light traps in the vicinity of the sentinel herd. Sentinel bovines were established in prefectures affected in previous years and in prefectures at risk. Five groups of 10 seronegative bovines, above six months of age, were placed in each prefecture. Heparinised and clotted blood samples were collected from each animal at day intervals. When a sentinel animal seroconverted, the viraemic animal was eliminated from the group (slaughtered) and replaced with a seronegative animal. Insect collections were made for virus isolation. Field samples All state field veterinarians collaborate with the national laboratory in active surveillance by submitting samples from BTV cases, establishing and seromonitoring the sentinel bovine herds and submitting Culicoides to the Entomology Laboratory. They are also responsible for applying the measures to safeguard domestic trade of live animals and for control measures. In a case of suspected BTV infection, samples are tested for foot and mouth disease, peste des petits ruminants, orf and sheep and goat pox viruses, in accordance with the methods recommended by the Office International des Épizooties (OIE). The samples are heparinised blood, serum and, in the case of the death of an animal, spleen and lymph nodes. Tissues are triturated and a suspension of 1:5 in buffered lactose peptone containing antibiotics is prepared. The red blood cells are washed (10) and are ruptured by lysing in sterile, distilled water. For isolation from insects in the Entomology Laboratory, the Culicoides are sorted into species and then into parous pools with blood-fed females pooled with non-parous females; the males are discarded. Pools of 50 to 100 parous biting midges of a single species or mixed species are placed into vials containing MEM with antibiotics and are sent to the virology laboratory for BTV isolation and identification. The lysed blood cells, tissues and insects are held at +4 C or 70 C until inoculated. Serological testing From 1979 to 1985, the serum samples were tested using the immunodiffusion test (16), while from 1986 to date, competitive enzyme-linked immunosorbent assay (c-elisa) is performed (1), using the serogroup reactive monoclonal antibody 3-17-A3. The BTV serotype-specific antibodies are determined using the serum neutralisation test in BHK-21 cells (8). Differential diagnosis Differential diagnosis is performed in clinical cases for the following purposes: 1 to detect antibodies against epizootic haemorrhagic disease of deer virus (EHDV), using the c-elisa (23) 2) to detect antibodies against orf and against sheep and goat pox viruses using the immunodiffusion or serum neutralisation test (15). Virus isolation and identification The samples thus prepared are inoculated intravenously into 11- to 12-day-old embryonated chicken eggs (7). Embryos found dead between 110 Veterinaria Italiana, 40 (3), 2004

81 Global situation 2 and 7 days post inoculation (pi) are collected for inoculation into BHK-21 tissue cultures and/or Vero cells (14). For identification of the isolated virus, indirect immunofluorescence was used (9, 10), but this was replaced with the antigen-capture ELISA for BTV (14) and for EHDV (22), as well as the polymerase chain reaction (PCR) test (3). The isolated BTV is serotyped with the microtitre virus neutralisation test (14). Each time a new BTV serotype is identified, results are confirmed by the OIE Reference Laboratories (Onderstepoort Veterinary Institute, 1979 for BTV-4 and the Institute of Animal Health, Pirbright Laboratory, 1998 for BTV-9, 1999 for BTV-16 and 2001 for BTV-1). Vector monitoring for Culicoides spp. The methodology of the vector monitoring system is as follows: 1) operation of at least one light trap in each affected area, with preference for affected premises, breeding sites, etc. 2) collection and classification of insects every 15 to 30 days 3) creation of a database. In regard to the trapping protocol, mainland Greece was composed of 59 quadrants of 50 km 50 km 2 (labelled 1-59), which were sampled for Culicoides over two years (Fig. 1). Two farms (at least 10 km apart) were sampled in each quadrant. During the summer (July to October), each farm was sampled for two nights During the winter (December to March), farms where C. imicola was found in the summer were sampled for a further five to seven nights. Results The results of serological monitoring of the sentinel bovines are presented in Table II and Figure 2. In the first incursion of BTV in 1979, BTV was confirmed as serotype 4 (24) and the vector responsible was Culicoides imicola (4, 5, 12). BTV serotype 9 was first isolated in 1998, apparently with C. imicola again involved in transmission (M. Patakakis, personal communication). During the 1999 epidemic, three serotypes were isolated, namely: BTV-4, BTV-9 and, for the first time, BTV-16. The three insect vectors involved were C. imicola, C. obsoletus, C. pulicaris on the eastern Culicoides imicola Culicoides pulicaris Culicoides obsoletus 30 quadrants sampled in quadrants sampled in 2002 Figure 1 Distribution of bluetongue virus vectors in Greece, Veterinaria Italiana, 40 (3),

82 Global situation islands and on mainland Greece. C. imicola was not found in northern Greece (M. Patakakis, personal communication). Only BTV-4 was isolated in the 2000 outbreak and C. imicola was not detected (M. Patakakis, personal communication). In 2001, BTV-1 was isolated for the first time in the western, north-western and eastern regions of Greece, while BTV serotypes 4 and 9 were isolated in north-west and central Greece. In the eastern and central prefectures, C. imicola was involved, while C. obsoletus was involved in the western, north-western and central prefectures (M. Patakakis, personal communication). Results of the BTV serotypes identified are presented in Table III and Figure 3. Table II Results of serological monitoring of sentinel bovines for bluetongue antibodies in Greece, Year No. of prefectures with sentinels Location of prefectures with sentinels Period of sampling starting date End date Results (positive/ total samples tested) Seroconversion months Eastern Greece June December 560/approx South-eastern Greece May 1999 End of September 12/639 August, September South-eastern, eastern, northeastern, north-central, central Greece March 2000 End of December / March until September South-eastern, eastern, northeastern, north western, north-central, central Greece June 2001 June /4 340 September until January North, north-eastern, northwestern, north-central, central Greece June 2002 End of November / South-eastern, eastern, northeastern, north-western, western, north-central, central Greece June 2003 End of September 0/ Seroconversion (colour = year) Figure 2 Serological monitoring of sentinels for bluetongue in Greece, Veterinaria Italiana, 40 (3), 2004

83 Global situation Table III Distribution of bluetongue virus serotypes in Greece, Prefectures Virus isolated Serology Virus isolated Serology Virus Virus Virus Serology Serology Serology isolated isolated isolated Dodekanissa (south-east) BTV-9 BTV-4, BTV-16 Samos (east) Chios (east) BTV-4 Lesbos (north-east) BTV-4 BTV-4 BTV-4, BTV-9 BTV-4, BTV-16 BTV-9, BTV-16 BTV-4, BTV-16 Evros (north-east) BTV-4 BTV-4 Rodopi (north) BTV-4 BTV-9 BTV-9 Drama (north) BTV-9 Kavala (north) BTV-4 Serres (north) BTV-9 Chalkidiki (north) BTV-9, BTV-16 BTV-9 Pieira (north) BTV-4 Grevena (north-west) BTV-1 BTV-4, BTV-9 Kastoria (north-west) BTV-4, BTV-9 Ioannina (north-west) BTV-1 BTV-4, BTV-9 Thesprotia (west) BTV-1 Arta (west) BTV-4 Larisa (mainland) BTV-9 BTV-9 Magnesia (mainland) BTV-4 BTV-4 Evia (mainland) BTV-4 BTV-9 BTV 1 BTV-1 BTV-1 BTV-4 BTV-9 BTV-16 Figure 3 Distribution of bluetongue serotypes in Greece, Veterinaria Italiana, 40 (3),

84 Global situation Results of the distribution of Culicoides are presented in Figure 1 and, finally, results of seroprevalence for 2002 are presented in Figure 4. Seroprevalence (%) N/A 0 < >50 Figure 4 Accumulative seroprevalence (%) of bluetongue in Greece, 2002 Discussion The major epidemics of and 2001 were due to new independent incursions of BTV. In October 1998, after an absence of almost 20 years, BT was confirmed on four islands in the south-east near the western coast of Turkey. Serotype 9 was identified and was the first incursion of this serotype into Greece, although in it had already been found to occur serologically in western and south Anatolian Turkey (21). In August 1999, BTV serotype 4 and serotype 9 were introduced into Greece from the north-east. Initially, BTV-4 outbreaks were reported from near the border with European Turkey but later BTV-4 was distributed more widely by prevailing winds. BTV-9 was isolated initially from outbreaks in northern Greece and later BTV-9 was also isolated more widely. In June 1999, BTV-9 was confirmed in southern and eastern Bulgaria (2), while in July 1999, Turkey reported the incursion of BT into European Turkey in provinces bordering Bulgaria and mainland Greece (13). BTV-4, BTV-9 and BTV-16 were also isolated from south-eastern prefectures in Greece; BTV-4 had previously been reported in Anatolian Turkey (21). BTV-16 occurs regularly in Israel (19, 21). The source of the introduction of BTV-4, BTV-9 and BTV-16 into Greece in 1999 is difficult to determine. BTV serotypes 2, 4, 6, 9, 10, 13 and 16 have been reported in previous years in Anatolian Turkey (21), Syria, Jordan (20) and Israel. The movement of some of these viruses or of others (e.g. Akabane) from east to west is well documented (21, 24). At this time, BTV-2, BTV-6, BTV-10 and BTV-13 have not been isolated in Greece. Work on the serotyping of BTV isolates from previous outbreaks is in progress. In September 2001, BTV serotype 1 invaded northwest Greece along the Albanian border and then subsequently spread to western Greece, although it was limited in its spread due to natural barriers (mountains). In late October 2001, a BTV-1 incursion was reported on the island of Lesbos in the same areas that were severely infected during the 1999 BT epidemic. The isolation of BTV-1 was made from sick animals that had recovered from clinical disease in BTV-1 has never been reported in neighbouring countries or in eastern Turkey, Syria, Jordan or Israel; BTV-1 has been isolated in India (17, 18). Efficient laboratories must be established in neighbouring countries as well as in all Middle Eastern and North African countries so that BTV can be isolated and typed. Overall, in 1979 and between 1998 and 2001, BTV-4 appears to have been the most widely distributed serotype, and was isolated consistently every year. BTV-16 was identified for the first time in 1999 and appears to be limited to the islands of the east Aegean Sea, with the notable exception of isolation from insects on mainland Greece in It seems that several independent incursions of BTV-4, BTV-9, BTV-16 and BTV-1 occurred in Greece between 1998 and The islands of the east Aegean Sea were exposed to repeated incursions throughout the period 1979 to Results of the nationwide seroprevalence survey undertaken in 2002 are shown in Figure 4 indicating that despite repeated waves of disease, a large proportion of uninfected susceptible animals remain. Results of vector monitoring indicate there is an abundant and persistent presence of competent and/or potential BTV vectors in most areas of Greece. Culicoides imicola, in particular, is abundant in south and eastern Greece and was not found in the north, north-west and west. Considering all the problems associated with the use of the only commercially available BTV vaccines manufactured in South Africa (13), the Greek Ministry of Agriculture refused to apply vaccination against BTV. The nationwide seroprevalence survey indicates that large parts of Greece remain free of BTV infection, which suggests that the control and safeguard measures applied were appropriate and efficient. 114 Veterinaria Italiana, 40 (3), 2004

85 Global situation Acknowledgements The authors are grateful to F. Gioka for her excellent technical assistance from 1979 until today. Gratitude is extended to M. Patakakis for useful discussions on vectors and for providing the excellent Culicoides maps. Our deepest thanks are extended to P.S. Mellor for fruitful discussions. References 1. Anderson J. (1984). Use of monoclonal antibody in a blocking ELISA to detect group specific antibodies to bluetongue virus. J. Immunol. Methods, 74, Anon. (1999). Bluetongue in Bulgaria: additional information. Dis. Info., 12, Billinis C., Koumbati M., Spyrou V., Nomikou K., Mangana O., Panagiotidis C.A. & Papadopoulos O. (2001). Bluetongue virus diagnosis of clinical cases by a duplex reverse transcription PCR: a comparison with conventional methods. J. Virol. Methods, 98 (1), Boorman J.P.T. (1986). Presence of bluetongue virus vectors on Rhodes. Vet. Rec., 118, Boorman J.P.T. & Wilkinson P.J. (1983). Potential vectors of bluetongue in Lesbos, Greece. Vet. Rec., 113, Dragonas P.N. (1981). Evolution of bluetongue in Greece. OIE Monthly Circ., 9 (417), Goldsmit L. & Barzilai E. (1968). An improved method for the isolation and identification of bluetongue virus by intravenous inoculation of embryonating chicken eggs. J. Comp. Pathol., 78, Herniman K.A.J., Boorman J.P.T. & Taylor W.P. (1983). Bluetongue virus in a Nigerian dairy cattle herd. Serological studies and correlation of virus activity to vector population. J. Hyg., Camb., 90, Jochim M.M., Barber T.L. & Bando B.M. (1974). Identification of bluetongue and epizootic hemorrhagic disease viruses by indirect fluorescent antibody procedure. In Proc. 17th Annual Meeting of the American Association of Veterinary Laboratory Diagnosticians (AAVLD). AAVLD, Ames, Koumbati M., Mangana O., Nomikou K., Mellor P.S. & Papadopoulos O. (1999). Duration of bluetongue viraemia and serological responses in experimentally infected European breeds of sheep and goats. Vet. Microbiol., 64 (4), Mastroyanni M., Axiotis I. & Stoforos E. (1981). Study of the first outbreak of bluetongue disease in sheep in Greece. Bull. Hell. Vet. Ass., 32, Mellor P.S., Jennings M. & Boorman J.P.T. (1984). Culicoides from Greece in relation to the spread of bluetongue virus. Rev. Elev. Méd. Vet. Pays Trop., 37, Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean Basin Vet. J., 164 (1), Office International des Épizooties (OIE) (2000). Bluetongue, Chapter In Manual of standards for diagnostic tests and vaccines. OIE, Paris, Office International des Épizooties (OIE) (2000). Sheep and goat pox, Chapter In Manual of standards for diagnostic tests and vaccines. OIE, Paris, Pearson J.E. & Jochim M.M. (1979). Protocol for the immunodiffusion test for bluetongue. In Proc. 22nd Annual Meeting of the American Association of Veterinary Diagnosticians (AAVLD). AAVLD, Ames, Prasad G., Jain N.C. & Gupta Y. (1992). Bluetongue virus infection in India: a review. Rev. Sci. Tech. Off. Int. Épiz., 11 (3), Prasad G., Garg A.K., Minakshi, Kakker N.K. & Srivaztava R.N. (1994). Isolation of bluetongue virus from sheep in Rajasthan State, India. Rev. Sci. Tech. Off. Int. Épiz., 13 (3), Shimshony A.S., Barzilai E., Savir D. & Davidson M. (1988). Epidemiology and control of bluetongue disease in Israel. Rev. Sci. Tech. Off. Int. Épiz., 7 (2), Taylor W.P. (1987). Bluetongue in Syria and Jordan. In Bluetongue in the Mediterranean region, (W.P. Taylor, ed.) Commission of the European Communities, Brussels and Luxembourg, Taylor W.P. & Mellor P.S. (1994). Bluetongue virus distribution in Turkey, Epidemiol. Infec., 112, Thevasagayam J.A., Wellby M.P., Mertens P.P.C., Burroughs J.N. & Anderson J. (1996). Detection and differentiation of epizootic haemorrhagic disease of deer and bluetongue viruses by serogroup-specific sandwich ELISA. J. Virol. Methods, 56, Thevasagayam J.A., Woolhouse T.R., Mertens P.P.C., Burroughs J.N. & Anderson J. (1996). Monoclonal antibody based competitive ELISA for the detection of antibodies against epizootic haemorrhagic disease of deer virus. J. Virol. Methods, 57, Vassalos M. (1980). Cas de fièvre catarrhale du mouton dans l île de Lesbos (Grèce). Bull. Off. Int. Épiz., 92, Veterinaria Italiana, 40 (3),

86 Global situation Vet. Ital., 40 (3), Bluetongue in Israel a brief historical overview A. Shimshony Koret School of Veterinary Medicine, Faculty of Agriculture, Hebrew University, Rehovot, Israel Summary Bluetongue (BT) was first observed in Israel in The author presents a brief history of BT in Israel and of selected topics which might be useful for the planning of control and prevention policies in newly affected areas within the Mediterranean Basin. Systematic epidemiological, virological and entomological monitoring has been ongoing in Israel since BT virus serotypes 2, 4, 6, 10 and 16 have been identified. The seasonality of the disease, susceptibility of selected sheep breeds, resistance of goats and of native sheep breeds, high seroconversion rate (combined with the absence of clinical signs or affected fertility) in bovines and the impact of vaccination with live-attenuated vaccines on pregnant sheep are presented and discussed. Keywords Bluetongue Culicoides History Israel Virus. History Some authors believe that bluetongue (BT) existed in the Middle East as early as 1924 (5, 9). Initially, the disease was clinically suspected in sheep in Israel in 1943/1944, affecting some 20 flocks in northern districts, with morbidity ranging from 2-55% (mean <10%) and mortality 1-8%. At about the same time, cases were reportedly suspected in Cyprus (severe), Turkey and Syria. There was no laboratory involvement in the said suspected outbreak in Israel (8). Between September and November 1950, sporadic, moderate clinical cases were reported in northern Israel by Komarov-Kimron and Goldsmit (10) in cattle (68 foci); they reported more severe cases in sheep (16 foci). A BT virus (BTV) was isolated by them and later identified at Onderstepoort by Howell as BTV-4. The affected cattle were reportedly of European and North American origin, imported several months previously. The sheep most severely affected were Sardinian, imported into Israel in October Local sheep breeds were much less affected (10). In the summer of 1963, BT was suspected, in two sites (north and south Israel), in German mutton Merino sheep, imported into Israel one to two years previously. No virological laboratory investigations were conducted but, based upon detailed recorded clinical and post-mortem observations, these outbreaks may, with certainty, be retrospectively attributed to BTV. The following year, 1964, a major outbreak, characterised by typical BT symptoms, was recorded in northern Israel, affecting a population of approximately sheep of the exotic (European) breeds and their crosses, in 47 villages. The reported morbidity was 20.5% and the case-fatality rate 31.4%. The outbreak began at the end of June and died out in December. The highest losses were encountered in the Taanach area (Valley of Yezreel), where in 7 villages with 134 small flocks of German mutton Merino sheep, imported two to three years previously from Germany a morbidity rate of 39.5% and case-fatality rate of 45.6% were recorded (11). Typically, no cases in lambs younger than four months and no clinical cases in cattle were recorded. BTV-4 was isolated and identified by the Kimron Veterinary Institute (KVI), and later confirmed by the Onderstepoort Laboratory. The KVI developed the intravenous embryonated egg technique for BTV isolation and identification (6, 7) later to become the international method of choice. Since 1964, BT has been identified and reported in Israel each year, either following virus isolations from sampled clinical cases in unvaccinated sheep of susceptible breeds, or by serological tests in cattle without clinical signs of disease. Only laboratoryconfirmed cases are published. Brief summary of selected observations, BTV appears throughout Israel each year, from July to December, except in the southern Arava (arid) 116 Veterinaria Italiana, 40 (3), 2004

87 Global situation region. Single exceptional cases have been recorded in January and June. The geographical distribution has been investigated, based mainly on serological surveillance in young sentinel cattle in various regions, sampled in June and December. This surveillance had been operational since 1980 but was discontinued after The incidence of clinical disease depends also upon man-related factors such as vaccination rates and the presence of susceptible breeds. Therefore, the incidence of clinical cases in sheep is unrelated to the prevalence of the virus in a given year. It varies from year to year, averaging 13 to 14 cases per annum. Throughout the period from 1968 to 1998, a total of 413 (laboratory confirmed) outbreaks were reported. Only in 1980 and 1981 was not a single clinical case reported in sheep (though bovines did seroconvert during these years). Observations from 1980 to 1985 showed a good correlation between the BTV serotypes from cattle (serology), sheep (isolations from clinical cases) and Culicoides (isolation) (13). The disease may be observed in cattle. Such cases seem to be extremely rare, involving 6- to 12-monthold dairy animals but with very mild symptoms (elevated temperature, hyperaemia with slight necrosis of the oral papillae, coronitis). Five serotypes of BTV have been identified to occur in Israel (first year of identification in brackets): BTV-4 (1950), BTV-10 (1965), BTV-16 (1966), BTV-6 (1972), and BTV-2 (1973). Since 1973, no new serotypes have been added to the list. The most prevalent is BTV-4 (13). Typically, in most years only one serotype has been encountered in clinically affected sheep and in seroconverted cattle, although there have been years when several or all five serotypes were identified in field samples. This has led to the view that new introductions of infected vectors from elsewhere might be involved in the epidemiology of the disease in Israel in certain years. Regular vaccination against BT was not practised in Israel until 1964, when a polyvalent (serotypes 1 to 14) South African live-attenuated vaccine was applied. Over 50% of the vaccine was used in infected flocks; it was claimed that the spread of the disease was generally halted eight days post vaccination. In May 1965, a monovalent, live-attenuated BTV-4 vaccine (prepared in liquid form from seed material obtained from Onderstepoort) was used. Some months later, cases in unvaccinated sheep were observed, from which BTV-10 was isolated and identified. Consequently, the use of the polyvalent Onderstepoort vaccine was resumed; this was discontinued in Since then, a quadrivalent, liveattenuated vaccine (serotypes 2, 4, 6, 10), obtained from Onderstepoort has been in use annually to vaccinate sheep of exotic, susceptible breeds. No local Awassi sheep, goats or cattle have been included in the vaccination scheme. In 1965/1966, pregnant Merino sheep imported from Germany were vaccinated upon arrival with the polyvalent vaccine. Breeders of the imported sheep complained about their reduced lambing rate. Based upon the results of an experimental vaccination trial in 1968/1969 (in susceptible young Merino ewelambs imported for this purpose from Germany), it was concluded that primary vaccinations of susceptible sheep during the first half of pregnancy could cause foetal death (up to 40%), though no central nervous system damage or malformations were observed in their offspring. No such deleterious effect was observed following revaccination of the same animals during the same stages (6th week) of their subsequent pregnancy. Consequently, primiparae are not vaccinated in Israel during pregnancy (12). In another observation, it was concluded that (naturally acquired) viraemia of pregnant young dairy cattle, caused by BTV-4 and/or BTV-16, did not affect their foetuses. It should be said that in spite of a seroconversion rate, which may exceed 80% in the cattle of certain districts during certain years, BTV has never been correlated with fertility problems in Israeli cattle (4). Although identified in Israel since 1966, BTV-16 was initially not included in the vaccine, because an attenuation of this serotype was not performed at Onderstepoort at the time. In addition, no involvement of this serotype was recorded in sheep between 1974 and 1992 (though it was identified serologically in healthy, seroconverted cattle in central-southern Israel, e.g. in 1982). However, during 1993, BTV-16 was found to be involved in serious outbreaks in sheep, including, in several cases, local Awassi sheep, usually regarded as refractile to clinical BT. During 1994, this phenomenon was repeated, with a larger number of outbreaks, some of them involving sheep vaccinated with the standard quadrivalent vaccine. Subsequently, the Onderstepoort laboratory kindly agreed to attenuate BTV-16 in order to include it in the vaccine. The attenuated vaccine strain was initially put to use in At present, a pentavalent Veterinaria Italiana, 40 (3),

88 Global situation (serotypes 2, 4, 6, 10 and 16) vaccine is applied annually on a voluntary basis in susceptible, nonpregnant sheep but the number of vaccinated animals is rather small: between 2000 and 2002, only per year. As previously stated, other species have never been vaccinated in Israel. Experimental infection trials have been performed in Israel in goats and gazelles, in addition to serological/virological surveillance in the said animals and in ibex (2, 3). Entomological surveillance is in progress. References 1. Barzilai E. & Tadmor A. (1971). Multiplication of bluetongue virus in goats following experimental infection. Ref. Vet., 28, Barzilai E., Tadmor A. & Shimshony A. (1971). Natural bluetongue infection in the mountain gazelle (Gazella gazella). Ref. Vet., 28, Barzilai E. & Tadmor A. (1972). Experimental infection of the mountain gazelle (Gazella gazella) with bluetongue virus. Ref. Vet., 29, Barzilai E., Bar-Tana U., Cohen R., Eyal Y., Shimshony A. & Trainin Z. (1975). Natural infection of pregnant cattle with the virus of bluetongue: the effect on their progeny. Ref. Vet., 32, Gambles R.M. (1949). Bluetongue of sheep in Cyprus. J. Comp. Pathol., 59, Goldsmit L. & Barzilai E. (1965). Isolation and propagation of a bluetongue virus strain in embryonating chicken eggs by the intravenous route of inoculation preliminary report. Ref. Vet., 22, Goldsmit L. & Barzilai E. (1968). An improved method for the isolation and identification of bluetongue virus by intravenous inoculation of embryonating chicken eggs. J. Comp. Pathol., 78, Goor S. (1950). Bluetongue of sheep in Palestine. Ref. Vet., 7, Howell P.G. (1963). Bluetongue. In Emerging diseases of animals. FAO, Rome, Agricultural Studies, 61, Komarov A. & Goldsmit L. (1951). A disease, similar to bluetongue in cattle and sheep in Israel. Ref. Vet., 8, Shimshony A. (1964). An outbreak of bluetongue in sheep in the Taanakh area of northern Israel in Ref. Vet., 21, Shimshony A., Goldsmit L. & Barzilai L. (1980). Bluetongue in Israel. Bull. Off. Int. Épiz., 92 (7-8), Shimshony A., Barzilai E., Savir D. & Davidson M. (1988). Epidemiology and control of bluetongue disease in Israel. Rev. Sci. Tech. Off. Int. Épiz., 7 (2), Veterinaria Italiana, 40 (3), 2004

89 Vet. Ital., 40 (3), Global situation Serological and clinical evidence of a teratogenic Simbu serogroup virus infection of cattle in Israel, J. Brenner (1), T. Tsuda (3), H. Yadin (2), D. Chai (2), Y. Stram (2) & T. Kato (3) (1) Prevention of Neonatal Diseases Unit, Kimron Veterinary Institute, Beit Dagan, Israel (2) Division of Virology, Kimron Veterinary Institute, Beit Dagan, Israel (3) Clinical Virology Section, Kyushu Research Station, National Institute of Animal Health, Chuzan 2702, Kagoshima, Japan Summary During the last 35 years, two major outbreaks of Akabane virus (AKAV) infection were recorded in cattle in Israel in 1969/1970 and 2002/2003. Congenital malformations of calves characterised by the appearance of an arthrogryposis and hydranencephaly syndrome first appeared in Israel in Based on epidemiological, clinical, pathological, histopathological and serological data, this syndrome was strongly correlated with seroreactivity to AKAV, a member of the Bunyaviridae, Simbu serogroup. In February 2002, the first cases of blind newborn calves (BNC) were observed on farms located in the northern valleys of Israel. Microtitre serum neutralisation (SN) tests of serum from malformed calves and their dams were conducted using Akabane and Aino viruses (AINOV). The first SN test was performed at the reference laboratory of the Clinical Virology Section, Kyushu Research Station, National Institute of Animal Health, Kagoshima, Japan. The clear-cut findings of seroreactivity to AKAV by cattle located in the affected zone, in contrast to negative findings in cattle from unaffected farms (87% and 3.7%, respectively) was indicative of AKAV infection. In contrast, seroreactivity to Aino virus was relatively low in both affected and non-affected areas during the 2002 outbreak. In order to establish Israeli laboratory standards for Simbu serogroup diagnosis, 57 serum samples tested by the Japanese laboratory were retested by SN in Israel. An almost complete homology (96.5%) was found between the two SN panels of sera (kappa = 0.92). SN and ELISA kits enabled the surveillance of this arbovirus epidemic in the second consecutive year (2003). Moreover, AKAV was identified in trapped midges by hemi-nested PCR and real-time PCR. With these techniques, the geographical limits of the BNC epidemic that appeared in some areas of Israel was identified for the first time and was recorded in the Arava Rift Valley, 400 km south of the epicentre of the 2002 outbreak. The reintroduction of AKAV into this region, together with some evidence of AINOV activity and epidemics of bluetongue (BT) in the southern parts of Europe and the eastern Mediterranean, and renewed outbreaks of West Nile virus infection in Israel, Italy and southern France, are all evidence of the potential spread of arbovirus activity into southern Europe from the Mediterranean Basin. Keywords Aino virus Akabane virus Arthrogryposis Congenital malformations Hydranencephaly Israel Neonatal ruminants Simbu serogroup. Introduction Akabane virus (AKAV) and Aino virus (AINOV) are known to cause epidemics of abnormal parturitions in cattle, including abortion, stillbirths and calf deformities. This is known as the congenital arthrogryposis-hydranencephaly syndrome (CAHS) affecting the musculo-skeletal and nervous systems, respectively (5, 8, 9). AKAV was originally isolated from mosquitoes and from the midge, Culicoides imicola, which is now considered to be the major vector (1, 5, 7). Serological studies have since classified AKAV and AINOV in the Simbu Veterinaria Italiana, 40 (3),

90 Global situation serogroup of the family Bunyaviridae (7). Congenital disease associated with AKAV is found principally at the extremities of its normal geographical distribution. Outbreaks of congenital malformations in dairy calves characterised by the CAHS were first recorded in Israel in 1969 and persisted until 1970 (6, 10, 12, 14). Similar malformations were seen in lambs and kids (14). Based on epidemiological (10), clinicopathological (10, 12), histopathological (12) and serological data (6), CAHS was diagnosed as being caused by AKAV (17). Thirty-two years later, in February 2002, the first cases of blind newborn calves (BNC) (Fig. 1) appeared in two neighbouring large dairy herds in northern Israel. During the four months that followed, dozens of farms had reported similar problems (Figs 1 and 2). All the BNC cases presented hydro- or micro-hydranencephalic signs, or both, on post mortem. In these cases, the cranium of the necropsied BNC contained an apparently normal cerebellum, but instead of the cerebrum there remained only 5-10 g of tissue encasing ml of liquid (Figs 1 and 2) (3). the southernmost point recorded in the 1969/1970 outbreak (10) (Fig. 4). While the 1969/1970 outbreak was contained above the latitude of (10), in 2003, BNC was noted on dairy farms located very close to the Red Sea (Fig. 3) about 400 km from the 2002 epicentre. Figure 2 A brainless hydranencephalic blind calf This paper describes the diagnosis of the AKAV infection in 2003 and defines the geographic limits of its spread. Materials and methods Figure 1 A blind neonatal calf with micropthalmia Micro-SN for AKAV and AINOV was performed in the Laboratory of the Clinical Virology Section of the Kyushu Research Station, National Institute of Animal Health, Kagoshima; AKAV was identified as responsible for the newly emerging epidemic. This conclusion was based on the clear serological findings of AKAV in the affected zone in contrast to the negative findings in cattle from unaffected zones during the 2003 outbreak (87% and 3.7%, respectively) (3). This 2002 outbreak persisted until the end of April 2002, suggesting that the activity of AKAV began in August 2001 and ended prior to November the same year (3). In February 2003, BNC was recorded again and this time the syndrome was observed throughout Israel (Fig. 3). Moreover, BNC appeared in 2003 beyond Defining the geographical boundaries of the epidemic Between February and May 2003 the 25 settlements that reported the birth of blind calves were located along the inner coastal plain, the Negev and the Arava Rift Valley. All these areas were below latitude and had not reported neonatal malformations during the 2002 epidemic (Figs 3 and 4). The blocking ELISA for Simbu serogroup antibodies was used to test serum samples from twelve of these settlements. Serology Serum neutralisation test A panel of 37 positive and 20 negative serum samples was chosen on which the diagnostic test used was established (Table I). These samples were tested previously by the Japanese laboratory to identify the causative agent responsible for the outbreak of congenital malformations seen in 2002 (3). Four of the positive sera included two precolostral sera of affected newborn calves and two from their dams. To compare results with those of the Japanese Reference Laboratory, an agreement test was performed and the kappa value was 120 Veterinaria Italiana, 40 (3), 2004

91 Global situation calculated. SN tests for AKAV and AINOV are described elsewhere (3). serogroup antibodies (Elizabeth Macarthur Agricultural Institute, Australia). The procedure was conducted in accordance with the instructions of the manufacturer (5). Table I Agreement between the serum neutralisation test results on a panel of serum samples taken during the outbreak of congenital malformations, February-May 2002 National Institute of Animal Health, Japan Location Seropositive Seronegative Total Kimron Veterinary Institute, Israel Kimron Veterinary Institute, Israel Seropositive Seronegative Serological and/or clinical evidence of Akabane virus infection recorded in 2002 Arava Rift Valley Figure 3 The epicentre of blind neonatal calves in the 2002 outbreak Serological and/or clinical evidence of Akabane virus infection recorded in 2003 Figure 4 The epicentre of blind neonatal calves in the 2003 outbreak Enzyme-linked immunosorbent assay The extent of the Simbu serogroup infection in Israel from February until May 2003 was determined by employing a blocking ELISA for Simbu Total Polymerase chain reaction and real-time PCR PCR and real-time PCR for AKAV, AINOV and Simbu group viruses were developed by Stram et al. (15, 16) and were used to identify and sequence AKAV from Culicoides (Fig. 5). Results Defining the geographical boundaries of the epidemic The latitude of was the southernmost line of the 2002 epidemic in Israel. From May 2002 to February 2003, no additional cases were reported in Israel. In February 2003, BNC reappeared south of this latitude (Fig. 4) reaching the Arava Rift Valley (Fig. 4), 400 km south of the 2002 epicentre. Moreover, the 2003 BNC outbreak appeared beyond the southernmost point recorded in the 1969/1970 epidemic (10) in areas where it had never previously been recorded. The 1969/1970 outbreak was contained above (10). In 2002 and 2003, BNC was noted on dairy farms located very close to the Red Sea (Fig. 4). In 2003, BNCs were also recorded from a very limited number of settlements in the northern part of Israel, and in some of them it appeared twice in two consecutive years (data not shown). Serology Serum neutralisation test Table I summarises the results of the comparison made between the SN tests that were conducted in Israel and in Japan. An almost complete homology (96.5%) was found between the two SN test locations with a kappa value of Veterinaria Italiana, 40 (3),

92 Global situation Delta Rn Cycle Figure 5 The real-time polymerase chain reaction of Akabane virus isolated from Culicoides biting midges Enzyme-linked immunosorbent assay Simbu serogroup antibodies were demonstrated in all 12 settlements, using the blocking ELISA. Polymerase chain reaction and real-time PCR Figure 5 depicts the real-time PCR results and the sequences (Fig. 6) of the Israeli AKAV recovered from trapped Culicoides (15, 16). Conclusion This paper describes the methods that were used to diagnose the causative agent of neonatal ruminant malformations in the northern part of Israel in 2002 and to trace its progression to other zones in the second year of the epidemic. The rationale behind the methodology employed to investigate the epidemic is described elsewhere (2). The clear findings of seroreactivity to AKAV in the affected zones are very suggestive of the causative effect of AKAV during the 2002 epidemic, and are consistent with the main period of activity of C. imicola between August and November in this region (3). The geographical distribution of the 2002 outbreak that was first noted in northern Israel was similar to the 1969 outbreak (10). During the second year of these two epidemics, the disease expanded southwards. During the 2003 epidemic, BNC appeared beyond the southernmost point recorded in the 1969/1970 outbreak (10) (Fig. 3). The 1969/1970 outbreak was contained above the latitude of (10), while in 2003, BNC was detected in some dairy farms located very close to the Red Sea at (Fig. 4) about 400 km south of the 2002 epicentre. The demonstration for the first time of AKAV in trapped C. imicola (15, 16) strengthens the assumption that teratogenic arboviruses were involved in epidemics and were probably involved in some of the sporadic births of defective neonatal ruminants in our region. 1 ATGGCAAATc aatttatttt caacgatggt ccacaacgga ATGCAGCTAC 51 ATTTAACCCG GATGCAGGGT ATGTGGCATT TATCAGTAAG TATGGGCAGG 101 AGCTCAACTT TACTGTTGCT AGAGTCTTCT TCCTCAAGGA GAAGAAGGCC 151 AAGATGGTCT TACATAAGAC GCCACAACCA AGTGTCGATC TTACTTTTGC 201 AGGGGTCAAA TTTACAGTGG TTAATAACCA TTTTCCCCAG TACACTGCAA 251 ATCCAGTGTC AGACACTGCC TTTACGCTCC ATCGCATCTC GGGCTACTTA 301 GCTCGCTGGG TTGCTGGAGCA GTGCAAGGCT AATCAGATCA AATTCGCAGA 351 GGCAGCTGCC ACAATTGTGA TGCCGCTGGC TGAGGTGAAG GGTTGCACCT 401 GGAGTGATGG GTATGCAATG TACCTAGGAT TTGCCCCTGG TGCTGAGATG 451 TTTTTGGAAA CCTTTGAGTT TTACCCACTG GTTATCGACA TGCACCGTGT 501 GCTAAAGGAT GGGATGGATG TCAACTTCAT GAGAAAGGTC TTGCGCCAGA 551 GGTACGGGCA GCTGACTGCA GAGGAGTGGA TGACATCTAA GTTGGATGCA 601 GTCAAGGCTG CATTTAGCTC AGTTGCCCAG ATATCCTGGG CCAAATCTGG 651 TTTCTCACCT GCAGCAAGAG CTTTCCTGGC TCAATTTGGT ATCCAGATCT 701 AAT Figure 6 The sequence of the S segment of the Israeli Akabane virus lineage ISR Veterinaria Italiana, 40 (3), 2004

93 Global situation Stram et al. (16) proposed the presence of a new AKAV lineage in Israel, different from those published in Japan and Australia. It would be beneficial to define the new lineage or to confirm its similarities with the African lineage, but unfortunately, the African lineage genome segment S has not yet been published. It was surprising to find relatively high seroreactivity towards AINOV in 29.6% of the sera that originated from the zone that served as controls during the 2002 epidemic. These sera were also seronegative to AKAV (3). In light of the maximal life span of the average Israeli dairy cow, it appears that the initial infection with AINOV must have occurred some five to six years ago. It is not known whether AKAV is endemic or sporadic in Israel. Other unanswered questions are whether AKAV infection appears alone or in combination with other arbovirus(es), and whether different viruses in the Simbu group have invaded Israel. In this study, the authors have confirmed that the probable (major) vector of AKAV in Israel and in the eastern part of the Mediterranean Basin (11) is Culicoides imicola; this vector is present in the Old World in a belt between 35 S and 40 N (11). On the other hand, the source of AKAV (or of any other arbovirus included in the Simbu group) has not yet been identified. The distribution of Culicoides spp. depends on climatic (macro-global and microregional) changes and reasons for its presence or absence in certain defined zones are as yet unsolved. This probable reintroduction of AKAV into this region, the evidence of the presence of AINOV and the epidemics of BT virus in southern Europe and the eastern Mediterranean Basin (11), together with the potential alert to the presence of West Nile virus in Israel (13) and Italy (4), all provide evidence of the potential and actual spread of arboviruses into previously uninfected areas. References 1. Braverman Y. & Chechik F. (1996). Air streams and the introduction of animal diseases borne on Culicoides (Diptera, Ceratopogonidae) into Israel. Rev. Sci. Tech. Off. Int. Épiz., 15, Brenner J., Malkinson M. & Yadin H. (2004). Application of diagnostic procedures to epidemiological situations with special reference to arboviral infections. In Bluetongue, Part II (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (4), Brenner J., Tsuda T., Yadin H. & Kato T. (2004). Serological evidence of Akabane virus infection in northern Israel in J. Vet. Med. Sci., 66 (4), Cantile C., Di Guardo G., Eleni C. & Arispici M. (2000). Clinical and neurological features of West Nile equine encephalomyelitis in Italy. Equine Vet., 32, Della-Porta A.J., White J.R., Gard G.P. & Kirkland P.D. (1993). Akabane disease: histopathology, virology and serology. In Australian standard diagnostic techniques for animal diseases (L.A. Corner & T.J. Bagust, eds). CSIRO Information Services, Melbourne, Kalmar E., Peleg B.A. & Savir D. (1975). Arthrogryposis-hydranencephaly syndrome in newborn cattle, sheep and goats: serological survey for antibodies against Akabane virus. Refuah Vet., 32, Kinney R.M. & Calisher C.H. (1981). Antigenic relationship among Simbu serogroup (Bunyaviridae) viruses. Am. J. Trop. Med. Hyg., 130, Kurogi H., Inaba Y., Takahashi E., Sato K., Satoda K., Goto Y., Omori T. & Matumoto M. (1977). Congenital abnormalities in newborn calves after inoculation of pregnant cows with Akabane virus. Infec. Imm., 17, Liao Y.K., Lu Y.S., Goto Y. & Inaba Y. (1996). The isolation of Akabane virus (Iriki) from calves in Taiwan. J. Basic Microbiol., 1, Markusfeld-Nir O. & Mayer E. (1971). An arthrogryposis/hydranencephaly syndrome in calves in Israel-1969/70 epidemiological and clinical aspects. Refuah Vet., 28, Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean basin Vet. J., 164, Nobel T.A., Klopfer-Orgad U. & Neumann F. (1971). Pathology of an arthrogryposishydranencephaly syndrome in domestic ruminants in Israel: 1969/70. Refuah Vet., 28, Perl S., Fiette L., Lahav D., Sheichat N., Banet C., Orgad U., Stram Y. & Malkinson M. (2002). West Nile virus encephalitis in horses in Israel. Isr. J. Vet. Med., 57, Shimshony A. (1980). An epizootic of Akabane disease in bovines, ovines and caprines in Israel, : epidemiological assessment. Acta Morphol. Acad. Sci. Hung., 28, Stram Y., Brenner J., Braverman Y., Banet-Nuack C., Kuznetzova L. & Guini M. (2004). Akabane virus in Israel: a new virus lineage Virus. Res., 104 (1), Stram Y., Kuznetzova L., Guini M., Rogel A., Meirom R., Chai D., Yadin H. & Brenner J. (2003). Detection and quantitation of Akabane and Aino viruses by multiplex real-time reverse-transcriptase PCR. J. Virol. Methods, 116 (2), Trainin Z. (1971). Active production of immunoglobulins in calf embryos. Isr. J. Med. Sci., 7, Veterinaria Italiana, 40 (3),

94 Global situation Vet. Ital., 40 (3), Epidemiological surveillance of bluetongue in Sicily S. Caracappa (1), M. Bagnato (2), P. Sghembri (2), A. Guercio (3), F. Prato (4), G. Tumino (5), A. Migliazzo (6), F. Geraci (6), S. Vullo (6), S. Agnello (6) & C. Di Bella (6) (1) Direzione Sanitaria, Istituto Zooprofilattico Sperimentale della Sicilia A. Mirri, Via G. Marinuzzi 3, Palermo, Sicily, Italy (2) Ispettorato Regionale Veterinario, Assessorato Regionale alla Salute Piazza Ottavio Ziino, Palermo, Sicily, Italy (3) Area di Diagnostica Virologica, Istituto Zooprofilattico Sperimentale della Sicilia A. Mirri, Via G. Marinuzzi 3, Palermo, Sicily, Italy (4) U.O. Sistema Informativo e Statistica, Istituto Zooprofilattico Sperimentale della Sicilia A. Mirri, Via G. Marinuzzi 3, Palermo, Sicily, Italy (5) Area Ragusa, Istituto Zooprofilattico Sperimentale della Sicilia A. Mirri, C.da Nunziata, Km 1,3, S.P. 40, Ragusa, Sicily, Italy (6) Area Sorveglianza Epidemiologica, Istituto Zooprofilattico Sperimentale della Sicilia A. Mirri, Via G. Marinuzzi 3, Palermo, Sicily, Italy Summary The authors describe the status of bluetongue (BT) since 13 October 2000, when the first outbreak was reported in Sicily. The results of the epidemiological surveillance programme, based on sentinel animals distributed over the entire region, are also given. In Sicily, the incidence of the disease is relatively low compared to some other areas in the Mediterranean Basin. Seventy-five outbreaks of the disease were recorded in the first three epidemics (October 2000 to May 2003). Overall morbidity was 13.25%, mortality 5.36% and the case fatality rate 41.49%. The Province of Catania seems to have been the worst affected; the incidence rate in August 2002 was 0.8%. The monthly incidence rate was calculated for sentinel animals of which the estimated total was 3 654, distributed in 63 areas. It is important to underline that in the period under consideration, a total of animals was examined. During the surveillance period, which extended from September 2001 to May 2003, the incidence of BT peaked in September 2002, at 5.91% ± The cumulative incidence rate from September 2001 to August 2002 and September 2002 to March 2003 was 4.53% ± 0.76 and 20.03% ± 1.85, respectively. The circulation of BT virus serotypes 2, 4, 9 and 16 is described, as revealed by seroconversion in sentinel animals. Keywords Bluetongue Epidemiology Italy Orbivirus Outbreaks Sentinel animals Serotypes Sicily. Introduction The aim of this study is to provide a description of the first three epidemics of bluetongue (BT) in Sicily, using the principal epidemiological indicators. The spread of infection was monitored through a control group of sentinel animals with seroconversions tracked through a data bank. Following the first appearance of BT during the summer of 2000, the disease spread from the islands and the southern tip of Italy up into the centre of the mainland and into parts of the northern regions (4). The presence of a biological vector through which the infection spreads, means that the distribution of the disease is strongly linked to geo-climatic conditions. During the summer and autumn of 2000, Italy was struck by the most severe epidemic of BT ever recorded in Europe, affecting Sardinia, Sicily and Calabria, i.e. in regions where 54.7% of the entire Italian sheep and goat populations are located. During the second epidemic (16 May 2001 to 13 April 2002) the virus spread to the north. The number of regions affected rose to seven, namely: Tuscany, Lazio, Basilicata, Campania, Calabria, Sicily and Sardinia. The third epidemic (15 April 2002 to 14 May 2003) affected eight regions, namely: Basilicata, Puglia, Campania, Lazio, Molise, Calabria, Sardinia and Sicily (8). 124 Veterinaria Italiana, 40 (3), 2004

95 Global situation Since attempting to combat vectors is not realistic, methods of preventing BT are essentially based on preventive measures using live-attenuated vaccines (2, 5, 6). To date, no country in the world affected by the disease has been able to definitively eradicate the infection except in rare instances where it has spontaneously disappeared. In September 2001, vaccination against BTV serotypes 2 and 9 was initiated throughout Italy. In Sicily, vaccination against BTV-2 commenced in October 2001 in the Provinces of Palermo, Agrigento and Trapani (7). Materials and methods As soon as the disease was reported in Sicily, a surveillance network was established throughout the region by the Azienda Sanitaria Locale (ASL) to swiftly identify any symptoms that might indicate the presence of BT. Between October and December 2000, sheep and goat farms were visited, totalling animals. Clinical surveillance in the affected areas was detailed and intensive. The sentinel animals were spread throughout the entire region, which was subdivided into areas of 400 km 2 with at least 58 sentinel animals in each area. A commercially available c-elisa kit was used to conduct tests, and was replaced in June 2003 by a c- ELISA product from the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche). All the sera that gave positive results to the c-elisa were confirmed by CESME using their method of seroneutralising with reference viral structures belonging to serotypes 2, 4, 9 and 16. The following epidemiological indicators were used to describe the infected areas: incidence rate, morbidity, mortality rate and fatality rate as calculated according to Thrusfield (9). MapInfo Version 7.0 was used to obtain geographical figures and plot thematic maps, whilst Epiinfo 2000 was used to measure frequency and tendencies to variation. The database was set up by Sigla (Sistema Informativo per la Gestione dei Laboratori di Analisi). For comparisons between percentages, the critical ratio (Z) test (3) was used with Software EpiCalc 2000 Version Results Disease outbreaks In Sicily, the first outbreak was registered in the Giardinello area of the Province of Palermo on 13 October 2000 (1). From that date until 28 November 2000 (date of the last outbreak in 2000), 16 outbreaks, all involving BTV-2, were recorded (Fig. 1): 12 in the Province of Palermo, 2 in Agrigento and 2 in Trapani. The clinical situation was most serious in the Province of Palermo. During the first epidemic, from 13 October 2000 to 15 May 2001, the level of morbidity was 9.93%, with a mortality rate of 5.92% and a fatality rate of 59.60% (Table I and Fig. 2). The second epidemic, dating from 16 May 2001 to 13 May 2002, comprised six outbreaks (Fig. 1): two in the Province of Messina and four in the Province of Siracuse. Morbidity was recorded at 12.61%, mortality at 2.75% and the fatality rate at 21.80%, with an overall decrease in all epidemiological indicators compared to the first epidemic (Table I; Fig. 2). During the second epidemic, 4.87% morbidity was recorded in the Province of Siracuse, with mortality at 0.78%, as opposed to the two outbreaks in Messina where the recorded levels stood at 34.46% and 8.11%, respectively. Outbreaks in the first epidemic Outbreaks in the second epidemic Outbreaks in the third epidemic Figure 1 Outbreaks of bluetongue recorded during the first, second and third epidemics in Sicily, November May 2003 Table I Number of outbreaks, animals present, symptoms, deaths and epidemiological indicators during the first, second and third epidemics in Sicily, November 2000-May 2003 Epidemic No. of outbreaks Animals present Animals showing symptoms Deaths Morbidity (%) Mortality (%) Fatality (%) First Second Third Veterinaria Italiana, 40 (3),

96 Global situation Percentagee 70% 60% 50% 40% 30% 20% 10% 0% First outbreak Second outbreak Third outbreak Outbreak Mortality Morbidity Fatality rate Figure 2 Comparison of epidemiological indicators for the three epidemics of bluetongue in Sicily, November 2000-May 2003 The third epidemic, from 14 May 2002 to 14 May 2003 affected 19 local councils with a total of 53 outbreaks (Fig. 1), of which 28 occurred in the Province of Messina, 21 in Catania, 2 in Palermo, 1 in Caltanissetta and 1 in Siracuse. During the third epidemic, morbidity was 17.23%, mortality 7.43% and the fatality rate 43.07% (Table I; Fig. 2). In each outbreak, the disease was caused by BTV-2 or BTV -9. It has not been possible to establish the exact role of the two serotypes in the outbreaks. Morbidity increased significantly in the second epidemic compared to that of the first (p<0.05) and again, in the third epidemic compared to the second (p<0.01). Seroconversion of sentinel animals Upon careful analysis, an increase in the overall monthly incidence (of all serotypes) can be traced from June 2002, reaching a peak in September 2002 where the incidence of seroconversion reached 5.87% ± 0.97% compared to that of all the sentinel animals (Fig. 3). After declining in the month of October 2002, the incidence increased to 3.81% ± 0.82% and 3.94% ± 0.95% in November and December of the same year. Analysing seroconversions in the bovine sentinel animals and recording the monthly incidence per serotype, it is clear that the trends and values of BTV-2, present in Sicily since the first outbreak, largely reflect the trends and values of overall seroconversions (Figs 4 and 5). Serotype 9 first appeared on 12 October 2001 during the second epidemic, reaching peak incidence in December 2002 and March 2003, with levels of 1.56% ± 0.61% and 1.43% ± 0.62%, respectively Seroconversions (%) 8% 7% 6% 5% 4% 3% 2% 1% 0% January March May July Septem Novemb January March May July Septem Novemb January March May Figure 3 Overall monthly incidence of seroconversions to bluetongue in sentinel animals in Sicily, January May 2003 (Fig. 6). Seroconversions caused by serotype 4 were recorded in November 2002 in the Province of Siracuse, whilst in December 2002, serotype 16 occurred in the same province. The highest incidence recorded for BTV-4 was between 0.09% ± 0.13% in November 2002 and 0.16% ± 0.23% in January 2003 (Fig. 7). The highest incidence for serotype 16 was recorded in December 2002 at a level of 0.25% ± 0.24% Seroconversions (%) 8% 7% 6% 5% 4% 3% 2% 1% 0% January March May July September November January March May July September November January March May Total BTV-2 BTV-4 BTV-9 BTV-16 Figure 4 Monthly incidence of seroconversions to each of four bluetongue virus serotypes in Sicily, January 2001-May Veterinaria Italiana, 40 (3), 2004

97 Global situation Seroconversions (%) 8% 7% 6% 5% 4% 3% 2% 1% Seroconversions (%) 0.5% 0.4% 0.3% 0.2% 0.1% 0% January March May July September November January March May July September November January March May % January March May July Septem Novemb January March May July Septem Novemb January March May Figure 5 Monthly incidence of seroconversions caused by bluetongue virus serotype 2 in sentinel animals in Sicily, January 2001-May 2003 (Fig. 8). In the period from September 2001 to August 2002, there was an overall cumulative incidence of 4.53% ± 0.76%; this value reached levels of 20.03% ± 1.85% between September 2002 and March Similarly, a noticeable increase in the cumulative incidence of the serotypes was observed from the first to the second epidemic. Cumulative incidence levels for serotype 2 rose from 4.02% to 18.12% while those for serotype 9 rose from 0.58% to 3.30%. The monthly incidence for each province during the months of June and July 2002, revealed that the sentinel animals experienced 5% Figure 7 Monthly incidence of seroconversions caused by bluetongue virus serotype 4 in sentinel animals in Sicily, January 2001-May 2003 seroconversions only in Messina (5.38% ± 2.96%,16.50% ± 5.07%, respectively). In August, seroconversions were observed in the provinces of Messina, Catania and Enna (13.56% ± 5.05%, 4.61% ± 1.88%, 8.96% ± 3.95% respectively) with few changes throughout September, October and November. In November, however, seroconversions were also detected in the Province of Siracuse (3.74% ± 1.86%). The seroconversions persisted throughout December in Siracuse and Messina, spreading to the Provinces of Ragusa and Agrigento (8.56% ± 3.03%, 3.85% ± 7.39%, 10.47% ± 3.61%, 11.56% ± 4.18%, respectively) (Fig. 9). 0.5% Seroconversions (%) 4% 3% 2% 1% 0% Seroconversions (%) 0.4% 0.3% 0.2% 0.1% January March May July September November January March May July September November January March May 0.0% January March May July Septem Novemb January March May July Septem Novemb January March May Figure 6 Monthly incidence of seroconversions caused by bluetongue virus serotype 9 in sentinel animals in Sicily, January 2001-May 2003 Figure 8 Monthly incidence of seroconversions caused by bluetongue virus serotype 16 in sentinel animals in Sicily, January 2001-May 2003 Veterinaria Italiana, 40 (3),

98 Global situation Seroconversions (%) 40% 35% 30% 25% 20% 15% 10% 5% 0% September November January March May July September November January March May July Agrigento Caltanissetta Catania Enna Messina Palermo Ragusa Siracuse Trapani Figure 9 Monthly incidence of bluetongue seroconversions in each affected province of Sicily, September 2001-July 2003 Incidence rates in August were significantly higher in the Provinces of Messina and Enna (p<0.01 and p<0.05, respectively), remaining highest in September in the Province of Messina (p<0.01). In October, the disease was spread equally through three provinces (Messina, Enna and Catania). In November, a significant peak was reached (p<0.01) in the Provinces of Messina and Enna. Finally, the seroconversions spread throughout the provinces of Ragusa and Agrigento during December without significant changes (p<0.05) (Fig.10). Discussion The epidemiological study conducted for each outbreak never indicated that animals had been introduced either from Sardinia or from any other region. In addition, there was no record of infected animals being moved from the western Province of Palermo (where the first outbreaks occurred) to the eastern areas around Trapani and Agrigento (where subsequent outbreaks occurred). The clinical surveillance network and epidemiological studies performed in all the regions have also shown that BT was confined to the known areas of the outbreaks, since no symptoms of BT were recorded in any other flocks visited. The epidemiological development of BT in Sicily was rather unusual, especially when compared to the much more extensive development of the disease in other regions affected. The results show that the overall increase in seropositivity of susceptible species during the three epidemics involved the entire island. Examining the three epidemics as a whole, the morbidity level was 13.25%, the mortality rate 5.36% and the fatality rate 41.49%. If the monthly incidence of disease is considered in relation to the regional sheep and goat populations, the highest peak was reached in August 2002, with 0.19%; this figure highlights the lack of association between the disease and actual loss of animals, although a significant increase in morbidity can be observed from the first to the third epidemic (Fig. 11). Incidence (%) 0.20% 0.18% 0.16% 0.14% 0.12% 0.10% 0.08% 0.06% 0.04% 0.02% 0.00% October January April July October January April July October January Figure 10 Spread of bluetongue and seroconversions in Sicily, June-December 2002 Figure 11 Regional incidence of bluetongue in Sicily, as a proportional percentage of the regional sheep and goat populations, October 2000-January Veterinaria Italiana, 40 (3), 2004

99 Global situation The different mortality levels and fatality rates observed for the two serotypes in the second epidemic may confirm the lower pathogenicity of serotype 9. During 2003, seroconversions seem to have decreased dramatically, although it is important to note that the number of tests on sentinel animals in the Province of Messina was also reduced. However, these observations continue to be of relevance to the preceding years of study, since any changes in overall climatic conditions could undermine theories on seasonal or territorial risk. References 1. Caracappa S., Di Bella C., Guercio A., Prato F. & Torina A. (2001). Emergenza bluetongue in Sicilia: controllo del territorio ed attività di sorveglianza. Atti della Giornate di Studio bluetongue stato dell arte e possibili strategie di contenimento. Perugia, 16 November, European Commission (EC) (2000). Possible use of vaccination against bluetongue in Europe. Report of the Scientific Committee on Animal Health and Animal Welfare. EC, Brussels, Fleiss J.L. (1981). Statistical methods for rates and proportions, 2nd Ed. John Wiley & Sons, Chichester, Lelli R., Calistri P., Rolesu, S., Patta C., Sulis F. & Uleri R. (2000). Bluetongue: un problema anche italiano. Praxis Vet., 21 (4), Ministero della Salute (2001). Ordinanza del 11 maggio: Misure urgenti di profilassi vaccinale obbligatoria contro la febbre catarrale degli ovini (bluetongue). Disposizioni per lo spostamento degli animali. Allegato 1, parte 4, Ministero della Salute (2002). Ordinanza del 11 maggio Rev. 3 del 18/03/2002: Misure urgenti di profilassi vaccinale obbligatoria contro la febbre catarrale degli ovini (bluetongue). Piano di sorveglianza entomologica Anno Allegato 1, parte 2, Migliazzo A. (2002). Epidemiologia della Bluetongue in Sicilia; I e II epidemia (tesi di laurea A.A. 2001/2002, Università degli Studi di Parma, Facoltà di Medicina Veterinaria, Dipartimento di Malattie Infettive degli Animali). PhD thesis, Parma University. 8. Pini A. & Prosperi S. (1999). Bluetongue. In Manuale di malattie esotiche. Vet. Ital., 31, (Monografia 20), Thrusfield M. (1997). Veterinary epidemiology, 2nd Ed. Blackwell Science, Oxford, 4, Veterinaria Italiana, 40 (3),

100 Global situation Vet. Ital., 40 (3), Results of current surveillance of likely bluetongue virus vectors of the genus Culicoides in Catalonia, Spain V. Sarto i Monteys (1), C. Aranda (2), R. Escosa (3), N. Pagès (1) & D. Ventura (1) (1) Departament d Agricultura, Ramaderia i Pesca Fundació CReSA/Entomologia Universitat Autònoma de Barcelona, Campus de Bellaterra, edifici V, Bellaterra (Barcelona), Spain (2) Servei de Control de mosquits, Consell Comarcal del Baix Llobregat Parc Torreblanca CN 340 s/n, Sant Feliu de Llobregat, Spain (3) CODE Avinguda I. Soriano Montaunt 86, Amposta, Spain Summary Following the outbreaks of bluetongue (BT) disease in sheep on the Balearic islands in 2000, a survey was conducted for Culicoides vectors along the eastern Catalonian cost of continental Spain where the presence of only C. obsoletus (Meigen) and C. pulicaris (Linnaeus) was known. Light-trap collections made at eight sites in 2002 yielded nine species of Culicoides, including C. imicola Kieffer (represented by a gravid female caught at Dosrius at a latitude of N) and C. scoticus Downs and Kettle. The following season (2003), C. imicola was captured consistently at all sites and in greater numbers (maximum catch of 46) from August to November. The findings suggest that the distribution of C. imicola is extending northwards into Europe. The presence of four bluetongue vectors (C. imicola, C. obsoletus, C. pulicaris and C. scoticus) in Catalonia is of concern. Keywords Bluetongue virus Culicoides Catalonia Surveillance Spain Vector. Midges of the genus Culicoides (Diptera: Ceratopogonidae) were trapped weekly at eight sites located along the Catalonian coast of the Mediterranean in Spain. This trapping programme was performed within the framework of an investigation, co-ordinated by the Catalonian Department of Agriculture, Livestock and Fisheries, into the occurrence/absence of possible vectors of bluetongue virus (BTV) in Catalonia. The surveillance was initiated as a consequence of the outbreak of bluetongue (BT) in 2000 and the discovery of C. imicola on the Balearic islands. On account of several difficulties encountered while selecting the farms (all stocked with either sheep or goats, or both), the date of first trapping varied considerably between farms (Table I). The locations of the traps are shown in Figure 1. To date, knowledge of the Culicoides species present in Catalonia was extremely poor. In fact, the only published references concerning species from this region are those by Strobl who almost 100 years ago reported C. pulicaris and C. obsoletus in Malgrat (a Table I Trapping dates at eight locations in Catalonia, Spain, in 2003 Trapping sites (from north to south) First trapping date Peralada 31 July 2003 Vilanova de la Muga 31 July 2003 Dosrius 1 January 2003 Bellaterra 1 January 2003 Sant Just Desvern 21 May 2003 Vallirana 14 May 2003 Vinallop 21 May 2003 La Galera 21 May 2003 coastal town in the Province of Barcelona) (6), and by Havelka who reported C. circumscriptus, C. impunctatus, C. longipennis, C. minutissimus and C. punctatus in Lloret de Mar (Girona) (1). Later, as a result of the 1990 African horse sickness (AHS) epizootic that affected southern Spain, the Spanish Veterinary Services set 70 light traps in 17 provinces, 130 Veterinaria Italiana, 40 (3), 2004

101 Global situation it was found in Catalonia, almost at the northernmost point of the Iberian Peninsula (parallel N). Previously, in continental Spain, the northernmost record for C. imicola was Talavera de la Reina (Toledo), at N (2, 3). To date (2003), the ongoing surveillance programme has shown that C. imicola seems to be spread along the entire coastal region of Catalonia, as several specimens have now been found at all trapping sites. Specimens were first collected at the end of August, after the first heavy rains in Catalonia, and were also collected throughout September, October and November. Catches were not very large, with a maximum of 46 specimens per trap; however, the continuous presence of C. imicola in 2003 appears significant, especially on account of the fact that only a single specimen was captured in Sampling sites in Peralada 5. Vallirana 2. Vilanova de la Muga 6. Sant Just Desvern 3. Dosrius 7. Vinallop 4. Bellaterra 8. La Galera Figure 1 Location of Culicoides traps in Catalonia, Spain, 2003 mostly in southern localities. However, three of these were set in coastal Catalonia as follows: Castelló d Empúries (Girona), 18 m above sea level, Viladecans (Barcelona), 17 m above sea level and Amposta (Tarragona), 8 m above sea level. A total of 699 Culicoides were collected at the three trapping sites in 1990 and in 1991, and comprised nine species. Culicoides obsoletus and C. pulicaris were the only likely BTV vectors found (the primary vector, C. imicola, was not found). Very recently, the presence of nine Culicoides species, trapped between May 2001 and December 2002, in Dosrius and Bellaterra, in the Province of Barcelona, was reported (4). These included the first records of C. scoticus and C. imicola in Catalonia. In regard to C. imicola, only a single female was collected (on 8 August 2002) at Dosrius; it was heavily gravid and might have been blown on the wind from the Balearic islands of Majorca or Menorca where large populations of C. imicola exist and where an epizootic of BT occurred in It is known that adult Culicoides can be carried on the wind for long distances, as far as 700 km (5). The west coasts of these two islands are only 209 km (Menorca) and 200 km (Majorca) from the Barcelona coast of mainland Spain. The finding of the principal vector of BT and AHS viruses in Europe is in itself significant but what makes it more important is that The current situation in Catalonia seems to correspond to that foreseen by Wittmann et al. (7), who developed a logistic regression model to identify locations where C. imicola could become established in Europe. The model indicated that under current conditions, the distribution of C. imicola in Spain, Greece and Italy could extend northwards and that other eastern Mediterranean countries could be invaded. The presence in Catalonia of C. pulicaris, C. obsoletus, C. scoticus and especially C. imicola, all likely BTV vectors, is a concern that must be addressed adequately by regional animal health authorities, since the risk of a BT outbreak in Catalonia and neighbouring regions cannot be overlooked. References 1. Havelka P. (1982). Neue Ceratopogonidenfunde von der Iberischen Halbinsel. Eos, 63, Ortega M.D., Mellor P.S., Rawlings P. & Pro M.J. (1998). The seasonal and geographical distribution of Culicoides imicola, C. pulicaris group and C. obsoletus group biting midges in central and southern Spain. Arch. Virol. [Suppl.], 14, Rawlings P., Pro M.J., Pena I., Ortega M.D. & Capela R. (1997). Spatial and seasonal distribution of Culicoides imicola in Iberia in relation to the transmission of African horse sickness virus. Med. Vet. Entomol., 11, Sarto i Monteys V. & Saiz-Ardanaz M. (2003). Culicoides midges in Catalonia (Spain), with special reference to likely bluetongue virus vectors. Med. Vet. Entomol., 17, Sellers R.F. (1992). Weather, Culicoides, and the distribution and spread of bluetongue and African horse sickness viruses. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Veterinaria Italiana, 40 (3),

102 Global situation Symposium, Paris, June CRC Press, Boca Raton, Strobl G. (1906). Spanische Dipteren. II: Beitrag. Memorias de la Real Sociedad Española de Historia natural, Tomo III, Memoria 5ª, 1905, Wittmann E.J., Mellor P.S. & Baylis M. (2001). Using climate data to map the potential distribution of Culicoides imicola (Diptera: Ceratopogonidae) in Europe. Rev. Sci. Tech. Off. Int. Épiz., 20 (3), Veterinaria Italiana, 40 (3), 2004

103 Vet. Ital., 40 (3), Global situation Bluetongue surveillance in Switzerland in 2003: a serological and entomological survey A. Cagienard (1), F. Dall Acqua (2), B. Thür (3), P.S. Mellor (4), E. Denison (4), C. Griot (3) & K.D.C. Stärk (1) (1) Swiss Veterinary Office, Schwarzenburgstrasse 161, PO Box, 3003 Berne, Switzerland (2) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy (3) Institute of Virology and Immunoprophylaxis, National Reference Laboratory for exotic diseases, 3147 Mittelhäusern, Switzerland (4) Institute for Animal Health, Pirbright Laboratory, Ash Road, Pirbright, Woking, Surrey GU24 ONF, United Kingdom Summary At present, Switzerland is considered officially free from bluetongue (BT) disease. Recently reported outbreaks have recorded BT moving north as far as latitude N in Europe and 49 N in Kazakhstan. The absence of clinical disease does not prove freedom from BT virus (BTV) infection. In addition, the occurrence and distribution of the only known biological vector, certain species of Culicoides biting midges (Diptera: Ceratopogonidae), is poorly understood for Switzerland. Consequently the Swiss Veterinary Office initiated a project on BT surveillance in April 2003 on cattle farms. The study comprised serological and entomological activities; initial results are presented. Keywords Blacklight trap Bluetongue Competitive enzyme-linked immunoassay Culicoides Serological surveys Surveillance Switzerland Vector. Introduction Bluetongue (BT) is an infectious, non-contagious vector-borne viral disease of ruminants (5) that has been designated a List A disease by the Office International des Épizooties (OIE). To date, certain species of Culicoides biting midges (Diptera: Ceratopogonidae) are the only known biological vectors of bluetongue virus (BTV). Areas in which the climatic and environmental conditions favour the survival of Culicoides spp. are considered to be at risk; these areas occur approximately between 40 N and 35 S although in parts of western North America and in China may extend to almost 50 N (3). In Europe, the major vector species is C. imicola, but the C. obsoletus and C. pulicaris species groups of midges may also be involved in some areas (4). Recently, outbreaks of BT have been reported from as far north as latitude N in Europe (4) and 49 N in Kazakhstan (2); the absence of clinical disease does not prove freedom from BTV infection (7). At present, Switzerland is considered to be officially free from BT disease. However, the occurrence and distribution of vector Culicoides in Switzerland is poorly understood. With climate as a major determinant of insect distribution and vector competence, BTV competent populations of Culicoides may already be present in Switzerland or could become established as global warming proceeds. Furthermore, livestock populations in Switzerland have not been tested for antibodies against BTV. Consequently, the Swiss Veterinary Office initiated a BT surveillance project in Switzerland in April As C. imicola, the major Old World BTV vector apparently prefers feeding on cattle rather than on sheep (6), the system was based around selected cattle farms. The study was divided into serological and entomological activities. The objectives of this project were to investigate the immune status of cattle regarding BTV and to obtain baseline data on vector populations, predominantly C. imicola, C. pulicaris and C. obsoletus, and their distribution within areas of Switzerland considered to be most at risk. Veterinaria Italiana, 40 (3),

104 Global situation Materials and methods Serological study, test validation and testing protocol Blood samples were taken from cattle on farms selected randomly. The sample size for the surveillance study was calculated using OIE recommendations, so that a minimal prevalence of 2% could be detected with a probability of 95% (8) assuming that Switzerland is free from BT. Switzerland was divided into 46 equal quadrants (40 km 40 km). Taking into account that quadrants have a different acreage because of lakes and mountains, a proportion of km 2 pasture/quadrant was calculated for each quadrant. The number of farms to be sampled within each quadrant was calculated according to pasture size. A two-stage cluster sampling scheme was used. Of cattle farms sampled for routine disease surveillance purposes in 2003, 660 farms with cattle over 24 months of age were selected randomly and five blood samples taken on each. Almost half of a total of blood samples were analysed by competitive ELISA (c-elisa); prevalence of BT antibodies was calculated using a survey toolbox (1). As the prevalence of BTV-specific antibodies in the Swiss cattle population was expected to be very low, a sensitive serological test for antibody detection was required. Two commercial ELISAs, c-elisa VMRD (Pullmann, Washington, USA) and c-elisa BDSL (Irvine, UK) were compared for sensitivity. Comparison of the sensitivity of each test was conducted using international standards (weak positive control: Institute of Animal Health [IAH] Pirbright [OIE Bluetongue Reference Laboratory], United Kingdom, and weak positive control: Onderstepoort Veterinary Institute [OVI], Onderstepoort, Republic of South Africa). Initial serological screening was performed with c-elisa BDSL as it was more sensitive than c-elisa VMRD. Positive results thus obtained were to be re-tested with the c-elisa VMRD using a lower cut-off than given in the instructions of the manufacturer so as to also detect weak positive samples or international standards, respectively. As a consequence of a c-elisa VMRD positive result, all animals of the herd in which a positive sample originated would be sampled and tested. Because prevalence of BTV-antibodies is expected to be very low, and because imperfect serological tests were to be used, it would be necessary to confirm positive herd screening results by virus neutralisation (VN) tests. Therefore, the two strongest positive samples would be sent to the IAH in Pirbright to confirm seropositivity and to determine the BTV serotype involved. Test specificity and its 95% confidence interval were determined, based on serum samples originating from the 1998 serum bank consisting of samples obtained from animals over 24 months of age during routine serological surveillance. It was assumed that the Swiss cattle population was free of BT in Entomological study Vector trapping was performed on cattle farms within areas considered to be at most risk as calculated by Rawdon (9) in a mathematical model. For the entomological study, blacklight traps manufactured by the OVI were used to capture Culicoides for one night in high-risk areas. The locations of the trapping sites are shown in Figure 1. Trapping sites : catches analysed Trapping sites : catches not analysed Figure 1 Locations of 36 Culicoides trapping sites in Switzerland, Veterinaria Italiana, 40 (3), 2004

105 Global situation High-risk areas were defined as areas with the following: annual average temperature 12.5 C the most intensive rainfall adjacent to Italy without natural barriers (e.g. high mountains) to prevent the introduction of insects. Trapping site criteria were as follows: cattle farm (>3 cattle on farm during collection) altitude lower than m above sea level. Farm managers were asked to participate by telephone interview. As July (C. obsoletus) and September (C. imicola) is the peak season for midge activity, traps were set during this period. On each selected farm, a trap was set for one night (single point/single night collection modus). At each trapping site, the same trapping protocol (modified trapping protocol of the OVI) was followed. In addition to collection site parameters, an interview based on a questionnaire was conducted with the farm manager to obtain information on livestock management. The midges collected were analysed by the IAH to identify the abundance of Culicoides species found, specifically C. imicola, C. obsoletus and C. pulicaris. Meteorological data were obtained from the national Swiss weather service; daily temperatures are recorded at 25 weather stations distributed across Switzerland. Results Preliminary results of the serological study Analysis of sera collected in 2003 Of sera analysed by BDSL ELISA, were negative, 303 gave a non-specific result, while 33 were positive. To date, 40 samples giving either questionable results or a positive response to the BDSL ELISA have been re-analysed with the VMRD ELISA, resulting in 11 positive samples. For a definite positive result, additional analyses using the VN test are required; the results of these tests are pending. Preliminary results of the entomological study A total of Culicoides were caught at 21 trapping sites on 21 trapping nights (range in number of Culicoides per trapping site: ). The total Culicoides catch consisted of C. obsoletus (range per trapping site: ), 957 C. pulicaris (range per trapping site: 0-240) and other Culicoides of unidentified species. Discussion Serological study Based on the information available at present, c-elisa VMRD positive results are likely to be false positives, because both tests are not 100% specific. Additionally, positive samples originated from different farms distributed across Switzerland. However, all samples need to be re-analysed and confirmed by the IAH in Pirbright before conclusions can be made. Entomological study A further 15 Culicoides collections remain to be analysed before final conclusions can be drawn. Acknowledgements Grateful thanks are extended to Tullio Vanzetti (veterinarian of the Canton of Ticino) and H. Russi (veterinarian in the Valley of Poschiavo) for their help in the search for appropriate farms and for contacting farmers. Thanks are extended also to Lucia Polini (Museo cantonale di storia naturale, Lugano), Rudy Meiswinkel (Istituto Zooprofilattico Sperimentale, Teramo) and Gert Venter (OVI, Republic of South Africa) for providing assistance with entomological questions. References 1. Cameron A. (1999). Survey toolbox for livestock diseases: a practical manual and software package for active surveillance of livestock diseases in developing countries. Australian Centre for International Agricultural Research (ACIAR), Canberra, Lundervold M., Milner-Gulland E.J., O Callaghan C.J. & Hamblin C. (2003). First evidence of bluetongue virus in Kazakhstan. Vet. Microbiol., 92, Mellor P.S., Boorman J. & Baylis M. (2000). Culicoides biting midges: their role as arbovirus vectors. Ann. Rev. Entomol., 45, Mellor P.S. & Wittmann E.J. (2003). Bluetongue virus in the Mediterranean Basin Vet. Rec., 164, Mertens P.P.C. (1999). Orbiviruses and coltviruses. In Encyclopaedia of virology (A. Granoff & R.G. Webster, eds). London, Nevill E.M. (1978). The use of cattle to protect sheep from bluetongue infection. J. Sth Afr. Vet. Assoc., 49, Office International des Épizooties (2003). Bluetongue in Taipei China: laboratory findings. Dis. Info., 1 August, 16 (31) (oie.int/eng/info/hebdo/ AIS_08.htm accessed on 28 July 2004). Veterinaria Italiana, 40 (3),

106 Global situation 8. Office International des Épizooties (2003). Recommendations applicable to specific diseases, Part 2. In Terrestrial animal health code (oie.int/eng/ normes/mcode/a_00038.htm accessed on 12 July 2004). 9. Rawdon T. (2002). Predictive mapping of Culicoides imicola targeting bluetongue control measures in Europe. MSc Report in Veterinary Epidemiology, University of London. 136 Veterinaria Italiana, 40 (3), 2004

107 Vet. Ital., 40 (3), Global situation The current situation of bluetongue in Turkey A. Ertürk (1), N. Tatar (1), O. Kabakli (1), S. Incoglu (2), S.G. Cizmeci (1) & F.M. Barut (1) (1) Etlik Central Veterinary Control and Research Institute, Etlik, Ankara, Turkey (2) Bornova Veterinary Control and Research Institute, Bornova, Izmir, Turkey Summary The first reported outbreak of bluetongue (BT) was in 1977 in the Aydin Province in the west of the country. Disease spread between 1977 and 1979 and became endemic in the provinces bordering the Aegean and Mediterranean Seas. The causative agent was isolated in samples from sheep and calves and was identified as bluetongue virus (BTV) serotype 4. Epidemiological investigations showed that not only sheep, but also goats and cattle, were involved in these outbreaks. The vector was Culicoides imicola. The disease was controlled successfully by vigorous control measures (quarantining, animal movement control, disinfection, insecticide treatment and vaccination campaigns) in sheep in the western provinces. Attenuated BTV-4 vaccine, produced in the Etlik Central Veterinary Control and Research Institute, was used in the vaccination campaigns. Unexpected BT outbreaks occurred in the Edirne Province, northwest of Thrace, on 20 July 1999 and spread to adjacent villages. The disease was controlled successfully by the measures described above. The last case was recorded in August Serotypes were reported as BTV-9 and BTV-16 by the Institute for Animal Health (IAH) in Pirbright. Diagnosis was based on clinical findings, serological surveillance and virus isolation. Keywords Bluetongue Cattle Culicoides Goats Serology Sheep Turkey Vaccine Virus isolation. Introduction Bluetongue (BT) is a viral disease of domestic (sheep, cattle and goats) and wild ruminants; the 24 distinct serotypes are transmitted by Culicoides biting midges. The distribution of BT virus (BTV) is virtually global between latitudes 35 S and 40 N; Turkey is located within the latitudes of 36 N and 42 N and with seven international borders is located in a unique geographical and cultural position at the crossroads between Europe and Asia. There are approximately 11 million cattle and 34 million sheep and goats in 81 provinces. Clinical signs of BT range from subclinical to an acute, febrile response characterised by inflammation and congestion, leading to facial oedema and haemorrhages and ulceration of the mucous membranes. The authors describe the current situation of BT in Turkey. History The last Near East epizootic of BT occurred between 1977 and Western Anatolia was involved in this epizootic and the causative agent was identified as BTV serotype 4, isolated from sheep on several occasions and also from a calf with congenital arthrogryposis and hydranencephaly (2, 9). The most likely vectors of BTV in western Turkey have been identified as Culicoides imicola, C. obsoletus and C. schultzei (5). Serological and vector surveillance for BT was conducted between 1986 and 1989 in southern Turkey. Surveillance confirmed the widespread presence of C. imicola in the region. Although the virus was not isolated, surveillance results showed that southern Turkey would be vulnerable to BTV incursions in the future (3). After many years of no evidence of BTV, the disease was observed unexpectedly in the Edirne Province in It spread to western Anatolia during the summer and autumn of 1999 when vector populations were active. Typical clinical signs were observed in sheep flocks (pyrexia, anorexia, serous or catarrhal nasal discharge, oral hyperaemia, excoriation of the lips, erosion of the gums, swollen tongue and oedema of the head and neck). Diagnosis was based on clinical findings, serology and virus isolation. Isolates were identified as BTV serotypes 9 and 16 by the Institute for Animal Health (IAH) in Veterinaria Italiana, 40 (3),

108 Global situation Pirbright. The BT outbreaks that occurred in 1999 and 2000 are summarised in Table I. Materials and methods Virus isolation In 1999 and 2000, blood samples in oxalate-phenolglycerin were collected from sick sheep that had a high fever (>41 C) and were submitted to the Etlik Central Veterinary Control and Research Institute for virus isolation (4). Results Blood samples were inoculated into embryonated chicken eggs (ECE) by the yolk sac route and haemorrhages were observed in the embryos that died. Suspensions of affected embryonic tissues were inoculated into BHK-21 cell cultures. A total of 16 suspect isolates were submitted to the BT reference laboratory for identification and typing. The results are summarised in Table II. Serological surveillance The sera of cattle from villages in the Edirne Province where the disease was detected, were tested for antibodies against BTV using the competitive ELISA (c-elisa) (1). Cattle had not been vaccinated against BT; the results indicated that BTV had circulated in the region. The BT outbreaks spread north-west to western Anatolia and serological results indicated that not only sheep but also cattle were involved in the outbreak in the Thrace region. The sera of cattle from the villages of the Edirne Province were tested for antibody against BTV virus using the c-elisa; 40 of the 78 sera tested positive. Likewise, 12 of 22 sera taken randomly from cattle in the villages of Kirklareli Province, where the vector insects were known to exist, tested positive for antibody against BTV. The results for cattle are summarised in Table III. The provinces of Turkey infected with BT are shown in Figure 1. Discussion The Mediterranean Basin is a potential source of BTV for Europe (6). Earlier outbreaks occurred mostly in the southern and western regions of Turkey (8). However, more recent outbreaks were seen in Thrace and the disease moved from Thrace to some parts of the Aegean region. BTV-4 was recorded in earlier outbreaks and BTV-9 and BTV-16 were identified in later outbreaks in Turkey Table I Bluetongue outbreaks recorded in Turkey, Province District Village Date Species Susceptible Cases Deaths Destroyed Slaughtered Edirne Lalapasa Vaysal Ovine-caprine Haciansment Ovine-caprine Kalkansöğüt Ovine-caprine Manisa Selendi Havaoglu Ovine-caprine Salihi Kirdamlar Ovine-caprine Kemer Damlar Ovine-caprine Aydin Karacasu Palamutcuk Ovine-caprine Denizli Akköy Haytabey Ovine-caprine Kavakbasi Ovine-caprine Izmir Aliaga Çaltidere Ovine-caprine Helvaci Ovine-caprine Dikili Demirtaş Ovine-caprine Esentepe Ovine-caprine Menemen Yeniköy Ovine-caprine S. Hisar Beyler Ovine-caprine Izmir Ödemis Inönü Ovine-caprine Ödemis Mescitli Ovine-caprine Veterinaria Italiana, 40 (3), 2004

109 Global situation Table II Identification of bluetongue isolates from different provinces of Turkey, Province Submission date to the Pirbright Laboratory Town/village No. of passages in ECE and BHK-21 Serotype isolated Edirne 1999 Kalkansöğüt ECE/1, BHK-21/4 BTV-9 Izmir 2000 Seferihisar ECE/3, BHK-21/7 BTV-16 Manisa 2000 Salihi ECE/2, BHK-21/7 BTV-9 Denizli 2000 Akköy ECE/2, BHK-21/7 BTV-9 Manisa 2000 Selendi ECE/3, BHK-21/6 BTV-9 ECE embryonated chicken eggs BHK baby hamster kidney Table III Results of bluetongue serological surveillance testing in cattle in Turkey, 1999 Province District Village Total Positive Negative Edirne Lalapasa Vaysal H. Danişment K. Sögüt Kirklareli Vize Kizilagaç Hamidiye Demirköy Yigitbaşi Hamdibey Avcilar Total and 2000 Disease not reported Figure 1 Provinces of Turkey affected by bluetongue between 1977 and 2000 and neighbouring countries (6). C. imicola continues to be the most prevalent vector in Turkey (5). Global warming and wind systems may be factors which promote the spread of the disease (7). Since the control of disease in the region requires co-operation and co-ordination with neighbouring countries, a United Nations Food and Agriculture Organization project involving Turkey, Bulgaria and Greece was developed and project activities commenced in Regional international collaboration and cooperation is vital to determine and identify vector species and to eradicate the disease from the region. Veterinaria Italiana, 40 (3),

110 Global situation References 1. Anderson J. (1984). Use of monoclonal antibody in a blocking ELISA to detect group specific antibodies to bluetongue virus. J. Immunol. Methods, 74, Anon. (1980). Epidemiology, diagnosis and control of bluetongue in Turkey. Bull. Off. Int. Épiz., 92, Burgu I., Urman H.K., Akça Y., Yonguç A., Mellor P.S. & Hamblin C. (1992). Serologic survey and vector surveillance for bluetongue in southern Turkey. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Gard G.P. & Kirkland P.D. (1993). Bluetongue virology and serology. In Australian standard diagnostic techniques for animal diseases (L.A. Corner & T.J. Bagust, eds). CSIRO Information Services, Melbourne, Jennings M., Boorman J.P.T. & Ergun H. (1983). Culicoides from Western Turkey in relation to bluetongue disease of sheep and cattle. Rev. Elev. Med. Vet. Pays Trop., 36 (1), Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean Basin, Vet. J., 164 (1), Sellers R.F. & Pedgley D.E. (1985). Possible windborne spread to western Turkey of bluetongue virus in 1997 and of Akabane virus in J. Hyg. (Lond.), 95 (1), Taylor W.P. & Mellor P.S. (1994). Distribution of bluetongue virus in Turkey Epidemiol Infec., 112 (3), Yonguç A.D., Taylor W.P., Csontos L. & Worrall E. (1982). Bluetongue in western Turkey. Vet. Rec., 111, Veterinaria Italiana, 40 (3), 2004

111 Vet. Ital., 40 (3), Global situation Incidence and isolation of bluetongue virus infection in cattle of the Santo Tomé Department, Corrientes Province, Argentina I.A. Lager (1), S. Duffy (1), J. Miquet (1), A. Vagnozzi (1), C. Gorchs (1), M. Draghi (1), B. Cetrá (1), C. Soni (1), C. Hamblin (2), S. Maan (2), A.R. Samuel (2), P.P.C. Mertens (2), M. Ronderos (3) & V. Ramirez (4) (1) INTA-Castelar, CP 1712, Hurlingham, Buenos Aires, Argentina (2) Institute for Animal Health, Pirbright Laboratory, Ash Road, Woking, Surrey, GU24 0NF, United Kingdom (3) Dpto Entomologia, Museo La Plata, Paseo del Bosque s/n, 1900, La Plata, Argentina (4) SENASA, Av. Fleming 1653, CP 1640, Buenos Aires, Argentina Summary Sentinel herds were monitored for the detection of bluetongue (BT)-specific antibodies and virus over two periods, namely: June 1999 to August 2000 and September 2000 to April Herds were located in Santo Tomé (Herds 1 and 2) where BTV activity was known to occur. From June 1999 to August 2000, the cumulative incidence (CI) of bluetongue virus (BTV) infection was 0% and 35% in Herds 1 and 2, respectively. In the second period, the CI of BTV infection was 10% and 97% in Herds 1 and 2, respectively. The virus was isolated from red blood cells of animals that seroconverted and was identified as serotype 4. Averages of the monthly maximal temperatures were always above 19 C. However, averages of the monthly median temperatures were below 19 C and averages of the monthly minimal temperatures were below 15 C from May 2000 to August There was no viral activity detected at that time. Culicoides insignis was identified as the predominant potential vector species (99%) trapped near sentinel herds. Although clinical disease has never been reported in Argentina, viral activity was detected and the virus has been isolated in sentinel herds. Keywords Argentina Bluetongue virus Culicoides Infection Disease Serotypes. Introduction and objectives Bluetongue (BT) virus is present worldwide in the tropics and subtropics and can cause disease in domestic and wild ruminants. The virus is transmitted by some species of haematophagous Culicoides midges (2). In South America almost all the countries have serological evidence of BT virus (BTV) infection (1) but only four outbreaks of the disease have been reported. The importance of BT resides in economic losses due to the restrictions to international movement of ruminant livestock and germplasm. Although 24 serotypes of BTV have been recognised in the world, BTV distribution seems to follow ecological factors (4). The objective of this study was to investigate the pattern of transmission of BTV in cattle in Argentina and to isolate the virus. Materials and methods Sentinel herds were monitored serologically from June 1999 to August 2000 and from September 2000 to April Herds were located in Santo Tomé (Herds 1 and 2) in the Corrientes Province, areas where BTV activity was known to occur. In the first period, 30 BTV competitive ELISA (c-elisa)- negative, 6-9-month-old female cattle were selected from each herd (Herds 1 and 2). The c-elisa kits were supplied by Veterinary Diagnostic Technology, Wheat Ridge in Colorado (USA). Blood samples were collected monthly except for January 1999 and July In the second period, 40 seronegative 6-9-month-old female cattle were selected from each herd (Herds 1 and 2). Samples were collected in September 2000, December 2000, March 2001 and April Light traps were located close to the sentinel animals to collect potential vectors of BTV Veterinaria Italiana, 40 (3),

112 Global situation from Herd 1. Red blood cells from those animals that seroconverted were processed for virus isolation by inoculation into embryonated chicken eggs and cell cultures (5). Results From June 1999 to August 2000, the cumulative incidences (CI) of BTV infection were 0% (0/34) and 35% (11/31) in Herds 1 and 2, respectively (Figs 1 and 2). Cattle July August September October November December February March April May June August identified between March and April 2001 in Herd 1 (Fig. 3) and between December and March in Herd 2 (Fig. 4) The CI for these herds were 10% (5/50) and 97% (39/40), respectively (Figs 3 and 4). Cattle August September October November December January March Positive Negative April Figure 3 Serological monitoring of cattle in Argentina (Herd 1), Negative Figure 1 Serological monitoring of cattle in Argentina (Herd 1), Cattle July August September October November December February March April May June Positive Negative Figure 2 Serological monitoring of cattle in Argentina (Herd 2), August Cattle were seropositive from October to May in Herd 2. The highest monthly CI were recorded from March to April in Herd 2. No seroconversion was detected later than May. Seroconversions were Cattle August September October November December January March Positive Negative April Figure 4 Serological monitoring of cattle in Argentina (Herd 2), Averages of the monthly maximal temperatures were always above 19ºC. However, averages of the monthly median temperatures were below 19ºC and averages of the monthly minimal temperatures were below 15ºC from May 2000 to August 2000 (Fig. 5). No viral activity was detected during this period. Culicoides insignis was identified as the predominant potential vector species (99% of midges trapped close to the sentinel herds) (6). 142 Veterinaria Italiana, 40 (3), 2004

113 Global situation Temperature ( C) March May July September November January March May July September November January March 3 kb 1 kb 2.2 kb expected PCR product Maximum Median Minimum Figure 5 Monthly temperatures in Ituzaingó-Santo Tomé, Argentina, March 1999-April 2001 BHK cells inoculated with four samples developed CPE and were positive by direct and indirect immunofluorescence with BTV-specific reagents. Those samples examined by electron microscopy showed virus particles with BTV morphological characteristics. Blood samples and tissue culture supernatants were positive with the RT-PCR technique with primers corresponding to segment 3 of the BTV genome. Using microseroneutralisation and RT-PCR, with primers corresponding to the segment 2 of the BTV genome, the four samples were identified as serotype 4 (Fig. 6). Three out of four Argentinian virus samples tested (20, 99 and 102) gave positive results in RT- PCR assays with BTV-4 specific primers, generating a cdna band of the expected size (2.2 kb) which was detected after agarose gel electrophoresis (Fig. 6). The cdna from virus sample 20 was subsequently used as template for sequencing reactions and analysis (accession number AJ585169). A comparison of 687 nucleotides near to the 5 end of genome segment 2 was made with sequence, that was also derived from a South African reference strain of BTV-4 (accession number AJ585125) and a 9.4% nucleotide sequence difference was detected. According to previous phylogenetic analyses of genome segment 2 from different BTV isolates (3), this level of homology indicates that the two strains are from the same serotype, although not closely related. Data concerning these virus isolates is available at: iah.bbsrc.ac.uk/dsrna_virus_proteins/ ReoID/btv-4.htm. Figure 6 Agarose gel electrophoresis of cdnas generated by RT- PCR, using BTV-4 specific primers and RNA templates extracted from infected BHK cells Lane 1 1 kb marker Lane 2 negative control Lane 3 BTV-4 South Africa Lane 4 BTV-4 Argentina No. 20 Lane 5 BTV-4 Argentina No. 82 Lane 6 BTV-4 Argentina No. 99 Lane 7 BTV-4 Argentina No. 102 Discussion Although clinical disease has never been reported in Argentina, viral activity was detected and the virus has been isolated from sentinel herds. This first isolation of BTV in Argentina was identified as serotype 4. There was a marked variability in the CI of BTV infection among herds and between years. Absence of BTV activity from May to September suggests that low temperatures resulted in low or no vector activity. Culicoides insignis seems to be the most likely vector of BTV in this area. Acknowledgement This project was supported by SECYT, PICT/ References Gibbs E.P.J. & Greiner E.C. (1994). The epidemiology of bluetongue. Comp. Immun. Microbiol. Infec. Dis., 17, Veterinaria Italiana, 40 (3),

114 Global situation 2. Gorchs C. & Lager I. (2001). Lengua azul. Actualización sobre el agente y la enfermedad. Rev. Arg. Microbiol., 33, Mertens P.P.C. (1999). Orbiviruses and coltviruses (Reoviridae): general features In Encyclopaedia of virology, Vol. 2 (A. Granoff & R.G. Webster, eds). Academic Press, London, Mo C.L., Thompson L.H., Homan E.J., Oviedo M.T., Greiner E.C., Gonzalez J. & Saenz M.R. (1994). Bluetongue virus isolations from vectors and ruminants in Central America and the Caribbean. Am. J. Vet. Res., 55, Office International des Épizooties (2000). Manual of standards for diagnostic tests and vaccines. OIE, Paris, 957 pp. 6. Wirth W.W., Dyce A.L. & Spinelli G.R. (1988). An atlas of wing photographs, with a summary of the numerical characters of the neotropical species of Culicoides (Diptera: Ceratopogonidae). Contrib. Am. Entomol. Inst., 25, Veterinaria Italiana, 40 (3), 2004

115 Vet. Ital., 40 (3), Epidemiology and vectors Culicoides and the global epidemiology of bluetongue virus infection W.J. Tabachnick Florida Medical Entomology Laboratory, Department of Entomology and Nematology, University of Florida IFAS, 200 9th St. SE, Vero Beach, Florida 32962, United States of America Summary The distribution of the bluetongue viruses (BTV) is limited to geographic areas containing competent vector species. All BTV-competent species belong to the genus Culicoides. In the New World, two different BTV epidemiological systems (episystems) occur. Culicoides sonorensis is responsible for transmitting BTV serotypes in North America that differ from South American serotypes transmitted by C. insignis. There are other episystems in the world. The role of different Culicoides vector species and the underlying mechanisms governing their vector capacity for BTV are unknown. It is likely that these vary between Culicoides species and episystems. As a result, our ability to predict and/or mitigate BTV in different episystems will remain problematic. Several complex issues need to be resolved to provide risk assessment and mitigation for BTV. This will require a substantial investment in new research paradigms that investigate details of underlying controlling mechanisms in several species of Culicoides. Keywords Bluetongue virus Culicoides Epidemiology Episystem Genetic control Vector capacity. Since the first report of bluetongue (BT) disease as malarial catarrhal fever in South African sheep in 1902 (3), the role of BT virus (BTV) and its impact on animal health and animal economies worldwide has become increasingly important. BTVs are distributed worldwide, and indeed the distribution of the viruses has been described as limited to geographic areas containing competent vector species (13). The major issues that have impeded efforts to reduce the impact of BTV on animal health and the economics of animal industries continue to be as follows: to develop the capability to predict regions of the world with high risk for BTV transmission to livestock to develop the capability to predict temporal periods in at-risk regions in advance of a BT outbreak to develop strategies to reduce the potential for a BT epidemic in at-risk regions, reduce the risk of introducing BTV or new BTV serotypes to BTVfree regions, and strategies that will interrupt and mitigate the impact of ongoing BTV transmission in a region. It is clear that the ability to predict and mitigate BT outbreaks in animals is at a very rudimentary stage. In particular, a major difficulty is that the ability to define potential regions that are at risk for BT epidemics is primarily based on the identification of geographic areas containing competent vector species. Although it is well established that the vectors of the BTVs are all biting midges in the genus Culicoides, the identification of suitable Culicoides vector species is fraught with uncertainties and unknowns. There is now a large body of research on several species of Culicoides and their role in BTV endemic regions (7, 10). This information has been obtained largely due to substantial BTV epidemiological history in a specific region, coupled with intensive field and laboratory confirmation on vector capacity. This has enabled identification of the major Culicoides vector species in some areas. Figure 1 shows major Culicoides BTV vector species in different regions of the world, based on information reviewed elsewhere (7, 10). The information depicted in Figure 1 is very superficial and not of much specific use in predicting or mitigating the risk for BTV transmission on a regional or local level. It is important to note that there are many localities where populations of many of these known vector species can be found, yet Veterinaria Italiana, 40 (3),

116 there is no evidence of BT. Alternatively, livestock producers in regions that have yet to experience BT are concerned that their local Culicoides species place them at risk for a BT epidemic. The distribution of known vector species and BTV serotypes is only a first cursory level for predicting and mitigating the risk for a BT epidemic. It is very likely that a good portion of the ability of various species of Culicoides to serve as effective epidemic BTV vectors is associated with the large population sizes that can be achieved under appropriate weather conditions. Therefore, weather and climatic conditions are very important (7, 15). Climate and weather models can provide a measure to predict vector distribution and epidemic risk that is discussed elsewhere in these Proceedings. However, these models are usually based on data from one Culicoides species, and extrapolation to other species may not be appropriate. Weather and climate conditions can provide indications of the risk for BT though there are many additional factors that need to be evaluated to improve risk assessment. It is essential to understand the details of the influence of many other factors on the vector capacity of a species in order to provide accurate prediction and mitigation of potential epidemic risk beyond the cursory prediction based on suitable weather conditions for the presence of a vector species. Unfortunately not much progress has been achieved in obtaining this type of information since recommendations for research on Culicoides vector species were made at the Second International Symposium on bluetongue (13). There is little information about the details of the mechanisms contributing to the role of any of the known Culicoides vector species, little to no information on intraspecies or population differences within any species of Culicoides for BTV transmission and the consequent effects on the epidemiology of BTV infection, and little information to assess the potential role of any species of Culicoides in BTV transmission in advance of a BT outbreak. The current status of information on the Culicoides species vectors of BTV worldwide requires study, as does the information needed to enable greater predictability and mitigation for BTV transmission and the risk for epidemics. Episystem concept The concept of an episystem is used here, consisting of the species and environmental aspects of an epidemiological system in a particular ecosystem which affects the distribution and dynamics of a pathogen and disease. This concept will be particularly valuable in considering complex systems such as the BTVs. Figure 1 shows that the major vector species for BTV transmission differ in different broad geographic regions of the world, namely: C. sonorensis (formerly C. variipennis) (2) in North America, C. insignis in Central and South America, C. imicola in Africa and C. wadai and C. brevitarsis in Australia. Certainly other species, particularly in Africa, Asia and Australia play a role in BTV transmission in specific regions and at specific times, such as C. fulvus and C. schultzei in Asia and C. bolitinos in South Figure 1 Worldwide distribution of the bluetongue viruses and the major Culicoides vectors 146 Veterinaria Italiana, 40 (3), 2004

117 Africa, among others. The role of these and other species, i.e. C. nubeculosus, C. obsoletus, C. pulicaris, C. actoni, C. fulvus needs to be defined and, specifically, it is important to understand how populations of these species contribute to BTV epidemiology in the presence or absence of populations of the major vector species. As a first step, accurate species descriptions, taxonomy and systematic relationships of the species in the genus Culicoides will be an essential step in addressing these issues. Information on the phylogenetic relationships in the genus Culicoides can provide clues to understanding the evolution of vector capacity for the BTVs. Phylogenetic relationships among Culicoides are only beginning to be explored using molecular phylogenetics (4, 5). The different broad geographic regions of the world contain different suitable Culicoides vector species and it is interesting to note that particularly in the New World, there are associated BTV serotypes that are apparently confined to specific regions. In the New World, two broad episystems can be defined for the BTVs. In the North American system C. sonorensis is the vector of BTV serotypes 2, 10, 11, 13 and 17, while in the South American ecosystem C. insignis is the vector of BTV serotypes 1, 3, 6, 8, 12 and 14. The factors that govern these ecosystems are unknown and are likely to be a combination of vector-virus and host interactions and differences that have yet to be explored. However, the New World does suggest that there are two BT ecosystems with different viruses and vector species. The epidemiologically significant factors that contribute to the maintenance of the two systems are unknown, possibly the result of a complex interaction between environment and all the biological components existing in each region. The result is different epidemiological systems or episystems. The term episystem is used here to include all those aspects of the ecosystem specifically relevant to the epidemiological distribution of the BTVs. New World bluetongue episystems The New World BTV episystems provide a useful illustration of the complexities in providing information to predict and mitigate vector borne animal disease like BT. The North American BTV episystem is perhaps the best understood of the BTV episystems. Tabachnick (10) reviewed the substantial information collected over the past 50 years on this system. North American BTV epidemiology, field and laboratory assessments of vector competence, and studies on genetic differentiation have shown that C. sonorensis is the primary North American vector. Close relatives of C. sonorensis in the North American Variipennis Complex, i.e. C. variipennis and C. occidentalis have not been implicated in BTV transmission. Indeed, the absence of C. sonorensis and the presence of only C. variipennis populations with lower vector competence in the north-east United States has been the basis for recommendations that the north-east United States be recognised as a BTVfree region (14). The question of whether the information from the New World episystems might be useful in understanding other episystems, and enable predictions and mitigation of BT elsewhere in the world should be examined. One might assume that C. insignis and C. sonorensis share certain characteristics that enable them to be efficient BTV vectors they both are indeed vectors of these viruses. However, it is unlikely that these two species share similar vector capacity control mechanisms. C. insignis in the subgenus Hoffmania and C. sonorensis in the subgenus Monoculicoides are not closely related phylogenetically. Since the close North American neighbours of C. sonorensis in Monoculicoides, i.e. C. occidentalis, C. variipennis and C. gigas, are not BTV vectors, and there are no close neighbours in New World Hoffmania that are BTV vectors, it is difficult to imagine that these two distantly related species (C. insignis and C. sonorensis) share vector characteristics due to a shared common ancestry. This would require that the ancestral characteristics have been maintained in two distantly related species while being lost in all other close neighbours. It is more likely that C. insignis is capable of fulfilling its BTV transmission role in the BTV South American episystem due to either different mechanisms and/or adaptations to this ecosystem that also have an impact on vector capacity. This may be related to different ancestral hosts or pathways evolved during the adaptation to cattle as hosts, which occurred under different conditions in North and South America. These secondary traits might be considered exaptations (1) or traits that evolved originally for a particular function, but were later recruited to fulfil new functions, i.e. vector capacity in the BTV episystem of South America. Similarly, C. sonorensis vector capacity traits possibly evolved independently in the North American ecosystem. BTV vector capacity traits may be exaptations that evolved in the North American episystem. What characteristics and mechanisms, whether homologous or completely different, allowing each species to fulfil the role of BTV vectors in their respective episystems are unknown. It is important to note that the previous discussion illustrates the significance of having an appreciation of the phylogenetic relationships in the genus Culicoides and using this information to obtain Veterinaria Italiana, 40 (3),

118 information on the evolution of vector capacity. Many Culicoides vectors are distantly related, yet have close relatives that are not vectors. The distribution of vector capacity in species complexes needs to be resolved. Only C. sonorensis in the Variipennis Complex is a BTV vector. When considering the Imicola, Shultzei and Pulicaris Complexes (16), it would appear that it is unlikely that there will be shared mechanisms between the disparate Culicoides vectors for the BTVs, and that there are likely many different mechanisms capable of producing the same phenotypes contributing to vector capacity for BTV. For example, genetic studies with C. sonorensis identified a single locus controlling susceptibility to C. sonorensis infection with BTV (8). There is no reason to suspect that this locus is the controlling locus that causes resistance in the close neighbour C. variipennis. Furthermore, there is no evidence or reason to assume that this locus is largely responsible for intraspecies variation or the differences observed in BTV susceptibility between populations of C. sonorensis. The identified locus contributed to the differences between individuals in a C. sonorensis laboratory colony. The extent of its significance in nature is unknown. There is no reason to expect that the North American and South American BTV episystems exist because there are similar mechanisms in each system. If C. sonorensis vector capacity is contingent on its presence and evolution in the North American ecosystem, it is problematic that C. sonorensis would even be a BTV vector if somehow transported to South America, or if it had evolved in the South American episystem. The same could be said of the entry of C. insignis into North America. Unfortunately, with the limited understanding of these complex episystems, our ability to predict the behaviour of a vector species in a different episystem, or the behaviour of any species of Culicoides encountering BTV for the first time, i.e. in a new potential episystem, is problematic. Unfortunately, the ability to predict the potential Culicoides BTV vectors in Italy, the United Kingdom, or anywhere else is equally rudimentary. The BTV episystems in each of these new regions simply do not exist so there is little information available to provide critical parameters of the system. Similarly, the information to predict the consequences of introducing different strains of BTV into an existing episystem is lacking. Although laboratory experiments and evaluations as currently undertaken, i.e. studies of vector competence and capacity, with colonies or even field collected populations may provide clues, these efforts may also be uninformative. In the absence of information about the episystem and the details of the mechanisms controlling vector ability for a particular species, such laboratory experiments provide little information about natural situations. The long-term history of the epidemiology of BTV infection in North America was essential to the interpretation of laboratory vector capacity information that resulted in understanding the role of C. sonorensis populations in BTV transmission. However, even in this situation, there is no understanding of the details of any controlling mechanisms of C. sonorensis capacity to transmit the BTVs. Hence, although there are some populations of C. sonorensis with low vector competence to infection for BTV there is no ability to assess whether these populations actually pose less of a risk for BTV transmission in an episystem. For example, there are populations of C. sonorensis in Weld County (Colorado) with approximately 1-2% susceptibility to infection for BTV in the laboratory. This is similar to the susceptibility of all C. variipennis populations tested to date (10) but whether there is less risk of BTV transmission in the Colorado episystem as a result cannot be predicted. There may be episystem factors that influence these populations and would support BTV transmission. Information on the intraspecies variation or population variation in vector capacity traits is virtually non-existent for other Culicoides species. Requirements for bluetongue risk prediction and risk mitigation It is essential to identify the critical factors governing a BTV episystem that have an impact on various geographic areas containing competent vector species. A competent Culicoides vector species must be capable of being infected with BTV, must be capable of transmitting BTV, must bite susceptible animal hosts, and must occur in numbers sufficient to sustain epidemic transmission. What is missing is information on the details of the specific underlying biological-genetic controlling mechanisms for vector capacity in each of the known BTV vector species. The following questions need clarification: 1) What specific genes govern BTV vector competence and capacity in each Culicoides vector species? 2) Are there BTV vector capacity genes that are homologous between distantly related Culicoides vector species? 3) What are shared vector capacity mechanisms between vector species and how have they evolved? Common ancestry? Convergent evolution? 4) What is the nature of intraspecies variation for vector capacity traits and how does this variation influence the risk for BTV transmission? 148 Veterinaria Italiana, 40 (3), 2004

119 5) What are common and different features of the major BTV episystems? 6) What are potential new BTV episystems? 7) What critical features of particular BTV episystems can be targeted to reduce the risk of the establishment of BTV in the system or the risk of a BT epidemic in the system? The ability to predict and mitigate an arthropodborne disease such as BT is fundamentally a very difficult issue. Similar difficulties have been raised elsewhere for understanding other arthropod-borne diseases, such as West Nile virus in North America, malaria in Africa, and other emerging arthropodborne diseases in general (10, 11, 12). There are genetic factors that control Culicoides vector competence and capacity and environmental influences (9, 10). The dynamics of the interactions probably vary according to the different species and ecologies. It is naive to believe that the same genetic factors, environmental factors and interactions between them are the same in different Culicoides species in different ecologies, particularly when species are phylogenetically disparate. The mechanisms allowing C. sonorensis to be effective as a vector of BTV serotype 11 may be absent and completely different from the mechanisms influencing C. insignis as a vector of BTV serotype 6. In C. sonorensis, the reaction of a colony to infection with one BTV serotype does not even predict the susceptibility of this colony to a different BTV serotype (6). There is little ability to predict interspecies abilities for the BTVs in northern Europe, and there is even less information to use to gauge intraspecies or population vector potential. Even in the best evaluated BTV episystem, the North American episystem, C. sonorensis interpopulation variation is virtually ignored in assessing the potential for sustaining BTV and epidemic transmission. The ability to predict and mitigate BT epidemics will require a substantial investment to obtain essential information on the mechanisms controlling vector capacity in different vector species. This will require a change in the research focus from evaluating vector capacity phenotypes in laboratory experiments with no understanding of underlying mechanisms, to elucidating the fundamental genetic and environmental factors governing vector capacity traits in several different species of Culicoides. A well documented phylogeny for the genus Culicoides is essential to understand the evolution of vector capacity and for use to predict potential vector species. Genetic, environmental and phylogenetic information will make it possible to identify with assurance the competent Culicoides vector species in specific geographic areas in different episystems. Such information will also provide opportunities to develop new strategies to interrupt introduction of the BTVs into new regions, and to reduce the impact of epidemic BT. Acknowledgements The helpful comments of C.C. Lord and G.F. O Meara on an earlier draft of the manuscript were much appreciated. Thanks are also extended to A. Borkent and W.C. Grogan for sharing their knowledge of Culicoides taxonomy. This is University of Florida Agricultural Experiment Station Publication N References 1. Gould S. & Vrba E. (1982). Exaptation a missing term in the science of form. Paleobiology, 8, Holbrook F.R., Tabachnick W.J., Schmidtmann E.T., McKinnon C.N., Bobian R.J. & Grogan W.C. (2000). Sympatry in the Culicoides variipennis complex (Diptera: Ceratopogonidae): a taxonomic reassessment. J.Med. Entomol., 37, Hutcheon D. (1902). Malarial catarrhal fever of sheep. Vet. Rec., 14, Li G.Q., Hu Y.L., Kanu S. & Zhu X.Q. (2003). PCR amplification and sequencing of ITS1 rdna of Culicoides arakawae. Vet. Parasitol., 112, Linton Y.-M., Mordue (Luntz) A.J., Cruickshank R.H., Meiswinkel R., Mellor P.S. & Dallas J.F. (2002). Phylogenetic analysis of the mitochondrial cytochrome oxidase subunit I gene of five species of the Culicoides imicola species complex. Med. Vet. Entomol., 16, Mecham J.O. & Nunamaker R.A. (1994). Complex interactions between vectors and pathogens: Culicoides variipennis (Diptera: Ceratopogonidae) with bluetongue virus. J. Med. Entomol., 31, Mellor P.S., Boorman J. & Baylis M. (2000). Culicoides biting midges: their role as arbovirus vectors. Ann. Rev. Entomol., 45, Tabachnick W.J. (1991). Genetic control of oral susceptibility to infection of Culicoides variipennis for bluetongue virus. Am. J. Trop. Med. Hyg., 45, Tabachnick W.J. (1992). Genetics, population genetics, and evolution of Culicoides variipennis: implications for bluetongue virus transmission in the USA and its international impact. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Tabachnick W.J. (1996). Culicoides variipennis and bluetongue virus epidemiology in the United States. Ann. Rev. Entomol., 41, Veterinaria Italiana, 40 (3),

120 11. Tabachnick W.J. (1998). Arthropods and pathogens: issues for emerging diseases. In Emerging infections (R.M. Krause, ed.). Academic Press, San Diego, Tabachnick W.J. (2003). Reflections on the Anopheles gambiae genome sequence, transgenic mosquitoes and the prospect for controlling malaria and other vector-borne diseases. J. Med. Entomol., 40, Tabachnick W.J., Mellor P.S. & Standfast H.A. (1992). Working team report on vectors: recommendations for research on Culicoides vector biology. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Walton T.E., Tabachnick W.J., Thompson L.H. & Holbrook F.R. (1992). An entomologic and epidemiologic perspective for bluetongue disease regulatory changes for livestock movement from the United States and observations on bluetongue in the Caribbean Basin. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Wittmann E.J., Mellor P.S. & Baylis M. (2001). Using climate data to map the potential distribution of Culicoides imicola (Diptera: Ceratopogonidae) in Europe. Rev. Sci. Tech. Off. Int. Épiz., 20, Wirth W.W. & Dyce A.L. (1985). The current taxonomic status of the Culicoides vectors of bluetongue viruses. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Prog. Clin. Biol. Res., 178, Veterinaria Italiana, 40 (3), 2004

121 Vet. Ital., 40 (3), Epidemiology and vectors The taxonomy of Culicoides vector complexes unfinished business R. Meiswinkel (1), L.M. Gomulski (2), J.-C. Delécolle (3), M. Goffredo (1) & G. Gasperi (2) (1) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Campo Boario, Teramo, Italy (2) Department of Animal Biology, University of Pavia, Piazza Botta 9, Pavia, Italy (3) Zoology Museum, University of Strasbourg, 29 bd de la Victoire, Strasbourg, France Summary The thirty species of Culicoides biting midges that play a greater or lesser role in the transmission of bluetongue (BT) disease in the pantropical regions of the world are listed. Where known, each species is assigned to its correct subgenus and species complex. In the Mediterranean region there are four species of Culicoides involved in the transmission of BT and belong in the subgenera Avaritia Fox, 1955 (three species) and Culicoides Latreille, 1809 (one species). Using both morphological and molecular second internal transcribed spacer (ITS2) sequence data, the authors reappraise the taxonomy of these four species and their congeners. A total of 56 populations of Culicoides collected from across Italy and representing 17 species (18 including the outgroup taxon C. imicola Kieffer, 1913) were analysed. The findings revealed the following: C. imicola is the only species of the Imicola Complex (subgenus Avaritia) to occur in the Mediterranean region in Europe the subgenera Avaritia and Culicoides (usually, but not quite correctly, equated with the C. obsoletus and C. pulicaris groups, respectively) are both polyphyletic, each comprising three or more species complexes (including a hitherto unknown complex) about half the species studied could not be identified with certainty; furthermore, the results indicate that at least three previously described species of Palaearctic Culicoides should be resurrected from synonymy finally, a high level of taxonomic congruence occurred between the morphological and the molecular data. One of the new vector species, C. pulicaris, was described by the father of taxonomy, Carl Linnaeus, in 1758, but today, almost 250 years later, no monograph has appeared that treats the Culicoides fauna of the northern hemisphere as a whole. At a time when such economically important livestock diseases as BT are affecting ever larger areas of Europe, it would seem appropriate to commence the production of such a monograph to aid in the field identification of vector Culicoides. This unfinished business might best be achieved through a collaborative network embracing all ceratopogonid specialists currently active in both the Palaearctic and Nearctic faunal realms. Keywords Bluetongue Culicoides biting midges Internal transcribed spacer 2 sequencing Mediterranean Taxonomy Vectors. Introduction The explosive outbreaks of bluetongue (BT) that cyclically decimate livestock in the Mediterranean Basin were understood, until recently, to be fuelled by the widely distributed Afro-Asian biting midge vector C. imicola only. In the past five years ( ) the involvement of C. imicola has been reaffirmed following large outbreaks of this disease affecting sheep in the central Mediterranean, especially Italy. However, BT also penetrated areas devoid of C. imicola and, subsequently, multiple isolations of the causative virus were made from one or more Culicoides of the Obsoletus and Pulicaris Veterinaria Italiana, 40 (3),

122 species complexes (9, 41). This would seem to confirm (finally and conclusively) their longsuspected (34, 36) involvement in the transmission of BT in southern Europe. Of significance is the fact that these two complexes contain species of Culicoides that are adapted to temperate climes with some extending as far north as the 70th parallel, and throughout this range occur in abundance, attacking both man and livestock. Whilst these new vectors are widely referred to as C. pulicaris, C. obsoletus (Meigen), 1818, and C. scoticus Downes and Kettle, 1952, some doubt is still attached to these identifications as they fall within species complexes, the member taxa of which are notoriously difficult to identify. It is essential that their identity is clarified, given the spread of BT in Europe. The authors re-evaluate the taxonomy of the C. obsoletus and C. pulicaris groups of earlier research workers and which equate, broadly, to the subgenera Avaritia and Culicoides, respectively. The principal aim is to clarify further the phylogenetic relationships that exist between the various taxa, and to discover new morphological characteristics facilitating their more rapid and reliable identification in the field. The strategy used here was to appraise each taxon both morphologically and molecularly, and, at the populational level, to obtain data from multiple series of specimens and from more than one geographical location. Initial results are briefly presented here. Materials and methods Light-trap collections made throughout Italy and its islands were screened randomly (across an altitudinal and a latitudinal transect) for specimens of the member species of the two subgenera Avaritia and Culicoides. Selected species pools were split in two: one half was reserved for slide-mounting in Canada Balsam for detailed morphological studies, whilst the other half was used for the extraction of DNA for the sequencing of the ribosomal DNA second internal transcribed spacer (ITS2). A total of 56 populations were studied and included 17 species (18 with the outgroup species C. imicola). Phylogenetic analyses were performed using Kimura two-parameter genetic distances and the neighbourjoining (N-J) algorithm (40) in PAUP*4b10 (45). The reliability of the resultant tree topology was determined by bootstrap replications (15). The dendrograms presented here are pared down versions of the larger trees originally generated (i.e. only one representative of each species is shown). The complete N-J trees will be published elsewhere. These results are compared with the morphological results obtained from the slidemounted adults. These adults, after preparation, were examined before a sometimes tentative identification was agreed upon. The formerly used species group category was abandoned as it has been used mostly (and duplicatively) in lieu of the subgenus category. The neutral (and more modern) term species complex is preferred and is employed in the strictest cladistic sense (i.e. to group closely related terminal taxa, presumably recently evolved, and united phylogenetically in that they share one or more synapomorphic features). Results Approximately 30 of the species of Culicoides across the world have been incriminated to varying degrees in the transmission of BT disease. These are listed in Table I. The more clearly proven vectors (eight) are shown in bold. (Authors are cited only where isolations of BTV have been made from unidentified species.) These 30 species can be assigned to 8 of the 36 subgenera currently deemed to comprise the genus Culicoides; 14 of the vector species belong in the subgenus Avaritia alone, and can be subdivided further amongst seven species complexes (Table I). However, nearly all these species complexes are poorly defined (47); the situation is reviewed elsewhere in this volume (31). In regard to the four species linked to the transmission of BT disease in the Mediterranean region the most important is C. imicola and probably accounts for 90% of disease transmission. The three remaining vectors are C. obsoletus and C. scoticus, also of the subgenus Avaritia, but placed within the Obsoletus Complex, and C. pulicaris (subgenus Culicoides). The taxonomy of these four vectors, and of those species deemed most closely related to them, is appraised following the table of world vectors presented below. Subgenus Avaritia Fox, 1955 Type-species: Culicoides obsoletus (Meigen), 1818: 76. Europe The Imicola Complex This complex is restricted to the Old World and comprises at least twelve species; the nine species that have been formally described (26, 27, 28, 29, 32, 33) are listed elsewhere (31), and include three important vectors of BT, namely: C. imicola, C. brevitarsis Kieffer, 1917 and C. bolitinos Meiswinkel, Three morphological studies (3, 12, 32) indicate that C. imicola is still the only species of this complex to be found in the Mediterranean, a conclusion that has been recently confirmed also molecularly (10). However, constant vigilance must be maintained in 152 Veterinaria Italiana, 40 (3), 2004

123 regard to the possible introduction of additional members of this species complex from either the east (Oriental region: C. brevitarsis) or from the south (Afrotropical region: C. bolitinos). Table I The 30 species of the genus Culicoides Latreille, 1809, that play a greater or lesser role in the transmission of bluetongue disease across the world These are assigned to their correct subgenus and species complex (where known); the species given in bold are those more clearly implicated in the field transmission of BT virus Subgenus Species complex Species Avaritia Fox, 1955 Imicola C. imicola C. brevitarsis C. bolitinos Obsoletus Orientalis Grahamii Pusillus Suzukii Gulbenkiani C. obsoletus C. scoticus C. fulvus C. dumdumi C. orientalis C. actoni C. pusillus C. wadai C. brevipalpis C. gulbenkiani C. tororoensis Culicoides Latreille, 1809 Pulicaris C. pulicaris C. magnus Silvicola Mirzaeva and Isaev, 1990 Cockerellii Species unknown (21) Monoculicoides Khalaf, 1954 Variipennis C. sonorensis Nubeculosus C. nubeculosus C. puncticollis Remmia Glukhova, 1977 Schultzei C. oxystoma C. nevilli Species unknown (6) Hoffmania Fox, 1948 Guttatus C. insignis C. filarifer Haematomyidium Goeldi, 1905 Peregrinus Milnei Complex unknown C. peregrinus C. milnei C. stellifer Oecacta Poey, 1853 Furens C. furens Subgenus unknown Complex C. trilineatus unknown Culicoides imicola was first recorded from the Mediterranean region (Egypt) in 1943 (25) but only in the 1980s was it found to occur also in the north of the Mediterranean (35). Remarkably, it was only discovered in Italy and on the islands of Sardinia, Sicily, Corsica and the Balearics in 2000 (11, 16, 37) creating the impression that C. imicola is a recent invader, spreading rapidly northwards into Europe. This seems supported by a molecular study (10) showing C. imicola to be segregated into western, central and eastern Mediterranean populations with little evidence for gene flow between them. However, further studies are required to more fully expose the haplotype structure of C. imicola, both in Africa and in south-west Asia, before it can be stated with greater certainty that this vector has only recently invaded Europe. Despite the undoubted importance of C. imicola as a vector of livestock diseases, a comprehensive inventory of its distribution in southern Europe is far from complete. Since 2000, over light-trap collections have been made across Italy and its distribution mapped in considerable detail (17). This work has shown the occurrence of C. imicola to be unexpectedly variable (and often across short distances) and, in many instances, has contradicted (7) the predictions made in risk maps utilising satellite imagery and climatic variables (1, 48). This implies that for the purposes of modelling, all the variables that determine the local prevalence of C. imicola have still not been identified. For example, differences in soil type might have a significant impact on its ability to breed successfully (30), and so deserve consideration in future modelling efforts. These efforts would be served well if more countries around the Mediterranean Basin were to initiate weekly or bi-weekly surveys across four seasons. The protocols for such surveys have been developed (18). The accurate identification of Culicoides to the level of species is what plagues all epidemiological investigations, principally because intra- and interspecific variation is always being confused. For seven species of the Imicola Complex, their status as discrete genetic entities (good species) was supported in two molecular studies (23, 42) that confirmed each species to possess stable morphological markers that, once quantified, allowed for their reliable identification in the field. The capacity to identify species correctly is essential when species-linked virus isolations are being sought during outbreaks of disease. (Unfortunately species of the Obsoletus and Pulicaris Complexes cannot be identified with similar confidence, as discussed further below.) Having stated that C. imicola is the only member of the Imicola Complex to occur in the Mediterranean region, it must be noted that another species, C. pseudopallidipennis Clastrier, 1958, was recently recorded in the Yemen (4). Although it would appear to have been misidentified (33), it nevertheless represents a second record of the Imicola Complex on the doorstep of Europe. However, until more is known about its ecology, it remains futile to speculate further on its possible northward penetration. Veterinaria Italiana, 40 (3),

124 The Obsoletus Complex In the Palaearctic region, the 30 or more described species of the subgenus Avaritia are usually referred to collectively as the C. obsoletus group (2, 8, 22, 24), or as the C. chiopterus group (20), i.e. these groups are used in the broad sense (and thus incorrectly) in lieu of Avaritia. An analysis of the ITS2 sequence (and of the morphological data) revealed that the subgenus Avaritia (at least as it occurs in Italy) is polyphyletic, and that the six species (excluding C. imicola) form three well-supported clades as follows (Fig. 1). C. montanus C. obsoletus Unidentified species C. dewulfi C. imicola s.s. C. scoticus C. chiopterus Figure 1 Phylogenetic tree of the subgenus Avaritia and related species in Italy, based on ITS2 sequences, using Culicoides imicola as an outgroup The uppermost clade (Fig. 1) includes three tightly grouped species namely C. montanus Schakirzjanova, 1962, C. obsoletus and C. species unidentified. Together, these appear to represent the Obsoletus Complex sensu stricto; its monophyly is also strongly intimated morphologically. Figure 1 illustrates that C. scoticus falls a little outside the Obsoletus Complex clade; from a morphological point of view this is not entirely unexpected as its male genitalia (a reliable indicator of phylogenetic relationships) differ somewhat in both form and size from those of C. obsoletus (and its close congeners). Unfortunately, these differences are not reflected in the female sex and so it is close to impossible to separate these two vector taxa (based upon morphology) in the field. It would seem that their identification in the future may depend upon the development of speciesspecific molecular probes. Thus, of the 20 species of Avaritia known to occur throughout the Holarctic, it is considered that only seven fall within the Obsoletus species complex sensu stricto. These are the four species listed above, plus C. sinanoensis Tokunaga, 1937, C. gornostaevae Mirzaeva, 1984 and C. sanguisuga Coquillett, Thirteen synonyms exist for C. obsoletus alone, illustrating how difficult it is to identify these species morphologically. Until the Obsoletus Complex is appraised over a wider geographic area, it will be impossible to judge whether any of these 13 names should be raised from synonymy. Lower down the dendrogram (Fig. 1) appear two species: C. chiopterus (Meigen), 1830, and C. dewulfi Goetghebuer, 1936, which the authors prefer to keep separate from the above-mentioned Obsoletus Complex sensu stricto, and which, provisionally, are referred to as the Chiopterus Complex and the Dewulfi Complex. The species C. chiopterus is collected very rarely in Italy. Indeed, doubt exists in regard to its identity because specimens were too large to be C. chiopterus (considered to be one of the smallest Culicoides of Europe). This implies that the taxonomic status of the larger C. dobyi Callot and Kremer, 1969, currently considered a synonym of C. chiopterus by some authors (5) but not by others (38), must be re-evaluated. The second taxon, C. dewulfi, was more commonly encountered in the present study but was found to occur far less abundantly than either C. obsoletus or C. scoticus. The male genitalia of C. dewulfi somewhat resemble those of C. imicola, and consequently the two species may have been confused by some authors in the past. Whether this explains why two widely different forms of C. dewulfi appear in the literature is debatable. One form has a strongly patterned wing (39), the other (as seen in material collected in this study) has a weakly patterned one (11). Finally, C. imicola is a clearly separate outlier at the base of the dendrogram (Fig. 1), and demonstrates that the Palaearctic Avaritia fauna has an evolutionary history largely distinct and separate from that which has given rise to the Avaritia faunas of tropical Africa and Asia. Subgenus Culicoides Latreille, 1809 Type species: Culicoides punctatus (Meigen), 1804: 29. Europe The precise number of species that comprise the subgenus Culicoides in the Palaearctic region is unknown, as various authors lump an agglomeration of some 50 disparately related taxa into it (8, 11, 22). This includes representatives from at least four subgenera and/or species complexes, most of which are represented in Italy; the validity of these complexes (with minor adjustments) was borne out by the sequencing of the ITS2 region. Also included by most authors is the notorious pest species C. impunctatus Goetghebuer, 1920, a species whose true affinity has long confused taxonomists. 154 Veterinaria Italiana, 40 (3), 2004

125 Unfortunately no material of this species became available to us for study. Grouped at the top of the dendrogram are the six species C. lupicaris Downes and Kettle, 1952, C. pulicaris, C. punctatus, C. newsteadi Austen, 1921, C. deltus Edwards, 1939, and dark pulicaris (Fig. 2). These species comprise the subgenus Culicoides sensu stricto; all (except C. deltus) possess the distinctive dark spot in the cubital wing cell. Unfortunately this highly diagnostic character is homoplastic and is thus unreliable for the determination of monophyly. However, as shown by the distinct clade formed by C. lupicaris and C. pulicaris (Fig. 2), it is obvious that species complexes also exist within the subgenus Culicoides. However, their reality can only be assessed once more populations are studied from across a wider geographic range. The tantalising hint that C. pulicaris may comprise two species, a lowland (Grosseto) and a highland (Campotosto) form (Fig. 2) is also worthy of note. C. lupicaris (Val Badia) C. lupicaris (Campotosto) C. pulicaris (Campotosto) C. pulicaris (Grosseto) C. punctatus (Val Badia) C. deltus C. newsteadi dark pulicaris C. fagineus C C. remmi C. fagineus A C. fagineus B C. flavipulicaris C. imicola s.s. Figure 2 Phylogenetic tree of the subgenus Culicoides and related species in Italy, based on ITS2 sequences, using Culicoides imicola as an outgroup Bootstrap values above 50% are indicated The species identified to be C. lupicaris is currently considered a synonym of C. deltus (5). The dendrogram shows clearly, however, that they are separate taxa; this was confirmed following detailed morphological study of the female sex. This indicates that C. lupicaris should be raised from synonymy but can be done only once material of both species, collected in their respective type localities (in the British Isles), has also been molecularly analysed and the results compared. This is because the series used for the current molecular study could not be identified as C. lupicaris when published keys (8) were employed. In addition, morphological similarity is insufficient cause for establishing synonymy; it could equally well transpire that the Italian material of supposed C. lupicaris represents a second, closely related, taxon. This inability to identify species using keys developed in earlier studies (8), and in those of high quality, is cause for concern. The authors have similar reservations about the identity of C. newsteadi; material from this study represents a species that is both smaller and darker than that deemed to be C. newsteadi sensu Delécolle (11). Furthermore, C. newsteadi material was captured at lower altitudes in warmer, frost-free areas only. This places into question the true status of the very similar C. halophilus Kieffer, 1924; it is currently deemed a synonym of C. newsteadi but appears to occur allopatrically being consistently reported only from the cooler climes of northern continental Europe. The second clade in Fig. 2 is represented by a species identified to be C. remmi Damien-Georgescu, 1972, which is closely related to C. grisescens Edwards, The latter (not included in this study but which is known to occur in Italy) is the type species of the subgenus Silvicola Mirzaeva and Isaev, The ITS2 data from this study, together with the morphological data, strongly support the validity of Silvicola. Five species of this subgenus have been recorded from the Palaearctic and 12 from the Nearctic, where they are assembled in the Cockerellii group; this group is clearly synonymous with the Grisescens subgroup (22). Of potential relevance is that in North America, BTV has been isolated from a species of the Cockerellii group (21), and so introduces into the vectorial arena a subgenus of Culicoides that is adapted to cooler northerly climes. At least one species of Silvicola has penetrated into more southerly latitudes; this includes the central mountainous backbone of Italy that is favoured by local shepherds for their annual transhumance treks. Near the base of the dendrogram occurs a trio of species of distinctive morphology for which the hitherto unrecognised Fagineus Complex is created. The adults can be distinguished from those of the subgenus Culicoides sensu stricto in that they lack a dark central spot in the cubital wing cell, and in that they usually have the cibarium armed with robust teeth (unfortunately both these characters are homoplastic). Species of the Fagineus Complex are niche specialists breeding in container habitats (phytotelmata) and consequently are captured more rarely than species of the subgenera Silvicola and Culicoides. Finally, and unexpectedly, the molecular data indicate that a new record for Italy, the enigmatic C. flavipulicaris Dzhafarov, 1964, should Veterinaria Italiana, 40 (3),

126 perhaps be included also in the Fagineus Complex. Until now, this taxon, because it possesses the highly diagnostic (but homoplastic) dark spot in the centre of the cubital wing cell, has always been assigned to the subgenus Culicoides. However, various other morphological data would support its placement nearer to, or in, the Fagineus Complex. Conclusions Eight points emerged from this study, as follows: 1) Culicoides imicola remains the only member of the Imicola Complex to have penetrated into the Mediterranean region where a study on its matrilineal structure indicates it to be segregated into western, central and eastern populations, with little evidence of gene flow between them. Such findings suggest C. imicola to have recently invaded southern Europe, and that outbreaks of BT are mediated prinicipally by the dispersal of infected midges on the wind (43, 44). In this context, it is relevant to note that only two of 157 known species of Afrotropical Culicoides are found also in the Mediterranean region, suggesting that the migration (and subsequent establishment) of species outside their faunal homes is the exception rather than the rule. (Whether this rule will hold under the influence of global warming is but one of many questions still to be answered.) Nevertheless, it is now becoming clearer that further molecular studies are required on the haplotype structure of C. imicola across its incredibly vast Old World range to better understand the dispersal patterns of this important vector; this knowledge should, in turn, throw further light on the role played by livestock movement in the dissemination of BT into the Mediterranean Basin. 2) The subgenus Avaritia in the Palaearctic region is divisible into four species complexes, namely: the Obsoletus, Chiopterus, Dewulfi and Imicola Complexes. However, they remain to be resolved more fully through the study of additional species and populations from across a wider geographical area, including the Nearctic region. 3) In Italy, multiple isolations of BT virus (BTV) have been made from field populations of Culicoides that contained a mixture of the two species C. obsoletus and C. scoticus (41), and in areas devoid of C. imicola. Of particular concern is that field specimens of the former two could not be reliably separated into distinct species pools. It was only the presence of the vastly dissimilar males in the same light-trap collections that revealed the sympatric occurrence of these two species; this was confirmed subsequently by the molecular analysis of a subsample of females (data not shown). Thus, for the present, both species have to be considered to be vectors of BT in Europe. Of added concern is that the molecular data revealed both taxa to occur also widely in Italy, up to altitudes approaching m. This places almost the entire country at risk to the transmission of BT disease; another implication is that if C. obsoletus is represented in the Palaearctic region by one and the same species across more than 50 of latitude, this would place a significant part of the region at potential risk to incursions by BT. 4) As noted above, the accurate morphological identification of species within the Obsoletus Complex sensu stricto has eluded taxonomists over a number of decades. Until a system for easy and rapid identification is developed, it will remain difficult to refine present knowledge at all investigative levels. It would seem inevitable that the future identification of C. obsoletus and C. scoticus (and their reliable separation from sister taxa) will depend upon the development and use of species-specific PCR assays. 5) In Italy, the mapping of the distribution and relative abundance of the Obsoletus Complex (excluding C. dewulfi and C. chiopterus), has commenced (17). Although the maps use conflated data, it is almost certain that 95% of the records (gleaned from light-trap collections) are represented by C. obsoletus and C. scoticus alone. Howewer, the relative proportions in which these two taxa occur is the key question still to be answered. It is also imperative to establish whether only one, or both, of them are involved in the transmission of BT in southern Europe. The distribution of C. imicola, the principal vector of BT in Italy, has been mapped separately and in considerable detail (16, 17). Still requiring explanation is its highly anomalous distribution across Italy. Whether this is the result of recent invasion or is due to specific larval habitat preferences remains to be investigated. 6) The C. pulicaris group or the Pulicaris group sensu lato of earlier authors (8) is clearly a moderately disparate agglomeration of 11 species or more which, in the past, were subdivided into the pulicaris, impunctatus and grisescens subgroups (22). Although these subdivisions were neither adopted nor refined further, the 11 species in Italy were found to be similarly subdivisible; these subdivisions are referred to here as the subgenera Culicoides sensu stricto, the subgenus Silvicola and the Fagineus Complex (a hitherto unrecognised species assemblage). Whether or not 156 Veterinaria Italiana, 40 (3), 2004

127 C. flavipulicaris should be included in the Fagineus Complex requires further study. 7) At the species level, the data suggest that C. lupicaris and C. remmi (and possibly also C. halophilus) should be raised from synonymy. The authors also have reservations regarding the identity of C. newsteadi in that it is suspected that a number of taxa are hidden under its wing. If the lowland and the highland forms of C. pulicaris are eventually shown to be two species, which then should be taken to be C. pulicaris sensu stricto? This raises a further question: which of the two was involved in the transmission of BT in Sicily (9)? These are just some of the taxonomic issues still to be resolved but require our urgent attention as misidentifications in biology are worse than useless. All corrective nomenclatural decisions to be taken in future should, however, be based upon the study of fresh material collected from the respective type localities. A joint strategy in which each taxon is redescribed both morphologically and molecularly will strengthen greatly the taxonomic foundation of the genus Culicoides as it will provide the data needed to distinguish between intra- and interspecific variability. 8) The mapping of the geographic distribution and relative abundance of the subgenus Culicoides sensu stricto (i.e. excluding the subgenus Silvicola and the Fagineus Complex) in Italy has also been initiated (17). However, these maps (gleaned from light-trap collections) conflate data representative of at least six species; it will require considerable effort to refine these maps to the single species level. The very recent discovery of large and extensive populations of C. imicola in Italy, and the incrimination of two (perhaps three) additional BT vectors, illustrates the paucity of our knowledge on the biting midges of the Mediterranean region. Furthermore, nearly 250 years have passed since the father of taxonomy, Carl Linnaeus, described C. pulicaris, but yet no monograph has emerged that treats the Palaearctic Culicoides fauna as a whole. Although a number of very good studies exist (e.g. 3, 4, 8, 11, 12, 13, 14, 19, 24, 38, 39, 46, 47), they remain fragmentary, have been published in a number of languages, and vary greatly in style. This regional individuality has introduced a subjective element into species identifications, and if taxonomic studies on the genus are not refined further, errors in identification are likely to persist. This small study illustrates this point as evidence has been found for undescribed species and an unrecognised species complex, and shows also that some previously described species should be resurrected from synonymy. A refined taxonomy is essential to modelling the distributions of known and potential vectors across the Palaearctic region, and more especially so for the Mediterranean Basin, the crucible in which three species of Culicoides are now known to circulate BT disease. Revisionary studies should embrace the entire Holarctic as some of the 40 Nearctic species of the subgenera Avaritia, Culicoides and Silvicola, could also display increased vectorial competence under the influence of global warming. This inclusion of the Nearctic would bring with it a host of competent Culicoidologists, and in so doing would harmonise further the large collaborative venture Unfinished business. References 1. Baylis M., Mellor P.S., Wittmann E.J. & Rogers D.J. (2001). Prediction of areas around the Mediterranean at risk of bluetongue by modelling the distribution of its vector using satellite imaging. Vet. Rec., 149, Boorman J. (1988). Taxonomic problems in Culicoides of southwest Asia, in particular of the Arabian peninsula. In Biosystematics of haematophagous insects (M.W. Service, ed.). Systematics Association, Special Volume No. 37, Clarendon Press, Oxford, Boorman J. (1991). A review of Culicoides subgenus Avaritia species (Insecta, Diptera, Ceratopogonidae) vectors of viruses of sheep, cattle and horses, with particular reference to Culicoides imicola in Europe and the Mediterranean region. A report prepared for the Overseas Development Administration. The Natural History Museum, London, 54 pp. 4. Boorman J. & Van Harten A. (2002). Some Ceratopogonidae (Insecta: Diptera) from the Arabian Peninsula, with particular reference to the Republic of Yemen. Fauna Arabia, 19, Borkent A. & Wirth W.W. (1997). World species of biting midges (Diptera: Ceratopogonidae). Bull. Am. Mus. Nat. Hist., 233, Braverman Y., Rubina M. & Frish K. (1981). Pathogens of veterinary importance isolated from mosquitoes and biting midges in Israel. Ins. Sci. Appl., 2, Calistri P., Goffredo M., Caporale V. & Meiswinkel R. (2003). The distribution of Culicoides imicola in Italy: application and evaluation of current Mediterranean models based on climate. J. Vet. Med. B, 50, Campbell J.A. & Pelham-Clinton E.C. (1960). A taxonomic review of the British species of Culicoides Latreille (Diptera: Ceratopogonidae). Proc. R. Soc. Edin. Sec. B (Biol.), 67, Caracappa S., Torina A., Guercio A., Vitale F., Calabro A., Purpari G., Ferrantelli V., Vitale M. & Mellor P.S. (2003). Identification of a novel Veterinaria Italiana, 40 (3),

128 bluetongue virus vector species of Culicoides in Sicily. Vet. Rec., 153, Dallas J.F., Cruickshank R.H., Linton Y.-M., Nolan D.V., Patakakis M., Braverman Y., Capela R., Capela M., Peña I, Meiswinkel R., Ortega M.D., Baylis M., Mellor P.S. & Mordue (Luntz) A.J. (2003). Phylogenetic status and matrilineal structure of the biting midge Culicoides imicola, in Portugal, Rhodes and Israel. Med. Vet. Entomol., 17, Delécolle J.-C. (1985). Nouvelle contribution à l étude systematique et iconographique des espèces du genre Culicoides (Diptera: Ceratopogonidae) du nord-est de la France. Thesis, Université Louis Pasteur de Strasbourg, UER Sciences, Vie et Terre, 238 pp. 12. Delécolle J.-C. & de La Rocque S. (2002). Contribution à l étude des Culicoides de Corse. Liste des espèces recensées en 2000/2001 et redescription du principal vecteur de la fièvre catarrhale ovine C. imicola Kieffer, 1913 (Diptera: Ceratopogonidae). Bull. Soc. Entomol. Fr., 107, Dyce A.L. (1983). Review of published records of Culicoides species in subgenus Avaritia (Diptera: Ceratopogonidae) from the New Guinea subregion. Int. J. Entomol., 25, Dyce A.L. & Wirth W.W. (1983). Reappraisal of some Indian Culicoides species in the subgenus Avaritia (Diptera: Ceratopogonidae). Int. J. Entomol., 25, Felsenstein J. (1985). Confidence limits on phylogenies: an approach using the bootstrap. Evolution, 39, Goffredo M., Conte A., Cocciolito R. & Meiswinkel R. (2003). Distribuzione e abbondanza di Culicoides imicola in Italia. Vet. Ital., 39 (47), Goffredo M., Conte A. & Meiswinkel R. (2004). Distribution and abundance of Culicoides imicola, Obsoletus complex and Pulicaris complex (Diptera: Ceratopogonidae) in Italy. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Goffredo M. & Meiswinkel R. (2004). Entomological surveillance of bluetongue in Italy: methods of capture, catch analysis and identification of Culicoides biting midges. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Holbrook F.R., Tabachnick W.J., Schmidtmann E.T., McKinnon C.N., Bobian R.J. & Grogan W.L. (2000). Sympatry in the Culicoides variipennis complex (Diptera: Ceratopogonidae): a taxonomic reassessment. J.Med. Entomol., 37, Khalaf K The speciation of the genus Culicoides. Ann. Entomol. Soc. Am., 47, Kramer W.L., Jones R.H., Holbrook F.R., Walton T.E. & Calisher C.H. (1990). Isolation of arboviruses from Culicoides midges (Diptera: Ceratopogonidae) in Colorado during an epizootic of vesicular stomatitis New Jersey. J. Med. Entomol., 27, Kremer M. (1965). Contribution a l étude du genre Culicoides Latreille particulièrement en France. Paul Lechevalier, Paris, 299 pp. 23. Linton Y.-M., Mordue A.J., Cruickshank R.H., Meiswinkel R., Mellor P.S. & Dallas J.F. (2002). Phylogenetic analysis of the mitochondrial cytochrome oxidase subunit I gene of five species of the Culicoides imicola species complex. Med. Vet. Entomol., 16, Liu J., Li Y. & Yu Y. (1993). Quantitative taxonomic analysis of female (sic) of Culicoides obsoletus group in China (Diptera: Ceratopogonidae). Contributions to blood-sucking Dipterous insects, Vol. 3. Shanghai Science and Technology Press, Shanghai, Macfie J.W.S. (1943). Ceratopogonidae (Diptera) from Egypt. Proc. R. Ent. Soc. London, Ser. B., 12, Meiswinkel R. (1989). Afrotropical Culicoides: a redescription of C. (Avaritia) imicola Kieffer, 1913 (Diptera: Ceratopogonidae) with description of the closely allied C. (A.) bolitinos sp. nov. reared from the dung of the African buffalo, blue wildebeest and cattle in South Africa. Onderstepoort J. Vet. Res., 56, Meiswinkel R. (1991). Afrotropical Culicoides: C. (Avaritia) miombo sp. nov. a widespread species closely allied to C. (A.) imicola Kieffer, 1913 (Diptera: Ceratopogonidae). Onderstepoort J. Vet. Res., 58, Meiswinkel R. (1992) Afrotropical Culicoides: C. (Avaritia) loxodontis sp. nov. a member of the Imicola group (Diptera: Ceratopogonidae) associated with the African elephant in the Kruger National Park, South Africa. Onderstepoort J. Vet. Res., 59, Meiswinkel R. (1995). Afrotropical Culicoides: biosystematics of the Imicola group, subgenus Avaritia (Diptera: Ceratopogonidae), with special reference to the epidemiology of African horse sickness. MSc (Agric.) thesis, University of Pretoria, South Africa. 30. Meiswinkel R. (1997). Discovery of a Culicoides imicola-free zone in South Africa: preliminary notes and potential significance. Onderstepoort J. Vet. Res., 64, Meiswinkel R. (2004). Adult characters defining and separating the Imicola and Orientalis species complexes of the subgenus Avaritia Fox, 1955 (Culicoides, Diptera: Ceratopogonidae). In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Meiswinkel R. & Baylis M. (1998). Morphological confirmation of the separate species status of Culicoides (Avaritia) nudipalpis Delfinado, 1961 and C. (A.) imicola Kieffer, 1913 (Diptera: Ceratopogonidae). Onderstepoort J. Vet. Res., 65, Veterinaria Italiana, 40 (3), 2004

129 33. Meiswinkel R. & Linton Y.-M. (2003). Afrotropical Culicoides Latreille (Diptera: Ceratopogonidae): morphological and molecular description of a novel fruit-inhabiting member of the Imicola Complex, with a re-description of its sister species C. (Avaritia) pseudopallidipennis Clastrier. Cimbebasia, 19, Mellor P.S. & Pitzolis G. (1979). Observations on breeding sites and light-trap collections of Culicoides during an outbreak of bluetongue in Cyprus. Bull. Entomol. Res., 69, Mellor P.S., Boorman J.P.T., Wilkinson P.J. & Martinez-Gomez F. (1983). Potential vectors of bluetongue and African horse sickness in Spain. Vet. Rec., 112, Mellor P.S., Boned J., Hamblin C. & Graham S. (1990). Isolations of African horse sickness virus from vector insects made during the 1988 epizootic in Spain. Epidemiol. Infec., 105, Miranda M.A., Borràs D., Rincón C. & Alemany A. (2003). Presence in the Balearic Islands (Spain) of the midges Culicoides imicola and Culicoides obsoletus group. Med. Vet. Entomol., 17, Mirzaeva A.G. (1984). A review of biting-midges of the subgenus Avaritia Fox (Diptera, Ceratopogonidae, genus Culicoides Latr.) from Siberia (in Russian, English summary). Entomol. Obozr., 63, Navai S. (1977). Biting midges of the genus Culicoides (Diptera; Ceratopogonidae) from Southwest Asia. PhD thesis, University of Maryland, 201 pp. 40. Saitou N. & Nei M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molec. Biol. Evol., 4, Savini G., Goffredo M., Monaco F., De Santis P. & Meiswinkel R. (2003). Transmission of bluetongue virus in Italy. Vet. Rec., 152, Sebastiani F., Meiswinkel R., Gomulski L.M., Guglielmino C.R., Mellor P.S., Malacrida A.R. & Gasperi G. (2001). Molecular differentiation of the Old World Culicoides imicola species complex (Diptera, Ceratopogonidae), inferred using random amplified polymorphic DNA markers. Molec. Ecol., 10, Sellers R.F., Pedgley D.E. & Tucker M.R. (1978). Possible windborne spread of bluetongue to Portugal, June-July J. Hyg., Camb., 81, Sellers R.F., Gibbs E.P.J., Herniman K.A.J., Pedgley D.E. & Tucker M.R. (1978). Possible origin of the bluetongue epidemic in Cyprus, August, J. Hyg., Camb., 83, Swofford L. (1999). Phylogenetic analysis using parsimony, version 4. Sinauer, Sunderland, Massachusetts. 46. Wirth W.W. & Dyce A.L. (1985). The current taxonomic status of the Culicoides vectors of bluetongue viruses. In Bluetongue and related orbiviruses (T.L. Barber & M.M. Jochim, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Wirth W.W. & Hubert A.A. (1989). The Culicoides of Southeast Asia (Diptera: Ceratopogonidae). Mem. Amer. Entomol. Inst., 44, Wittmann E.J., Mellor P.S. & Baylis M. (2001). Using climate data to map the potential distribution of Culicoides imicola (Diptera: Ceratopogonidae) in Europe. Rev. Sci. Tech. Off. Int. Épiz., 20, Veterinaria Italiana, 40 (3),

130 Vet. Ital., 40 (3), Epidemiology and vectors Infection of the vectors and bluetongue epidemiology in Europe P.S. Mellor Institute for Animal Health, Pirbright Laboratory, Ash Road, Pirbright, Woking, Surrey GU24 0NF, United Kingdom Summary The author describes some of the factors controlling the infection and transmission of bluetongue (BT) virus (BTV) by vector species of Culicoides. Also outlined are certain important features of the recent BT epizootic in the Mediterranean Basin, concentrating on those aspects involving vector transmission and overwintering of the virus. The regions affected by the outbreaks and the BTV serotypes involved are set out, the distribution of the major vector, C. imicola is described and the impact of novel vector species of Culicoides and a possible overwintering mechanism for the virus in Europe, are discussed. Keywords Bluetongue Culicoides Epidemiology Europe Infection Novel vectors Overwintering. There are some named species of Culicoides biting midges, in the world but only very few have been shown to act as competent vectors of bluetongue (BT) virus (BTV) (29). In North America, C. sonorensis is the major vector with C. insignis also being important in parts of some of the southern states (e.g. Florida). Further south in the Caribbean areas and in South America C. insignis and C. pusillus are thought to be the major transmitters of BTV. In Australia several vectors are thought to exist, with C. brevitarsis the most important but C. wadai, C. actoni and C. fulvus being of local significance. In most of the Old World, C. imicola is by far the most important BTV vector although recently one of its sibling species in Africa, C. bolitinos, has also been shown to be competent and, in Europe, C. pulicaris and C. obsoletus have both recently been implicated in BTV transmission. Vector competence The reasons why some species of Culicoides contain some individuals that are competent to support the replication and transmission of BTV subsequent to oral ingestion, and why others are not, are complex (reviewed in 22). Basically, when an arbovirus like BTV is ingested by a haematophagous insect during blood feeding, the virus passes into the lumen of the hind part of the mid-gut. It then has to gain access to the body of the insect proper before the potentially hostile environment in the gut lumen inactivates it or before it is excreted. If the virus is to be orally transmitted by the vector, as is BTV, it must reach the salivary glands with or without amplification in other susceptible tissues, multiply in them and finally be released with the saliva into the salivary ducts where it is available to infect a second vertebrate host during a subsequent bite. The details of this cycle (its duration, the tissues infected, the titre of virus produced, the proportion of insects infected, the transmission rate) are controlled by a range of inter-dependent variables (the virus, the insect host and environmental factors particularly temperature). In the case of BTV, a series of barriers or constraints are known to exist within the bodies of non-vector species of Culicoides and even within a variable proportion of individuals within vector species, that act to prevent infection or else restrict infection in such a way as to prevent transmission. Briefly, the major barriers that an arbovirus may have to surmount upon being deposited in the midgut of an haematophagous insect in order to develop a fully patent infection and so be available for oral transmission are as follows: infection of the mid-gut cells: mid-gut infection barrier (MIB) escape of progeny virus from the mid-gut cells into the haemocoel: mid-gut escape barrier (MEB) dissemination of virus through the haemocoel to the salivary glands (and ovaries if transovarial transmission is to occur): dissemination barrier (DB) infection of the salivary glands: salivary gland infection barrier (SGIB) Veterinaria Italiana, 40 (3),

131 release from the salivary glands into the salivary ducts: salivary gland escape barrier (SGEB). In addition, there is a further barrier to surmount if the virus is to be transmitted transovarially, i.e. the transovarial transmission barrier (TOTB). Figure 1 depicts a summary of these barriers to infection and transmission, and highlights those that have so far been identified in the BTV-Culicoides system. Within a vector species of Culicoides susceptibility to infection is a genetically heritable trait. This means that different populations of the same species may have widely varying oral susceptibilities to infection and transmission of a particular serotype or strain of BTV dependent upon the genotypes prevalent in the parental populations from which they were derived. Consequently, results obtained by testing one or several populations of a suspect vector species of Culicoides cannot necessarily be extrapolated across all or most populations of that species. This situation can make it difficult to estimate the importance of a suspect vector species unless exhaustive testing has been undertaken across many populations of that species using a range of BTV serotypes and strains. As may be seen when considering the current outbreaks of BT in Europe, in practice, this can result in incorrect assessments being made of the significance of such novel vectors. Regions affected by the epizootics in Europe The current epizootic of BT in Europe is presumed to have started in October 1998 on four Greek islands adjacent to the Anatolian coast of Turkey. The serotype involved was identified as BTV-9 and this was the first occasion that this serotype had ever been identified in Europe (30). As expected, transmission of BTV in the Greek islands seemed to end during December 1998, probably because of zoosanitary measures introduced by the Greek Veterinary Authorities and because adult vector populations of C. imicola in this region are at a minimum from this time of the year (M. Patakakis, personal communication). However, in June 1999, BTV 9 was again reported from eastern Europe, first in south-east Bulgaria and then in rapid succession from European Turkey and from the north-east of mainland Greece (30). The outbreaks in Greece were particularly active and extended across the north of the country from Evros on the Turkish/Bulgarian border to west of Thessaloniki and then south to Larissa, Magnisia and Evia. These outbreaks also continued sporadically in Greece during the summer of Of particular concern in Greece was the discovery that, in addition MIB midgut escape barrier MEB midgut escape barrier DB dissemination barrier SGIB salivary gland infection barrier SGEB salivary gland escape barrier TOTB transovarial transmission barrier Figure 1 Barriers to the infection and transmission of arboviruses by insect vectors Route of virus dissemination in the Culicoides-BTV system highlighted in red 168 Veterinaria Italiana, 40 (3), 2004

132 to BTV-9, serotypes 4 and 16 were also active in the country. All three serotypes have been identified historically from locations further east (e.g. Anatolian Turkey, Syria, Jordan and Israel), and BTV-16 was also isolated from Izmir Province in western Anatolian Turkey in August 2000 (30) (C. Hamblin, personal communication). Meanwhile, in January 2000, BT was reported for the first time ever from Tunisia and the virus involved was identified as BTV-2 (30). The origins of this incursion are obscure but are likely to be separate from the one involving Turkey, Greece and Bulgaria. As foot and mouth disease virus had also entered Tunisia (and Algeria) during 1999, probably via cattle imported from Côte d Ivoire and Guinea into Algeria (18) it is possible that BTV could have followed a similar route. Cattle in Africa often experience sub-clinical infections with BTV, and BTV-2 is common in several areas of sub-saharan West Africa (15, 16). From June 2000, additional, more widespread outbreaks of BT due to serotype 2 of the virus were reported from Tunisia and in July 2000 the same serotype was also identified in Algeria. These outbreaks continued until September and October In addition, seropositive animals were also recorded from 18 provinces stretching across northern Morocco (30). Then, in August 2000, came a new and rather dramatic turn of events, when BTV-2 was confirmed for the first time ever in Italy. The island of Sardinia was affected first but by October, BTV-2 had also spread to Sicily and southern mainland Italy. Outbreaks continued in Italy into December 2000 but were not reported later in the winter of , presumably because the cooler temperatures prevailing at those times significantly reduced vector abundance. Unfortunately the outbreaks in Italy were not the end of the matter and, in October 2000, BTV-2 was also recorded for the first time on the French island of Corsica, and on the Spanish islands of Menorca and Mallorca. Just as in Italy, outbreaks in the Balearics continued into November and December 2002 but were not recorded later in that winter (30). So at the end of 2000 there were two main foci of BTV infection established in Europe. One to the east, centred on Turkey, Bulgaria and Greece involving BTV serotypes 4, 9 and 16, and the second to the west, centred on Italy and certain French and Spanish islands (and parts of North Africa) involving BTV-2. Unfortunately both BT foci continued to be active in 2001 and Eastern focus In 2001, the eastern focus spread to new areas in north-west Greece and Lesbos during August and October, and also extended northwards to involve Kosovo and south-east Serbia in August, west Bulgaria and Macedonia in September and Croatia in December. Unofficially, seroconversions were also reported in sheep and cattle in European Turkey. Importantly in 2001 the Greek Veterinary Authorities identified an additional BTV serotype (BTV-1) in Greece. This brought the number of BTV serotypes active in Europe to 5 (BTV-1, -2, -4, -9 and -16). The origin of the new serotype is difficult to determine but recent studies by Mertens and his colleagues (iah.bbsrc.ac.uk/dsrna_virus_proteins/orbivirus-phyloge netic-trees.htm) have shown that, phylogenetically, the BTV-1 reported in Greece is much closer to topotypes of this virus from the east (India) than from Africa. This suggests that this incursion, like those involving serotypes 4, 9 and 16, originated to the east of Europe, unlike the incursion of BTV-2. In 2002 Greece was reported free from BTV but disease due to BTV-9 was still widely reported from the Balkans (southern Bulgaria, Kosovo, Montenegro and Bosnia). The presence of BTV-9 in the Balkans in 2002 makes this the fourth consecutive year that this serotype has been present in this region of Europe. Furthermore, the outbreaks in Serbia in 2001 and those in Bosnia in 2002 are, at N, the most northerly ever reported in Europe. Western focus In May 2001, Italy recorded renewed BTV activity in Sardinia. Further activity was reported in June (Calabria), August (Basilicata) and September (Lazio and Tuscany). The virus was also active for the second year running on the French island of Corsica but the Spanish authorities reported that the Balearics were free from infection. The vast majority of these outbreaks were due to BTV-2 but significantly and for the first time in this western focus, BTV-9 was also identified in the regions of Calabria and Basilicata in Italy. In 2002, BT continued to be widespread in Italy (Lazio, Tuscany, Campania, Abruzzo, Puglia, Basilicata, Calabria, Sardinia and Sicily) but was not reported from any other country previously involved in the western focus. Once more, most of the outbreaks in Italy seemed to be due to BTV-2 but BTV-9 was again detected and more widely so than in 2001 (Calabria, Puglia, Basilicata, Campania and Sicily) and a third serotype (BTV-16) was also reported from southern Italy (Calabria) (G.Savini 2002, personal communication). The occurrence in the western focus of BTV serotypes 9 and 16 that had previously only been reported from the eastern Veterinaria Italiana, 40 (3),

133 focus is a matter of concern and represents a clear link between the two foci. In the Mediterranean Basin, the general trend of BTV movement seems to be in a westerly direction. In the eastern focus, this is seen with all of the serotypes involved, which were first detected in the extreme east of the region but spread across Turkey, Greece and Bulgaria, and at least two of the serotypes have now moved as far west as Italy. In the western focus, BTV was first detected in Tunisia but then infected animals were reported in Algeria and Morocco in North Africa and, subsequent to Italy, on Corsica and on two of the Balearic islands of Spain. Despite the widespread outbreaks of BTV- 2 in Italy, this virus has not yet been identified in Greece, to the east, despite the geographical proximity of the two countries. The vector situation Culicoides imicola C. imicola is the major Old World vector of BTV and African horse sickness virus (AHSV). It has long been considered to be the only major vector of these viruses in the Mediterranean Basin because, prior to 1998, every incursion of BTV or AHSV into the Mediterranean Basin has been restricted to regions where C. imicola is known to be present. The known distribution of C. imicola in the Mediterranean Basin prior to 1999 is shown in Figure 2. This distribution is a composite of the findings of many investigations conducted since 1974 (3, 4, 5, 6, 11, 17, 21, 24, 25, 26, 31). However, following the 1998 incursion of BTV into Europe, further investigation into the distribution of C. imicola in the Mediterranean Basin has been galvanised and workers in Greece, Italy, France, Spain, Turkey and Bulgaria have all made significant contributions to our knowledge of where this species exists and, equally importantly, where it is apparently absent (1, 2, 4, 7, 8, 14, 20, 32) (J.-C. Delécolle 2002, personal communication; S. de La Rocque 2002, personal communication; M. Patakakis 2002, personal communication; I. Burgu 2002, personal communication; A. Martinez 2002, personal communication; J. Delgado and P. Collantes 2002, personal communication; N.K. Nedelchev and G. Georgiev 2002, personal communication). The second line on Figure 2 has been derived from these accumulated findings and depicts the northern limits of C. imicola, the so-called imicola-line, as known in It is likely that in the future this imicola-line will move even further northwards. When comparing the distribution of C. imicola in 2002 with earlier findings, it is at once apparent that the species is present in major areas of Europe where it was previously thought to be absent (e.g. south-eastern Spain, the Balearics, Corsica, Sardinia, Sicily, much of mainland Italy and much of mainland Greece). In Italy C. imicola has now been recorded as far north as 44 N (14). Whether this represents a real and general movement of C. imicola northwards and westwards perhaps in response to climate-change, or whether it is merely a reflection of more intensive sampling, is difficult to say although in some regions, at least, the former seems to be the most likely option. For example, in mainland Greece, a survey in 1983 included 16 collections of Culicoides comprising 19 species from locations where there was intense BTV transmission in 1999 but not a single isolate of C. imicola was recorded (25). Known northern limits of Culicoides imicola distribution prior to 1999 Known northern limits of Culicoides imicola distribution in 2002 Figure 2 The known northernmost limits of Culicoides imicola in the Mediterranean Basin prior to 1999 and in Veterinaria Italiana, 40 (3), 2004

134 However, more recent surveys conducted between 1999 and 2001 from locations in the same regions have found C. imicola to be common (M. Patakakis 2001, personal communication). Be that as it may, the newly discovered presence of populations of C. imicola in Menorca, Mallorca, Corsica, Sardinia, mainland Italy, mainland Greece (and Tunisia) are clearly the main reason that BTV was able to be transmitted in these regions. However, between 1999 and 2002, BTV was also transmitted in many other locations (northern Greece, European Turkey, Bulgaria, Serbia, Croatia, Macedonia, Montenegro and Bosnia) that are apparently beyond this new imicola-line. In some of these areas, no vector Culicoides surveys have been performed, but in others, C. imicola has been sought for since the BTV outbreaks started but has not been recorded (e.g. northern Greece [Thrace], European Turkey and Bulgaria). Indeed, in Bulgaria, Culicoides surveys have been conducted at intervals for over 10 years, 29 species of Culicoides have been recorded but not a single specimen of C. imicola has been identified (10, 12, 13). These findings suggest in the strongest possible terms that in some parts of Europe as yet unidentified, or unconfirmed, BTV vector(s) are present. Novel vector species of Culicoides in Europe In all of the non-imicola areas where BTV has been detected, Culicoides of the C. obsoletus and C. pulicaris groups are by far the most abundant and prevalent biting-midge species (30). C. obsoletus and C. pulicaris have long been suspected of being BTV vectors, mainly on the basis of BTV isolations from C. obsoletus made in Cyprus (23) and AHSV isolations from mixed pools of C. obsoletus and C. pulicaris made in Spain (28). In this context, it should be borne in mind that BTV and AHSV tend to utilise as vectors the same Culicoides species (21). However, vector competence studies carried out on a population of C. obsoletus and C. pulicaris in the United Kingdom (UK) during the 1980s recorded oral susceptibility rates of less than 2%, in comparison with 19.5% for a known major vector, C. sonorensis (= variipennis in part) (19, 27). At the time, these findings and the restriction of all previous BTV outbreaks in Europe entirely to imicola-areas (see above) suggested that these other species were likely to be of only minor importance as BTV (or AHSV) vectors. Nevertheless, as is explained below, in hindsight perhaps, more significance should have been attributed to these early findings. Firstly, the high abundance and high survival rates of C. obsoletus and C. pulicaris as exhibited in Bulgaria in 1999 (30) could compensate for low levels of vector competence. Precisely this situation exists in Australia where C. brevitarsis is considered to be the most important BTV vector, mainly on the basis of its high abundance and prevalence, even though its vector competence levels (0.3%) are very low indeed (34). Secondly, as expression of competence by a vector species for a particular virus is an hereditary trait (see section on Vector competence), populations with high, intermediate and low levels of competence are to be expected. It may be, therefore, that populations of these species in Bulgaria, northern Greece and European Turkey express higher levels of competence for BTV than had hitherto been suspected. In this context, the only previous vector competence information about these species, which was derived from testing single populations of C. obsoletus and C. pulicaris in southern England some 20 years ago, is clearly insufficient to extrapolate meaningfully across the whole of Europe or even the UK for that matter. Consequently, until very recently other populations of C. obsoletus and C. pulicaris throughout Europe have remained untested and their levels of competence for BTV were completely unknown. However, new and important information from workers on mainland Italy and in Sicily is now beginning to demonstrate the real significance of these two groups of Culicoides in the epidemiology of BT. Savini et al. (33) have drawn attention to the fact that outbreaks of BT have occurred in areas of southern Italy where C. imicola is scarce or absent. In these areas, Obsoletus Complex midges apparently comprised more than 90% of over collected Culicoides and more than 95% were parous individuals (cf. parous midges have taken and digested a blood-meal and so are the only specimens likely to harbour virus). These authors also reported that three isolations of BTV were made from tested Obsoletus Complex midges (estimated as 95% C. obsoletus sensu stricto and 5% C. scoticus) which gives a rate of one BTV isolation per approximately insects. Elsewhere on mainland Italy, other workers have also isolated BTV from pools of Obsoletus Complex collected in BT outbreak areas, at locations where C. imicola was either scarce or absent (C. De Liberato 2003, personal communication). These findings of Italian workers are the first to positively incriminate C. obsoletus midges as BTV vectors since 1977 when BTV was first isolated from this species group in Cyprus (23). Equally important are the recent findings of Sicilian scientists. Caracappa et al. (9) report that in Sicily, C. imicola is much less common than in most other regions of Italy where BT has occurred, and in 2002 he and his colleagues made five isolations of BTV from parous, non-engorged C. pulicaris at locations and times when C. imicola was absent. These are the Veterinaria Italiana, 40 (3),

135 first such field isolations ever reported from this species. Importantly, these five isolations were made from just 987 specimens, which gives an isolation rate of about 1 per 200 insects and indeed 1 isolation was made from a pool of only 10 individuals. This is a far higher rate than has been recorded from Obsoletus Complex midges elsewhere and may imply the presence of a highly competent population of C. pulicaris in Sicily. Significantly, Caracappa and his colleagues made no BTV isolations from 724 C. obsoletus midges collected at the same locations and at the same times, suggesting that this species may be a less competent vector, at least in parts of Sicily. These findings confirm the importance of C. pulicaris and Obsoletus Complex midges as probable BTV vectors in parts of Italy and suggest that these species are likely to be important in some locations, even within the latitudes where C. imicola occurs. Those responsible for implementing vector control campaigns might do well to be aware of this. The discoveries in Italy also strongly implicate C. pulicaris and Obsoletus Complex midges as the likely vectors in those more northerly parts of Europe where BT has occurred beyond the imicola-line (northern Greece, European Turkey and the Balkans). The recognition of probable new vector species of Culicoides in Europe is a matter of considerable concern. Although expanding its range, C. imicola is still restricted, climatically, to the more southerly parts of the continent, C. pulicaris and Obsoletus Complex midges exhibit no such restriction and are probably the most common Culicoides species across the entire central and northern regions of Europe. Theoretically, this would seem to put at risk most of these areas. However, BTV has not yet occurred across Europe, so the question is, why not? The reasons are likely to be complex but will probably include some or all of the following: a) vector abundance may be lower further north b) most northerly populations of the novel vectors may be less efficient BTV transmitters than most populations of C. imicola c) ambient temperatures further north tend to be lower. As BTV is transmitted more quickly and more efficiently at higher temperatures than lower, transmission may not be possible or may only be possible for a brief part of the year or in restricted locales in more northerly climes d) adult vectors may be completely absent for part or much of the year in northerly areas due to the harsh winters so that BTV-infected animals either recover or die before the new vector season begins. However, climate-change, through its effects upon the vectors, is likely to moderate all of these factors, allowing BTV to spread further northwards and be transmitted more efficiently over a greater proportion of the year than at present (30). Additionally, and specifically in connection with point (d) above, the apparent ability of BTV to overwinter in some of the more northerly, currently infected areas (e.g. Bulgaria, the Balkans, northwestern Greece and European Turkey) where adult Culicoides are absent for much of the year is extraordinary (30). If confirmed, this suggests that a novel overwintering mechanism may be involved that could extend the area perceived to be at risk to BTV significantly further north, irrespective of climate change. Interestingly, such a mechanism has recently been described which postulates the presence of covertly infected, seropositive ruminants in which BTV-persistently infected γδ T-cells are recruited into the skin during late summer and autumn in response to the biting activity of the new generation of vectors which thereby become infected and so initiate fresh transmission cycles (35). All of these recent findings, taken in concert with climate change, suggest that the current outbreaks of BT in Europe are unlikely to be a one-off aberration but may be a sign of things to come. The strong control measures presently being deployed against the virus in Europe may result in its temporary elimination from this continent but further, more frequent and widespread incursions are a real prospect. Novel aspects of the incursion into Europe There are a number of factors that make the current BTV incursions into Europe unique. These are as follows: 1) For the first time, multiple serotypes of BTV have been involved (i.e. 1, 2, 4, 9 and 16), most of which are new to Europe. 2) In some areas, live BTV vaccines of several different types have been deployed for the first time (i.e. including serotypes 2, 3, 4, 8, 9, 10 and 11). 3) The major vector, C. imicola, has been found in many new locations in and around Europe (e.g. Sardinia, Sicily, mainland Italy, Corsica, mainland Greece, Majorca, Menorca, eastern mainland Spain, European Turkey and Tunisia). 4) BTV has entered almost all of these locations but in some areas has penetrated beyond the imicola line, reaching further north than ever before (e.g N in Serbia and Bosnia). 172 Veterinaria Italiana, 40 (3), 2004

136 5) Novel vector species of Culicoides are therefore involved in certain outbreak areas (e.g. Bulgaria, northern Greece, European Turkey, Macedonia, Serbia, Kosovo, Albania, Croatia and Bosnia). Circumstantial evidence (e.g. abundance and prevalence) suggests that these novel vectors are likely to be members of the Obsoletus and Pulicaris species complexes. These are the most common Culicoides species across northern Europe. 6) Multiple isolations of BTV have now been made from C. pulicaris and species of the Obsoletus Complex in Italy, confirming their vector status and suggesting that they may also be involved in BTV transmission within the latitudes where C. imicola occurs. 7) In some of the more northerly areas, BTV has overwintered in the absence of adult vectors. A novel mechanism must be involved (e.g. the γδ T-cell mechanism). Acknowledgements The author is grateful to many colleagues throughout Europe and North Africa for sharing their unpublished information and their thoughts. This paper was produced while the author was in receipt of grants from the European Union (Contract Nos. QLK2-CT and QLK2-CT ) and the Department for Environment, Food and Rural Affairs (DEFRA) (Reference SE 2610 Effect of temperature and regional origin on the vector capacity of British and European Culicoides spp.). References 1. Anon. (2002). Arbovirus vectors and disease. First annual report of the EU-funded project QLK2-CT , 40 pp (plus annexes). 2. Anon. (2003). Arbovirus vectors and disease. Second annual report of the EU-funded project QLK2-CT , 43 pp (plus annexes). 3. Boorman J. (1974). Culicoides (Diptera: Ceratopogonidae) from Cyprus. Cah. ORSTOM. Entomol. Med. Parasitol., 12, Boorman J. (1986). Presence of bluetongue virus vectors on Rhodes. Vet. Rec., 118, Boorman J.P.T. & Wilkinson P.J. (1983). Potential vectors of bluetongue in Lesbos, Greece. Vet. Rec., 113, Bouayoune H., Touti J., El Hasnaoui H., Baylis M. & Mellor P.S. (1998). The Culicoides vectors of African horse sickness virus in Morocco: distribution and epidemiological implications. Arch. Virol., 14 [Suppl.], Capela R., Purse B.V., Pena I., Wittmann E.J., Margarita Y., Capela M., Romao L., Mellor P.S. & Baylis M. (2003). Spatial distribution of Culicoides species in Portugal in relation to the transmission of African horse sickness and bluetongue viruses. Med. Vet. Entomol., 17 (2), Caracappa S., Torina A., Loria G.R., Riili S., Monteverde P. & Goffredo M. (2001). Sorveglianza entomologica per il rischio di bluetongue in Sicilia: dati preliminari. Atti Soc. Ital. Buiatria, 33, Caracappa S., Torina A., Guercio A., Vitale M., Calabro A., Purpari G., Vitale F. & Mellor P.S. (2003). Identification of a novel bluetongue virus vector species of Culicoides in Sicily. Vet. Rec., 153, Dilovski M., Nedelchev N. & Petkova K. (1992). Studies on the species composition of Culicoides potential vectors of the virus of bluetongue in Bulgaria. Vet. Sci., 26, Gallo C., Guercio V., Caracappa S., Boorman J.P.T. & Wilkinson P.J. (1984). Indagine sieroentomologica sulla possibile presenza del virus bluetongue nei bovini in Sicilia. Atti Soc. Ital. Buiatria, 16, Georgiev G., Martinov S. & Veleva E. (2001). Studies on the distribution and epizootiology of bluetongue disease in ruminants in Bulgaria. Biotech. Biotech. Equip., 15, Glouchova V.M., Nedelchev N.K., Rousev I. & Tanchev T. (1991). On the fauna of blood-sucking midges of the genus Culicoides (Diptera: Ceratopogonidae) in Bulgaria. Vet. Sci., 25, Goffredo M., Satta G., Torina A., Federico G., Scaramozzino P., Cafiero M.A., Lelli R. & Meiswinkel R. (2001). The 2000 bluetongue virus outbreak in Italy: distribution and abundance of the principal vector Culicoides imicola, Kieffer. In Proc. 10th International Symposium of the American Association of Veterinary Laboratory Diagnosticians (AAVLD), Salsomaggiore, Parma, 4-7 July. AAVLD, Ames, Herniman K.A.J., Gumm I.D., Owen L., Taylor W.P. & Sellers R.F. (1980). Distribution of bluetongue virus and antibodies in some countries of the eastern hemisphere. OIE Bull., 92, Herniman K.A.J., Boorman J.P.T. & Taylor W.P. (1983). Bluetongue virus in a Nigerian dairy cattle herd. 1: Serological studies and correlation of virus activity to vector populations. J. Hyg., Camb., 90, Jennings M., Boorman J.P.T. & Ergun H. (1983). Culicoides from western Turkey in relation to bluetongue disease of sheep and cattle. Rev. Elev. Med. Vet. Pays Trop., 36, Knowles N.J. & Davies P.R. (2000). Origin of recent outbreaks of foot-and-mouth disease in North Africa, the Middle East and Europe. In Report of the Session of the Research Group of the Standing Technical Committee of the European Commission for the control of foot-and-mouth disease. Borovets, Veterinaria Italiana, 40 (3),

137 Bulgaria, 5-8 September. FAO, Rome, Appendix 3, Mellor P.S. (1992). Culicoides as potential orbivirus vectors in Europe. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Mellor P.S. (1993). African horse sickness surveillance in Algeria: insect vector identification and sentinel herds. Report to IAEA for Project ALG/5/014-06, Mellor P.S. (1993). African horse sickness: transmission and epidemiology. Vet. Res., 24, Mellor P.S. (2000). Replication of arboviruses in insect vectors. J. Comp. Pathol., 123, Mellor P.S. & Pitzolis G. (1979). Observations on breeding sites and light-trap collections of Culicoides during an outbreak of bluetongue in Cyprus. Bull. Entomol. Res., 69, Mellor P.S., Boorman J.P.T., Wilkinson P.J. & Martinez-Gomez F. (1983). Potential vectors of bluetongue and African horse sickness viruses in Spain. Vet. Rec., 112, Mellor P.S., Jennings M. & Boorman J.P.T. (1984). Culicoides from Greece in relation to the spread of bluetongue virus. Rev. Elev. Med. Vet. Pays Trop., 37, Mellor P.S., Jennings M., Wilkinson P.J. & Boorman J.P.T. (1985). Culicoides imicola, a bluetongue virus vector in Spain and Portugal. Vet. Rec., 116, Mellor P.S. & Jennings D.M. (1988). British vectors of bluetongue virus. In Orbiviruses and birnaviruses. Proc. dsrna Symposium, September Natural Environmental Research Council (NERC) Institute, Oxford, Mellor P.S., Boned J., Hamblin C. & Graham S. (1990). Isolations of African horse sickness virus from vector insects made during the 1988 epizootic in Spain. Epidemiol. Infec., 105, Mellor P.S., Boorman J. & Baylis M. (2000). Culicoides biting midges: their role as arbovirus vectors. Ann. Rev. Entomol., 45, Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean Basin Vet. J., 164, Ortega M.D., Mellor P.S., Rawlings P. & Pro M.J. (1998). The seasonal and geographical distribution of Culicoides imicola, C. pulicaris group and C. obsoletus group biting midges in central and southern Spain. Arch. Virol., 14 [Suppl.], Roger F.L. (2002). Emergence of bluetongue disease in the Mediterranean Basin: modelling locations at risk for potential vectors (Culicoides spp.) using satellite imagery. MSc thesis, University of London, 84 pp. 33. Savini G., Goffredo M., Monaco F., de Santis P. & Meiswinkel R. (2003). Transmission of bluetongue virus in Italy. Vet. Rec., 152, Standfast H.A., Dyce A.L. & Muller M.J. (1985). Vectors of bluetongue virus in Australia. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I.Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, Takamatsu H., Mellor P.S., Mertens P.P.C., Kirkham P.A., Burroughs J.N. & Parkhouse R.M.E. (2003). A possible overwintering mechanism for bluetongue virus in the absence of the insect vector. J. Gen. Virol., 84, Veterinaria Italiana, 40 (3), 2004

138 Vet. Ital., 40 (3), 175 Epidemiology and vectors Possible overwintering mechanism of bluetongue virus in vectors D.M. White (1, 2), W.C. Wilson (1), C.D. Blair (3) & B.J. Beaty (3) (1) United States Department of Agriculture (USDA)-Agricultural Research Service (ARS), Arthropod-Borne Animal Diseases Research Laboratory, Dept. 3354, 1000 E. University Avenue, Laramie, WY 82071, United States of America (2) Present address: Centers for Disease Control and Prevention, NCID/DVRD/Special Pathogens Branch, 1600 Clifton Rd NE, Building 15-SB, Mailstop G14, Atlanta, GA 30333, United States of America (3) Arthropod-borne and Infectious Diseases Laboratory, Department of Microbiology, Immunology, and Pathology, Colorado State University, Fort Collins, CO , United States of America Summary The overwintering mechanism of bluetongue virus (BTV) has eluded researchers for many years. While overwintering in the vertebrate host has been the main hypothesis, it has been shown that several arboviruses overwinter in their invertebrate vectors. Overwintering Culicoides sonorensis larvae were collected from long-term study sites in northern Colorado and examined for the presence of BTV nucleic acid by reverse transcriptase (RT)-nested polymerase chain reaction (PCR). Sequences from the S7 segment of BTV RNA were detected in 17 of 56 (30%) pools comprised of larvae and pupae collected in 1998, and in 32 of 319 (10%) pools comprised of adults reared from larvae collected in BTV was not isolated from the adult pools. Additionally, cell lines derived from culicoid embryos collected at the same site, or derived from material collected during a BTV outbreak, were positive for BTV nucleic acid. Interestingly, in contrast to the S7 segment, the L2 RNA segment could only rarely be detected in any of the field-collected larvae, and was not detected in the culicoid cell lines. These data suggest that BTV may not require abundant expression of the outer coat genes to persist in the insect vector. This could also explain the low rate of isolation of virus from insects. If the vertebrate cell receptor ligand VP2 (which is encoded by L2) is expressed at very low levels in the insect, traditional vertebrate cell-based isolation methods would be inefficient until the virus had amplified itself sufficiently to express all virus genes required for vertebrate cell infection. Further research in this area will define and characterise the role of the vector in the overwintering of BTV, and will help in focusing control efforts. Keywords Bluetongue Culicoides sonorensis Overwintering United States of America Vectors Virus. The full article of this contribution: Studies on overwintering of bluetongue viruses in insects is published in the Journal of General Virology, 86 (2), February 2005, (vir.sgmjournals.org/future/86.2.shtml). Veterinaria Italiana, 40 (3),

139 Vet. Ital., 40 (3), Modelling the distribution of bluetongue vectors M. Baylis, L. O Connell & B.V. Purse Institute for Animal Health, Ash Road, Pirbright, Surrey GU24 0NF, United Kingdom Summary Recent epizootics of Culicoides-borne disease in the Mediterranean Basin have stimulated the development of climate-driven models for vectors. Predictor variables come from two main sources, weather data and satellites. Generally, models for Culicoides imicola combine temperature and moisture variables. The best weather models explain 75-85% of the variance in observed data for C. imicola, but satellite models perform better (85-95% of variance). Predictions of models for other regions appear mixed, with successes and failures. The failures indicate the need to: explore and incorporate other factors that may affect Culicoides populations, such as soil characteristics, host type and wind speed develop more complex models, recognising that different climate variables affect different stages of the life-cycle e.g. biological models. The very rapid spread in the distribution of C. imicola in recent years suggests that global warming may be a less important driver of change than other, currently unknown, factors. Keywords Bluetongue Climate Culicoides imicola Global warming Land surface temperature Modelling Normalised difference vegetation index Satellite Weather. Recent epizootics of Culicoides-borne disease in the Mediterranean Basin, such as the outbreak of African horse sickness (AHS) in Spain, Portugal and Morocco (2), and the ongoing bluetongue (BT) epizootic (1), have stimulated the development of models of the spatial distribution of the Culicoides vectors, principally C. imicola. The models are climate-driven, as the life-history traits (e.g. survival and fecundity) of Culicoides are highly influenced by climatic factors, and this, in turn, affects their distribution and abundance, and the distribution and intensity of the diseases they transmit. The purposes of the modelling are as follows: to improve our understanding of the biotic and abiotic determinants of the distribution of the vectors to identify the limits to the distribution permitted by suitable climate, and thereby to define new areas at risk and areas that might remain diseasefree to investigate how the distribution might change under scenarios of future global warming. This review is limited to consideration of models of C. imicola. Many weather variables influence Culicoides populations via their effects on survival and fecundity (11). High temperatures yield smaller, less fecund adults and both high and low temperatures lead to larval and adult mortality (17). The pupae of C. imicola may drown if breeding sites become too wet following rain. Equally, lack of rain and the concomitant low soil moisture may desiccate larvae. Low relative humidity at high temperature causes low saturation deficit, which will desiccate adults. There is evidence that high windspeeds affect local Culicoides populations, perhaps via dispersal (blowing midges away, both literally and metaphorically), starvation (adults unable to find feeding opportunities during consistently windy conditions) or desiccation (3, 4). The development of climate-driven models for the distribution of Culicoides vectors requires uncovering the statistical relationship between climate variables and the presence/absence, or abundance, of the vectors. The vector picture to be modelled is blurred, however, by the understandable tendency to estimate 176 Veterinaria Italiana, 40 (3), 2004

140 the vector population using light traps. The number of Culicoides spp. caught per night in a light trap is proportional to the size of the local population (which is what needs to be measured), multiplied by the activity rate and multiplied again by the efficiency of the trap. The latter two quantities are themselves affected by the weather (11, 12), and this confounds attempts to determine the effect of climate on the population. Activity rate is defined as the proportion of the population that is active in a given night and, for Culicoides, this rate is highly variable (2). Thus, Culicoides tend to be less active when nights are very hot or cold or when relative humidity is very low; they tend not to fly when it is raining; and they are not active when windspeeds exceed certain levels. Activity is also dependent upon light levels. Trap efficiency is a measure of the ability of a trap to catch the vectors in its immediate area: if trap efficiency is 100%, the trap catches all vectors that approach it on a given night. For Culicoides, it is likely that trap efficiency is reduced by light sources other than that of the trap, such as moonlight, as this may distract the midges from approaching the trap. More significantly, the use of a suction device in light traps means that their efficiency reduces with an increase in windspeed. This effect has been known for other insects for many years (16) and the small size of Culicoides means that the effect is especially significant. Controlled experiments conducted at Pirbright (12), in which Culicoides were released into a room and caught in a trap exposed to different windspeeds, indicated a logarithmic relationship between windspeed and the reduction in trap efficiency (Fig. 1). A repercussion of this effect in the field is that trap catches outside a stable reduce at higher windspeeds, while those inside the stable do not (10). How can the precise relationship be determined between Culicoides population size and climatic variables when, for example, temperature also reduces activity rates and windspeed reduces both activity and trap efficiency? In most experiments, the best estimate of a population variable is taken to be the average of a number of samples. However, the approach used here was borrowed from the field of remote sensing, called maximum compositing. For satellite images of the earth, the detected level of radiance from the earth may be decreased but not increased (compared to the true level of radiance) by cloud cover or other atmospheric effects. The best estimate of the true level is, therefore, the maximum level across several images recorded at different times. Similarly, Culicoides trap catches, as a proportion of the local population size, can only be reduced (and not increased) by the effects of weather on activity rate and trap efficiency (2). Thus, the best estimate of population size will be the single greatest catch over a time period. Reduction in trap efficiency (%) Windspeed (m/s) y = Ln(x) R 2 = Figure 1 The effect of windspeed on the ability of a light/suction trap to catch Culicoides A known number of Culicoides nubeculosus were released into a chamber containing a trap, with fans positioned at different distances to create different windspeeds Reprinted with permission from L. O Connell (12) In order to include a range of environmental conditions in the spatial model, population size must be estimated at as many sites as possible. However, maximum compositing of Culicoides catches requires repeated trapping at a given site over short time periods. Given that trapping effort is always limited, it is pertinent to ask how many trapping occasions are needed to provide an accurate estimate of population size and/or presence? To address this question, data from twenty-two sites in Morocco were analysed. These were sampled weekly between 1993 and 1995 as part of a vector surveillance campaign set up in response to the epizootic of AHS. At sites of known C. imicola presence, the probability of a positive catch is about 0.22 in January, increasing to 0.77 in October. These monthly probabilities were used to estimate the number of zero catches that are required to give 95% confidence in the absence of C. imicola (Fig. 2a). In October in Morocco, two nights with zero catches are required; three nights in August, September and November; four nights in June and five or six nights in July and December; and seven nights in May. This analysis includes data from a small number of sites where only a handful of C. imicola were trapped over two years, and in areas that were free from AHS. If these sites are excluded, the probability of a positive catch (at the remaining sites, where disease risk is assumed) is 0.25 in January and 0.86 in October. The number of nights trapping for 95% confidence in absence (Fig. 2b) is two in September and October, Veterinaria Italiana, 40 (3),

141 three in June, August and November, four in July and five in April, May and December. In a country such as Morocco, which has experienced significant outbreaks of both AHS and BT, trapping for only one night is not sufficient at any time of year to give confidence that C. imicola, or disease risk, is absent. a) All known positive sites in Morocco Probability of positive catch (%) January February March April May June July August September October November December 1 Night Month 2 Nights 3 Nights Cut-off b) Excluding sites where only one or two Culicoides imicola were caught over two years, and which did not experience African horse sickness Probability of positive catch (%) January February March April May June July August September October November December 1 Night Month 2 Nights 3 Nights Cut-off Figure 2 The probability of catching at least one Culicoides imicola at known positive sites in Morocco, when setting traps for one to three nights The dashed line indicates the threshold above which there can be 95% confidence that Culicoides imicola is absent from a site, if none have been caught Vector abundance, for the purposes of disease riskmapping, must be measured at a spatial and temporal scale that is relevant to disease transmission. For Culicoides, population sizes change rapidly over time but generally show annual variation with one or two discrete peaks that for C. imicola tend to coincide with the seasonal peak of vector-borne disease in endemic areas. Maximum composited numbers over two-week intervals, averaged across two years, were used as measures of abundance for climate modelling in Morocco (3) but the significant trapping effort required precludes the use of this estimate of relative abundance across extensive sets of sites. Analysis of catch data from Morocco indicates that the maximum of catches over two nights during the late summer peak is significantly correlated with the fuller measure of abundance, and thus, this low effort trapping regime may permit abundance to be estimated at a large number of sites. Climatic predictor variables used for modelling Culicoides presence/absence or abundance have been derived from two main sources, ground-collected weather data and satellite imagery. The former have ready biological significance, but are recorded synoptically at a relatively small number of weather stations that are often distant from trap sites, and between which it is necessary to interpolate, to obtain a continuous layer of climate information. Weather stations are expensive to buy and the data can be laborious to process. Finally, it may be difficult to obtain comparable weather data for other regions to which predictions might be extended. In contrast, satellite images are usually free and give global coverage at scales ranging from many kilometres to a few metres, thereby requiring no interpolation and facilitating extensive prediction. A further benefit is that suitably processed imagery is generally a better predictor of Culicoides than are ground-collected climate data. For example, windspeed and the annual minimum normalised difference vegetation index (NDVI, a measure of vegetation biomass obtained from earth-orbiting satellites) were the best predictors of the abundance of C. imicola in Morocco (3). In South Africa, a model using NDVI and land surface temperature (LST, a measure of ground temperature, again from satellites) accounted for 67% of variance in the abundance of C. imicola, compared to only 45% for a model using temperature and rainfall recorded by weather stations (5). Nevertheless, models for Culicoides continue to be developed using weather data. Logistic regression was used to model the presence/absence of C. imicola in Iberia using historic ( ) weather data (as equivalent data were available for all of Europe, thereby facilitating extrapolation) (18). The best model, which correctly predicted presence/absence at 83% of sites in Iberia, comprised the temperature 178 Veterinaria Italiana, 40 (3), 2004

142 of the coldest month (the lowest mean daily minimum), the warmest month (the highest mean daily maximum) and the number of months with a mean temperature of 12.5 C. This model is unusual in lacking moisture variables, which other studies indicate to be very important for C. imicola. Not surprisingly, this temperature-driven model predicts C. imicola territory to occur in a broad band across southern Europe, with few predictions of absence at latitudes lower than Madrid. This model has recently been criticised (7). However, it should be noted that the predictions being criticised are substantially different from those presented by Wittmann et al. (18), as they were generated from a different source of weather data, without calibration. Another logistic regression model, of the presence/absence of C. imicola at >500 sites in Italy (9) was based on 10 km 10 km grid square data surfaces derived by interpolation between relatively few weather stations that were usually more than 40 km apart. At many sites, trapping was undertaken for a single night only, with the sampling regime biased towards areas in which BT virus (BTV) was present (and thus presumably suitable for vectors), and, indeed, this model performed less well than others (77.5% correct predictions). The best model comprised the annual mean daily minimum temperature, the annual mean daily minimum relative humidity and altitude. The model performed well for many parts of Italy (but not southern Sicily, where C. imicola is predicted to be prevalent), but was not externally validated by extrapolation to other countries. Extrapolation to other countries was a primary objective of an 8-variable model of the abundance of C. imicola in Portugal, Spain and Morocco developed by discriminant analysis and 8 8 km fourierprocessed satellite imagery (6). Three abundance ranges were considered, of which the lowest included 0 (i.e. absence). The best model correctly predicted the abundance range at 93.2% of sites and included, as the most important variables, proxies for both temperature and moisture as well as altitude. The high accuracy of predictions in Iberia/Morocco encouraged more extensive extrapolation (Fig. 3). Suitable conditions for C. imicola were predicted for eastern Spain, the Balearics, northern North Africa, Sardinia, Sicily, parts of Lazio and Puglia (Italy), eastern mainland Greece, the Peloponnese, Rhodes and Cyprus. All of these areas are now known to harbour C. imicola. There are false-positive predictions for southern Sicily and there are significant false-negative predictions. Most notably, it fails to predict the occurrence of C. imicola in Corsica and eastern Calabria (Italy). Highest Intermediate Lowest No prediction Figure 3 Abundance of Culicoides imicola predicted by a model derived from the observed abundances at 44 sites in Iberia and Morocco Reprinted with permission from Baylis et al. (5) Veterinaria Italiana, 40 (3),

143 A similar model, based on recently collected vector data from Portugal (maxima of two summer catches) and fine resolution (1 1 km) satellite imagery (15), correctly predicted presence/absence at 95% of sites, and abundance range at 87% of sites. As before, the model comprised correlates of both temperature and moisture. Predictions across Europe are broadly similar to those of Baylis et al. (6). A 10-variable model for Sicily correctly predicted the presence/absence of C. imicola at 87% of sites (13). The model correctly predicted the presence of C. imicola in parts of Iberia (including the Balearics), Sardinia and parts of Greece. However, the predicted distribution was generally much more restricted than the observed, with little prediction of presence in southern Spain, Corsica, mainland Italy and North Africa. This limitation probably results from the low number of presence sites for C. imicola in Sicily, such that only a restricted range of potential C. imicola habitats were included in the training set. This raises the question of why C. imicola is not more widespread in Sicily, despite most models predicting the island to be climatically suitable. It has been suggested that the porosity of the volcanic soils in Sicily may be unfavourable for breeding sites (7). Finally, a 4-variable model for Corsica correctly predicted the C. imicola abundance range at 78% of sites (14). The predictions across Europe were less restrictive than in the study conducted in Sicily. The presence of C. imicola was correctly predicted in southern Portugal, south-western Spain, the east coast of Spain, the Balearic islands, the south coast of mainland France, the Corsican lowlands, most of Sardinia, northern and eastern Sicily, the west coast of mainland Italy, much of Calabria and Basilicata and parts of Greece, Turkey and Cyprus. There were some significant false-positive results, however: a low risk is associated with western Greece and Crete. In summary, the best C. imicola models correctly predict 80%-95% of observations when built from satellite data and 75%-85% when based on weather data. In most cases, the models incorporate both temperature and moisture variables. Most models indicate that presence/abundance is most likely at moist, lower-altitude sites that are warm but not too hot. There is disparity between satellite-derived models in different regions as to whether mean, minimum, phase or amplitude NDVI variables are most important or whether minimum land surface temperature is an important predictor. While this may arise due to differences in the resolution of the imagery used or due to differences in the trapping regimes, it may also reflect biological differences in habitat requirements across a species range. This variation in importance of particular variables between regions and models may be unimportant for the production of preliminary risk maps, but detailed maps will require investigation of the relationship between satellite-derived variables and biological processes that determine population performance. To date, no spatial models have attempted to incorporate the effects of windspeed as a possible variable that may control population size. Satellitederived data layers for windspeed are becoming available, but their utility has not yet been confirmed. There is a further need to investigate and include certain soil characteristics that may mitigate against C. imicola breeding, irrespective of the suitability of the climate. To date, only one model has been used to examine the possible effects of global warming on the future distribution of C. imicola (18). With a putative 2 C increase in mean temperature, the distribution of C. imicola is predicted to expand to the north, with significant probabilities of occurrence in southern France and northern Italy. This prediction is not surprising. As described earlier, this model is entirely temperature-driven and any effects of global warming on moisture (rainfall, humidity) or other weather variables cannot be taken into consideration. Indeed, it is the temperature-dependence of this model that almost certainly explains why it is the only model that has been used to investigate possible effects of climate change. Most studies have detected effects of both temperature and moisture on the occurrence of C. imicola. All climate change scenarios include effects on rainfall and soil moisture, but these are still difficult to predict with accuracy, and vary considerably geographically. Until the expected effects of climate change are known more clearly, it is too early to derive predictions on the future distribution of the moisture-sensitive C. imicola. More importantly, the scale of change in recent years, with the supposed spread of C. imicola to the Balearics, Corsica, Sardinia, Sicily, mainland Italy, France and Greece and a significant northward spread, is very rapid compared to the observed rate of global warming and, to date, there has been no detectable spread in Portugal in the last decade (8). While climate change may be in-part driving the spread of C. imicola (1), it is important to consider that there may be other contributing factors, as yet unidentified, and also that the pattern may arise from increased surveillance effort such that there is an increased rate of discovery of C. imicola populations. Finally, recent preliminary models have indicated that climatic determinants of distribution differ between Culicoides species, probably due to their differing life histories. Predictive risk maps for Culicoides-borne disease must be based on species- 180 Veterinaria Italiana, 40 (3), 2004

144 specific spatial models for C. imicola, and novel vector species to avoid omitting extensive regions at risk of transmission by the latter. Acknowledgements Grateful thanks are extended to many people who have helped the authors to formulate ideas and undertake these studies, in particular: Hassan Bouayoune, Santo Caracappa, Ruben Capela, Hassan El Hasnaoui, Youssef Lhor, Rudy Meiswinkel, Philip Mellor, Michael Patakakis, Isabel Pena, Peter Rawlings, David Rogers, Paola Scaramozzino, Andy Tatem, Alessandra Torina, Jamal Touti, Emma Wittmann and many others too numerous to mention. This work was funded largely by European Union Grants 8001-CT , TS3-CT , IC18-CT and QLK2-CT References 1. Baylis M. (2002). The re-emergence of bluetongue. Vet. J., 164, Baylis M., El Hasnaoui H., Bouayoune H., Touti J. & Mellor P.S. (1997). The spatial and seasonal distribution of African horse sickness and its potential Culicoides vectors in Morocco. Med. Vet. Entomol., 11, Baylis M., Bouayoune H., Touti J. & El Hasnaoui H. (1998). Use of climatic data and satellite imagery to model the abundance of Culicoides imicola, the vector of African horse sickness virus, in Morocco. Med. Vet. Entomol., 12, Baylis M. & Rawlings P. (1998). Modelling the distribution and abundance of Culicoides imicola in Morocco and Iberia using climatic data and satellite imagery. In African horse sickness (P.S. Mellor, M. Baylis, C. Hamblin, C. Calisher & P.P.C. Mertens, eds). Arch. Virol. [Suppl.], 14, Baylis M., Meiswinkel R. & Venter G.J. (1999). A preliminary attempt to use climate data and satellite imagery to model the abundance and distribution of Culicoides imicola (Diptera: Ceratopogonidae) in southern Africa. J. Sth Afr. Vet. Assoc., 70, Baylis M., Mellor P.S., Wittmann E.J. & Rogers D.J. (2001). Prediction of areas around the Mediterranean at risk of bluetongue by modelling the distribution of its vector using satellite imaging. Vet. Rec., 149, Calistri P., Goffredo M., Caporale V. & Meiswinkel R. (2003). The distribution of Culicoides imicola in Italy: application and evaluation of current Mediterranean models based on climate. J. Vet. Med. B, 50, Capela R., Purse B.V., Peña I., Wittmann E.J., Margarita Y., Capela M., Romao L., Mellor P.S. & Baylis M. (2003). Spatial distribution of Culicoides species in Portugal in relation to the transmission of African horse sickness and bluetongue viruses. Med. Vet. Entomol., 17, Conte A., Giovannini A., Savini L., Goffredo M., Calistri P. & Meiswinkel R. (2003). The effect of climate on the presence of Culicoides imicola in Italy. J. Vet. Med. B, 50, Meiswinkel R., Baylis M. & Labuschagne K. (2000). Stabling and the protection of horses from Culicoides bolitinos (Diptera: Ceratopogonidae), a recently identified vector of African horse sickness. Bull. Entomol. Res., 90, Mellor P.S., Boorman J. & Baylis M. (2000). Culicoides biting midges: their role as arbovirus vectors. Ann. Rev. Entomol., 45, O Connell L. (2002). Entomological aspects of the transmission of arboviral diseases by Culicoides biting midges. PhD, University of Bristol. 13. Purse B.V., Tatem A.J., Caracappa S., Rogers D.J., Mellor P S., Baylis M. & Torina A. (2004). Modelling the distributions of Culicoides bluetongue virus vectors in Sicily in relation to satellite-derived climate variables. Med. Vet. Entomol., 18 (2), Roger F.L. (2002). Emergence of bluetongue disease in the Mediterranean basin: modelling locations at risk for potential vectors (Culicoides spp.) using satellite imagery. MSc, University of London. 15. Tatem A.J., Baylis M., Mellor P.S., Purse B.V., Capela R., Pena I. & Rogers D.J. (2003). Prediction of bluetongue vector distribution in Europe and North Africa using satellite imagery. Vet. Microbiol., 97 (1-2), Taylor L.R. (1962). The absolute efficiency of insect suction traps. Ann. Appl. Biol., 50, Wittmann E. (2000). Temperature and the transmission of arboviruses by Culicoides biting midges. PhD, University of Bristol. 18. Wittmann E.J., Mellor P.S. & Baylis M. (2001). Using climate data to map the potential distribution of Culicoides imicola (Diptera: Ceratopogonidae) in Europe. Rev. Sci. Tech. Off. Int. Épiz., 20 (3), Veterinaria Italiana, 40 (3),

145 Vet. Ital., 40 (3), Bluetongue surveillance methods in the United States of America D. Dargatz (1), K. Akin (2), A. Green (1), M. Herrero (1), S. Holland (3), A. Kane (1), D. Knowles (4), T. McElwain (5), K.M. Moser (6), E.N. Ostlund (6), M. Parker (1), E.T. Schmidtmann (7), A. Seitzinger (1), L. Schuler (8), G. Stevens (2), L. Tesar (9), L. White (10), L. Williams (11), N. Wineland (1) & T.E. Walton (1) (1) USDA/APHIS-VS Centers for Epidemiology and Animal Health, 2150 Centre Ave, Bldg B, MS2E7, Fort Collins, CO , United States of America (2) USDA/APHIS-VS Nebraska Area Office, PO Box 81866, Lincoln, NE 68516, United States of America (3) State of South Dakota, 411 S. Fort St., Pierre, SD 57501, United States of America (4) USDA/ARS, ADBF 3005, Washington State University, Pullman, WA , United States of America (5) Washington Animal Disease Diagnostic Laboratory, College of Veterinary Medicine, Washington State University, Pullman, WA , United States of America (6) USDA/APHIS-VS National Veterinary Services Laboratories, 1800 Dayton Road, Ames, IA 50010, United States of America (7) USDA/ARS Arthropod-Borne Animal Diseases Research Laboratory, 5031 Agriculture Bldg, University of Wyoming, Laramie, WY 82071, United States of America (8) State of North Dakota, 600 E. Boulevard Ave., Dept. 602, Bismarck, ND , United States of America (9) USDA/APHIS-VS South Dakota Area Office, 314 S. Henry, Suite 100, Pierre, SD 57501, United States of America (10) USDA/APHIS-VS North Dakota Area Office, 3509 Miriam Ave., Suite B, Bismarck, ND , United States of America (11) State of Nebraska, PO Box 94787, 301 Centennial Mall, South, 4th Floor, Lincoln, NE , United States of America Summary Historical surveillance for bluetongue virus (BTV) exposure in the United States of America (USA) has relied on periodical serological surveillance using samples collected from cattle at slaughter. Most of this surveillance has been focused on the north-eastern portion of the USA due to the lack of competent vectors of BTV in this region. For most of the states tested in this region, the prevalence of seropositive animals has been less than 2%. Recently, a study was conducted in north-central USA using sentinel cattle herds. Results of serological testing showed an increasing gradient of exposure from north to south. In addition, detection of Culicoides sonorensis showed a similar gradient with detection in the northern areas being relatively rare. The results of these studies indicate that cattle herds in the northern and north-eastern areas of the USA are likely to be free of BTV. Keywords Bluetongue virus Culicoides Sentinel herds Serology Surveillance United States of America Vector trapping. Bluetongue (BT) viruses (BTV) have been and continue to be a significant impediment to international livestock trade. Various activities have been undertaken over the years to develop information on the risk associated with movement of animals from certain geographic areas to other areas thought to be free of BTV infection. The United States of America (USA) has used a variety of methods to monitor the prevalence and distribution of BTV infection of cattle in various parts of the country. Historically, surveillance has been conducted in the USA on alternate years by testing serum collected from adult cattle at slaughter (1, 2). This programme has focused on the north-east segment of the country where the vector for the United States serotypes of BTV has been absent. Generally, the prevalence of seropositive animals in these studies has been 2% or less in each of the states with samples tested. The results for the most recent serological survey are reported elsewhere in these proceedings by Ostlund et al. (3). 182 Veterinaria Italiana, 40 (3), 2004

146 More recently, a BT Surveillance Pilot Project (BSPP) was designed and implemented to evaluate a sentinel herd surveillance system for BTV and to evaluate the ecology of BTV and the vector of BTV in the north-central part of the country. Cattle herds in North Dakota, South Dakota and Nebraska were selected to participate in the study. Blood samples were collected from cows in each of the herds before and after a vector season. Blood samples were tested for antibody to BTV using a commercial competitive enzyme-linked immunoassay (c-elisa). When only a single animal on an operation gave a positive result to the c-elisa, that sample was tested by virus neutralisation (VN) to five BTV serotypes (BTV-2, BTV-10, BTV-11, BTV-13 and BTV-17). Operations were considered positive if they produced two or more c-elisa positive samples or a single c-elisa positive sample that was also positive to the VN test. Insect traps were set on approximately half of the operations to determine the presence of Culicoides sonorensis, the primary North American vector of BTV. Data were also collected on operation management, proximity to vector habitats and animal characteristics. Overall, 149 operations were initially included in the study, most of which (93%) were beef cattle operations. Of the 144 operations where blood samples were collected prior to the vector season, samples were tested. Of these, samples (14.1%) gave positive results to the c-elisa test and 54 (37.5%) operations were considered positive (two or more c-elisa positive results or one c-elisa positive that was also positive by VN). Fewer operations were positive in the northern reaches of the study area than in southern latitudes. Of the 128 operations with blood samples collected after the vector season, 49 (38.3%) were classified as positive. Again, a similar geographic pattern of positive results was seen as before the vector season. For the second sampling, 975 (17.3%) of the animals tested were positive using the c-elisa. The results of the BSPP suggest the following: the prevalence of BTV infection in the more northern latitudes is very low or zero the distribution of C. sonorensis is consistent with previously drawn maps the sentinel cattle herd system is useful to explore the spatial distribution of infectious agents and vectors and to generate hypotheses regarding the ecology of animal diseases. Plans for the future include further descriptive analysis of the data and epidemiological modelling to evaluate risk factors of BTV. In addition, the data will be analysed for geo-spatial factors (weather and topographical) related to vector distribution. Acknowledgements This project was an extensive effort made possible by the co-operation and resources from many State and Federal agencies and groups. The collaboration of these groups, of the individuals within them, and of the participating operation owners is gratefully acknowledged. References 1. Mecham J. & Monke D. (1999). Report of the Committee on bluetongue and bovine retroviruses. Proc. US Anim. Hlth Assoc., 103, Mecham J. & Monke D. (2001). Report of the Committee on bluetongue and bovine retroviruses. Proc. US Anim. Hlth Assoc., 105, Ostlund E.N., Moser K.M., Johnson D.J., Pearson J.E. & Schmitt B.J. (2004). Distribution of bluetongue in the United States of America, In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Insect (vector) trapping was conducted on 68 operations. Of these, 32 (47.1%) had catch samples with C. sonorensis. The spatial distribution of C. sonorensis-positive catch samples was similar to that of the positive cattle blood samples, as fewer samples were positive in the more northern and eastern parts of the study area. Veterinaria Italiana, 40 (3),

147 Vet. Ital., 40 (3), Bluetongue surveillance methods in an endemic area: Australia L.F. Melville Department of Business, Industry and Resource Development, Darwin, NT 0800, Australia Summary Surveillance for bluetongue (BT) viruses (BTV) has been carried out in the Northern Territory, Australia since The number of sites, intensity of sampling and methods of testing have varied during this period. Monthly serology is conducted at a number of sentinel sites and intensive weekly sampling for virus isolation is conducted at the site of highest known arboviral activity. This has enabled the isolation of all eight BTV serotypes identified in Australia. Natural viraemias are between one and eight weeks. No additional serotypes have been isolated since However, genetic analysis of isolates has shown incursions of viruses of South-East Asian origin in 1992, 1994 and Trapping for Culicoides spp. has also been carried out at these sites on a regular basis. In recent years, an annual serological survey has supplemented the sentinel herds to more accurately define the BT zones described under OIE guidelines. Keywords Australia Bluetongue virus Culicoides Epidemiology Genetic analysis Surveillance Virus isolation. Introduction The Northern Territory of Australia lies in the semiarid tropics with the more northerly areas experiencing hot, wet summers and warm, dry winters. Cattle raising is conducted on large properties ranging in size from to one million hectares. There are no commercial sheep operations. Large areas in the north-east and south-west of the Northern Territory are unsuitable for raising cattle. The bluetongue (BT) endemic area is located between approximately 11 S and 17 S. Southerly extension to around 20 S occurs in some years, depending on seasonally variable risk factors. Surveillance for BT virus (BTV) has been conducted at between one and fourteen sites since the first isolation of a BTV in This isolation of BTV-20 was made from a mixed pool of Culicoides species collected approximately 50 km south-east of Darwin at Beatrice Hill (14). The number of sites and intensity of monitoring has varied during this period. Since the commencement of the National Arbovirus Monitoring Program (NAMP) in 1993 (9), between six and fourteen sentinel herds have been monitored each year. The most intensive studies have been conducted at Beatrice Hill. This site has previously been identified as having intense and varied arbovirus activity (15). Light traps for insect collections are also operated at most sites. In 2002 and 2003 sentinel herd monitoring was supplemented with an annual BTV serological survey at selected locations. This is designed to enable the BTV-free and surveillance zones to be more accurately defined under OIE guidelines. Materials and methods Sentinel herds Each sentinel herd consists of 10 to 25 young cattle initially seronegative for BTV antibodies. Cattle are replaced annually or earlier if seroconversion occurs. Serum samples are collected at regular intervals, usually monthly or quarterly. At Beatrice Hill, cattle are bled weekly throughout the year. Monitoring sites are confined to the areas of commercial cattle operations. Serological surveys Serological surveys are conducted on selected properties along six north/south transects. Up to three properties are chosen along each transect, concentrating on the margins of the endemic area. Between 80 to 100 cattle are bled at each site. Selected cattle must be between 18 and 24 months old, have been bred on the property and have never 184 Veterinaria Italiana, 40 (3), 2004

148 left the property. Between 10 and 15 locations are sampled each year. Serology Serum samples from sentinel animals and survey animals are tested for BTV antibodies by competitive enzyme-linked immunosorbent assay (c- ELISA) (10) and, if positive, by virus neutralisation (VN) tests (5). Virus isolation Lithium heparin blood samples collected weekly at Beatrice Hill are processed for virus isolation. Clots from serum samples collected at other sites are held for retrospective isolation following identification of seroconversion in individual animals. The isolation system uses embryonated chicken eggs (ECE) as described by Gard et al. (4), with the final cell passages through microtitre plates rather than cell culture tubes. The ECE homogenates are passaged onto mosquito cell cultures (Aedes albopictus C6/36). The second passage uses C6/36, BSR (13) and porcine stable equine kidney (PSEK) mammalian cell cultures. A third passage uses BSR and PSEK cell cultures to detect any cytopathology due to virus replication. Viruses are identified by a combination of BTV antigen capture ELISA (8) and VN (5). Molecular studies Each year, since 1992, a selection of BTV isolates have been subjected to genetic analysis (6, 7, 12). This has allowed identification of regional groupings of viruses and tracking of their movement (12). Entomology Light traps for insect collections are operated for three nights each month at most sentinel herd sites and other specific locations. Sampling is conducted throughout the year. Insects are collected into alcohol and the Culicoides species sorted and identified. More intensive monitoring by aspirating midges from cattle is also performed at Beatrice Hill. Data management Serological and entomological data are entered onto a web-based national database. This enables national zones to be developed for trade purposes (3). Results Virus isolation Since 1981, BTVs have been isolated from sentinel cattle at Beatrice Hill every year except 1990 (11). The level of activity and number of serotypes has shown marked annual variation. All eight BTV serotypes identified in Australia have been isolated at this site. BTV-1 is the most frequently observed. BTV-20, the original serotype isolated in 1975, was not seen again until Further isolates of BTV-20 were made in 1995, 1996 and BTV-21, isolated each year from 1981 to 1984, was not isolated again until Further isolates were made in BTV- 16, first isolated in 1986, has reappeared at regular intervals (1988, 1992 and 2001). Table I demonstrates the marked annual variation in infection rates and activity. Table I Characteristics of natural BTV infections, Year Detectable period of Bluetongue Sentinels viraemia (weeks) serotype infected Min. Max / / / / / / / / / / / / / / / / / / / Differences in the biological behaviour of these viruses are also demonstrated in Table I. In 1992, BTV-20 was isolated once from each of two animals of 75 monitored. Serology confirmed these were the only two of 75 infected. In 1995, BTV-20 was isolated from 45 of 68 animals monitored, with periods of viraemia up to four weeks, and in 1996, from 19 of 32 animals monitored with periods of viraemia lasting up to six weeks. Similar differences have been observed with BTV-21. During the period 1981 to 1984, isolations of BTV-21 were made from a small proportion of animals sampled and the periods of viraemia were less than two weeks. In 1995, BTV-21 was isolated from 59 of 68 animals monitored with periods of viraemia up to four weeks. Veterinaria Italiana, 40 (3),

149 No additional serotypes have been isolated since However, genetic analysis of isolates obtained since 1991 has shown incursions of BTVs of South- East Asian origin in 1992, 1994 and Virus isolation and genetic analysis has also been performed on selected samples from sentinel sites other than Beatrice Hill. BTV-1 and BTV-16 have been isolated from sites within 300 km from Darwin. BTV-1 isolates from these sites were also shown to be the same South-East Asian genotypes as the viruses isolated at Beatrice Hill. BTV-1 isolated from a site 600 km from Darwin was an Australian genotype. Serology Monitoring of sentinel sites within 300 km of Darwin shows BTV activity in most years. At inland sites and in lower rainfall areas, activity is observed infrequently. In these areas, serological surveys have been utilised to detect low levels of infection. Combined data from sentinel herds and serological surveys have shown that the boundaries of the endemic area can move considerable distances between seasons. Current sentinel and survey sites are shown in Figure 1. These sites are confined to the areas of commercial cattle enterprises. Sentinel Survey Figure 1 Bluetongue virus isolation and serology monitoring sites in the Northern Territory, Australia Entomology Five species of Culicoides are currently recognised as proven vector species in Australia (16). These species are C. brevitarsis, C. actoni, C. fulvus, C. wadai and C. dumdumi. All species except C. dumdumi have been identified at the sites within 300 km of Darwin, although C. wadai is present only in very low numbers and in occasional years at sites other than Beatrice Hill. C. dumdumi has only been identified at Beatrice Hill on limited occasions. Only C. brevitarsis has been identified at sites further inland. C. actoni, C. brevitarsis and C. fulvus are present at Beatrice Hill most months. C. fulvus numbers are higher during the wet summer months but C. actoni and C. brevitarsis numbers do not show any seasonal dependence. Ongoing studies aspirating midges from cattle have shown that peak activity of C. actoni occurs before sunset (1). This species is clearly under-represented in light traps. Discussion Over the past 25 years, monitoring and surveillance programmes in Australia have identified the BTV serotypes temporarily or permanently present and their distribution. While sentinel herds have been the basis of this monitoring, additional information, such as vector distribution, genetic analysis and results from serological surveys, has been incorporated to give a more detailed understanding of the viruses and their distribution. Data from an intensively monitored site at Beatrice Hill has demonstrated marked annual variation in the activity and infection rates of BTV. Despite the intensity of monitoring, the ecology and epidemiology of BTV is still poorly understood. Changes in the biological behaviour of the viruses have also been observed. These changes coincided with the detection of incursions of BTV viruses for which there is evidence in South-East Asia. Continued genetic analysis of a selection of BTV isolates obtained each year has shown at least one of these genotypes has become established in the Northern Territory. Analysis of isolates from elsewhere in Australia has shown these genotypes are currently confined to the area within 300 km of Darwin. In localised areas of Australia, where C. brevitarsis is the only vector species, attempts to model vector activity have been largely successful (2). In northern Australia where multiple vector species are present and incursions of viruses from South-East Asia are occurring, prediction of virus activity is very difficult. Similar problems are encountered when attempting to model BTV activity on a large scale where complex climatic and geographical interactions influence vector behaviour. Flexible surveillance methods are required to describe a dynamic BTV situation and to meet changing requirements for international trade. 186 Veterinaria Italiana, 40 (3), 2004

150 Acknowledgements The sentinel herds and vector monitoring are funded through the National Arboviruses Monitoring Program. This programme is managed by Animal Health Australia, with collaborative funding from the Australian State and Commonwealth governments, and the Australian livestock industries. The serological surveys are funded by Meat and Livestock Australia. References 1. Bellis G.A., Melville L.F., Hunt N.T. & Hearnden M.N. (2004). Temporal activity of biting midges (Diptera: Ceratopogonidae) on cattle near Darwin, Northern Territory, Australia. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Bishop A.L., Barchia I.M. & Spohr L.J. (2000). Models for the dispersal in Australia of the arbovirus vector, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae). Prev. Vet. Med., 47, Cameron A.R. (2001). Internet-based management of arbovirus monitoring data. In Arbovirus research in Australia (M.D. Brown, P. Ryan. & J.G. Aaskov, eds). Proc. 8th Symposium, South Stradbroke Island, 3-7 July CSIRO, Brisbane, Gard G.P., Weir R.P. & Walsh S.J. (1988). Arboviruses recovered from sentinel cattle using several virus isolation methods. Vet. Microbiol., 18, Gard G.P. & Kirkland P.D. (1993). Bluetongue virology and serology. In Australian standard diagnostic techniques for animal diseases (L.A. Corner & T.J. Bagust, eds). CSIRO Information Services, Melbourne, Gould A.R. (1987). The nucleotide sequence of bluetongue virus serotype 1 RNA 3 and a comparison with other geographic serotypes from Australia, South Africa, the United States of America and other orbivirus isolates. Virus Res., 7, Gould A.R. & Pritchard L.I. (1990). Relationships among bluetongue viruses revealed by outer coat protein nucleotide sequences. Virus Res., 17, Hosseini M., Hawkes R.A., Kirkland P.D. & Dixon R.J. (1998). Rapid screening of embryonated chicken eggs for bluetongue virus infection with an antigen capture enzyme linked immunosorbent assay. J. Virol. Methods, 75, Kirkland P.D., Ellis T., Melville L.F. & Johnson S. (1995). Australian National Arbovirus Monitoring Program a model for studying bluetongue epidemiology in China. In Bluetongue disease in South-East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Lunt R.A., White J.R. & Blacksell S.D. (1988). Evaluation of a monoclonal antibody blocking ELISA for the detection of group specific antibodies to bluetongue virus in experimental and field sera. J. Gen. Virol., 69, Melville L.F., Pritchard L.I., Hunt N.T., Daniels P.W. & Eaton B. (1997). Genotype evidence of incursions of new strains of bluetongue viruses in the Northern Territory. In Arbovirus research in Australia (B.H. Kay, M.D. Brown & J.G. Aaskov, eds). Proc. 7th Symposium, Surfers Paradise, November CSIRO, Brisbane, Pritchard L.I., Daniels P.W., Melville L.F., Kirkland P.D., Johnson S.J., Hunt R. & Eaton B.T. (2004). Genetic diversity of bluetongue viruses in Australasia. In Bluetongue, Part II (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Sato M., Tanaka H., Yamada T. & Yamamoto N. (1977). Persistent infection of BHK-21/WI-2 cells with rubella virus and characterisation of rubella variants. Arch. Virol., 54 (4), St George T.D., Standfast H.A., Cybinski D.H., Dyce A.L., Muller M.J., Doherty R.L., Carley J.G., Filippich C. & Frazier C.L. (1978). The isolation of a bluetongue virus from Culicoides collected in the Northern Territory of Australia. Aust. Vet. J., 54, Standfast H.A., Dyce A.L., St George T.D., Muller M.J., Doherty R.L., Carley J.G. & Filippich C. (1984). Isolation of arboviruses from insects collected at Beatrice Hill, Northern Territory of Australia, Aust. J. Biol. Sci., 37, Standfast H.A., Muller M.J. & Dyce A.L. (1992). An overview of bluetongue virus vector biology and ecology in the Oriental and Australasian regions of the Western Pacific. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Veterinaria Italiana, 40 (3),

151 Vet. Ital., 40 (3), Bluetongue virus surveillance in a newly infected area A. Giovannini (1), P. Calistri (1), A. Conte (1), L. Savini (1), D. Nannini (1), C. Patta (2), U. Santucci (3) & V. Caporale (1) (1) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy (2) Istituto Zooprofilattico Sperimentale della Sardegna G. Pegreffi, Via Duca degli Abruzzi N 8, Sassari, Italy (3) Ministero della Salute, Direzione Generale della Sanità Pubblica Veterinaria, Alimenti e Nutrizione, Piazzale Marconi, Rome, Italy Summary Keywords The occurrence of bluetongue virus (BTV) in areas in which intensive animal production is practised and where there is extensive movement of animals may have a substantial impact on both animal trade and husbandry. This situation occurred in Italy after the detection of bluetongue (BT) in August In such situations, surveillance can be used to delineate with precision those areas in which the virus is circulating and, consequently, to enforce the appropriate animal movement restrictions. Furthermore, surveillance can provide the data required to assess the risk associated with animal movement and trade. A structured surveillance system for the detection of BTV has been in place in Italy since August The system is based on the periodical testing of unvaccinated sentinel cattle that are uniformly scattered throughout Italy in a grid of 400 km 2 cells. The initial number of sentinel sites and sentinel animals, together with the width of the restricted area generated by the finding of a single seroconversion in a sentinel animal, were based on conservative criteria. Animal movement was restricted in a 20 km radius buffer zone around any positive serological result. This buffer area extends about km 2, equivalent to the area of three grid cells. After the commencement of the BT vaccination campaign in Italy, the sentinel surveillance system was the only way in which the effectiveness of vaccination and the incidence of infection in the non-immunised strata of ruminant animals could be estimated. Data collected over two years was used to assess the risks posed by the adoption of less conservative criteria for the delineation of infected areas and by the progressive relaxation of movement restrictions of vaccinated animals. In regard to the delineation of restricted areas, a new approach was tested and validated in the field, based on a Bayesian analysis of the positive and negative results obtained by the testing of sentinel animals from defined regions. For the risks related to animal movement, the surveillance data was used in risk assessment analyses to address the movement of slaughter and breeding animals from vaccinated/infected and surrounding areas to free areas. These risk assessments led to an amendment of the relevant European Union legislation. Finally, a Montecarlo simulation model was developed to simulate different sentinel system scenarios and to decrease the total number of sentinel animals and sites required by the surveillance system. The sentinel surveillance system was complemented by an entomological surveillance system based on the use of a number of permanent blacklight traps run weekly year-round and a number of mobile blacklight traps moved through the grid cells during the summer and autumn of each year. The aim of entomological surveillance was to define the maximum distribution of vectors and their seasonal population dynamics. Furthermore, the permanent trap system provides an early warning of the start of new epidemics. The data from the entomological surveillance system were also analysed to generate probability maps of the presence of the principal BTV vector (Culicoides imicola) and to define the geographical risk of BT on a nationwide basis, and to predict the geographical distribution and the short-term spread of C. imicola in Sardinia, using spatio-temporal data. The detection, since 2001, of BT outbreaks in the absence of C. imicola and the recent identification of BTV in midges of the Obsoletus Complex also stimulated investigations on other vector Culicoides, including C. obsoletus and C. pulicaris. Bluetongue Culicoides spp. Europe Italy Sentinel animals Surveillance. 188 Veterinaria Italiana, 40 (3), 2004

152 Since 1998, Southern Europe has been affected by the largest epidemic of bluetongue (BT) ever recorded. The disease initially affected Greece in 1998 with 84 outbreaks and cases and then spread to Bulgaria in 1999 (1 651 outbreaks in the two countries with cases), France (Corsica), Italy and Spain (Balearic islands) in 2000 (7 423 outbreaks in the three countries and cases, 98% of which were recorded in Italy), the Balkans and southern Italy in 2001 (7 358 outbreaks and cases, 94% of which were in Italy), and in 2002 the disease continued to spread to the Balkans and through much of southern Italy (Figs 1 and 2) Figure 1 Spread of bluetongue, caused by BTV-2, in Europe, Figure 2 Spread of bluetongue, caused by BTV-9, in Europe, In terms of both control actions and surveillance, neither the European Union (EU), nor the affected European countries were adequately prepared to cope with the problems posed by a vector-borne disease such as BT when it first appeared. This inadequate level of preparedness of Europe was reflected in existing regulations, because BT was given in Directive 92/119/EEC (22), together with other OIE List A diseases, such as rinderpest, sheep pox, swine vesicular disease etc., although Directive 1992/35/EEC (21) had already been issued to define measures against African horse sickness. Directive 92/119/EEC foresaw exclusive and direct control measures and the demarcation of a 3-km radius protection zone and a 10-km radius surveillance zone around each infected farm. Direct control measures included the slaughter of all susceptible animals on the farm and the possible extension of such measures to neighbouring farms suspected of housing infected animals. In November 2000, after the incursion of BT into Sardinia (Italy) on 18 August (9), the Balearic islands on 29 September (4) and Corsica on 18 October (3), the EU considered provisions already stipulated in Directive 92/35/EEC (21) and issued Directive 2000/75/EC (23) which fixes specific rules for the control and eradication of BT, and in particular the following: the demarcation of a protection zone with a radius of 100 km around the outbreaks or any farm on which virus circulation was confirmed the establishment of a surveillance zone of 50 km around the protection zone the slaughter of animals deemed necessary to prevent the spread of the epidemic and the destruction, elimination, incineration or burial of the carcasses of those animals the implementation of serological and entomological surveillance programmes in the protection and surveillance zones the prohibition of animal movement from protection and surveillance zones. To complement these prescribed measures, the Directive foresaw the possibility of a vaccination programme in the protection zone. Since no specific criteria were provided in the Directive for serological and entomological surveillance, each country can propose its own programme to the European Commission, taking into account specific needs and geographical features or husbandry practices. The application of Directive 2000/75/EC (23), through the adoption of Decision 2001/138/EC (13), has disrupted animal trade in at least a third of Italy. Long-term application of these regulations would probably have caused the irreversible decline of the entire cattle and small ruminant production system. Decision 2001/138/EC (13) instituted protection and surveillance zones in accordance with the criteria established by Directive 2000/75/EC Veterinaria Italiana, 40 (3),

153 (23) and more than a third of Italy was subjected to movement restrictions: 26% in the protection zone and 9% in the surveillance zone (Fig. 3). Animal movement was prohibited from the regions of Sicily, Sardinia and southern Italy (protection and surveillance zones) to all disease-free regions. Consequently, it was impossible to fatten and cull cattle, activities which traditionally take place in the Po Valley (free zone). Moreover, animal transhumance from the winter pastures in Puglia (surveillance zone) to the summer pastures in Abruzzi and Molise (free zones) was also stopped. Protection zone (26% of country) Protection zone (9% of country) Figure 3 Protection and surveillance zones identified in Italy, in accordance with Decision 2001/138/EC (13) All existing EU legislation regarding the compensation of farmers was developed in relation to contagious diseases of OIE List A, mainly foot and mouth disease and hog cholera (classical swine fever). The control strategy used in Europe for outbreaks of these diseases resorts to stamping-out of infected and in-contact animals and, since 1990, vaccination is only an ancillary measure. In this context, the principal economic losses are direct, due to the slaughter of infected and in-contact animals. According to European legislation, any compensation for indirect losses would perturb the market. In the case of vector-borne disease, especially when vaccination is the principal control measure, direct losses are virtually negligible, whereas indirect losses due to movement restrictions become considerable. Moreover, losses suffered by farmers whose livestock is subjected to movement restrictions also affect the earnings of farmers living in free areas. This has been recognised only very recently (July 2003) by the European Commission that enacted the Decision C(2003)2519fin (20), authorising the region of Sardinia to compensate cattle farmers for indirect losses due to movement restrictions imposed from 6 September 2000 to 31 December Prior to the 1998 outbreak in the Mediterranean (Figs 1 and 2), little information was available on the distribution of BTV vectors, i.e. of areas at risk of infection. In the early 1980s C. imicola was first identified in Spain, Portugal and on the Greek islands of Lesbos and Rhodes (5, 7, 29, 30). The presence of C. imicola was neither reported from the Balearic islands, Corsica, Sardinia, Sicily and Malta, nor was it found in the mainland territories of Greece and Italy (7). Further studies conducted in southern Italy, Sicily and the island of Pantelleria (Italy) in 1996 also did not detect the presence of C. imicola (33). It was only in June 2000 that C. imicola was first identified in western Sicily (26). It is noteworthy that, at least as far as Italy is concerned, the records show that in the past, C. imicola probably escaped notice in Italy as the majority of historical collections were made outside its current range (10). Even less was known regarding the distribution of other potential vectors of BTV, namely species of the Obsoletus and Pulicaris Complexes. Therefore, the Italian government decided to design a surveillance system that could delineate with precision the areas in which virus circulates and thus would enforce the appropriate movement restrictions on animal populations. The attempt to define, as precisely as possible, those areas that should be subjected to movement restrictions (i.e. less than the 100-km radius stipulated by the Directive), implies the existence of an effective early warning system, because the buffer territory of movement restrictions around areas of virus circulation is reduced. Furthermore, surveillance data were also used to provide information needed to assess the risks associated with animal movement and trade. The surveillance system comprised the following: clinical surveillance serological surveillance entomological surveillance. The surveillance programme Clinical surveillance Bluetongue spreads quickly after it enters a susceptible population. In 2000, the virus initially spread at a rate of around 30 km per week in Sardinia (Fig. 4) (9). When BTV serotype 4 (BTV-4) invaded Sardinia in August 2003, it spread at a rate similar to that recorded three years previously with BTV-2 (Fig. 5). 190 Veterinaria Italiana, 40 (3), 2004

154 First outbreak 18 August 26 August 2 September 14 August 21 August 28 August 9 September 16 September 23 September 4 September 11 September 18 September 30 September 7 October 14 October 25 September 2 October Figure 4 Weekly spread of bluetongue outbreaks in Sardinia, 18 August-14 October 2000 Figure 5 Weekly spread of bluetongue outbreaks in Sardinia, 14 August-2 October 2003 The incubation period of clinical BT disease is between 5 and 20 days. Thus, diseased animals are detected more rapidly by clinical examination than by serology. Therefore, in the summer and early autumn of 2000 (the epidemic peak), BT surveillance in Italy was based on clinical examination of sheep. Serological and virological diagnoses were only used as confirmatory tools on randomly sampled animals in newly infected areas (9). As the disease becomes established in a given territory, the importance of clinical surveillance decreases and the importance of serological surveillance increases. The presence of clinical BT was verified in suspected outbreaks; all susceptible animals were subjected to clinical examination and appropriate samples were collected for laboratory confirmation. Weekly clinical visits were made to farms to monitor the disease pattern and to update the census. Clinical visits were discontinued when diagnosis was not confirmed or when there was no evidence of virus circulation. When the presence of BT was confirmed in an area, clinical visits were extended to all sheep flocks within a radius of 20 km or 4 km in cases of clinical illness or sub-clinical infection, respectively. Clinical visits, sampling of animals and insect trappings were also performed on farms in BT-free regions that contained animals that had been imported from recently infected areas. During the first epidemic (August 2000-May 2001), a total of clinical visits (84% in Sardinia) were made. During the second epidemic (May 2001-April 2002), a total of clinical visits (2.4% in Sardinia) were made. Serological surveillance Based on the Italian experience, structured serological surveillance in a newly infected area can only start in the decreasing phase of the epidemic peak (Fig. 6). As already stated, clinical evaluation is more reliable than serology in detecting incursion of BTV in the initial phases of an epidemic. A sentinel network was already in place for more than 18 months prior to the 2003 epidemic of BTV-4 infection in Sardinia when 740 outbreaks involving animals in most of the southern and western coast of Sardinia were detected in 49 days (Fig. 5), whereas only 2 of more than tested sentinel animals seroconverted to BTV-4. Serological surveillance was based on a set of ad hoc surveys and a network of sentinel animals. Ad hoc surveys were the first serological activities; Veterinaria Italiana, 40 (3),

155 New outbreaks subsequently, based on the results of clinical surveillance and of ad hoc surveys, the network of sentinel animals was developed and implemented Start of ad hoc surveys Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May Figure 6 Monthy outbreaks of bluetongue, showing the peak of the first epidemic in Italy and ad hoc surveys, August 2000-May 2001 Ad hoc surveys The objectives of the ad hoc surveys conducted in the winter of were to determine the actual geographic distribution and the prevalence of BTV infection in sheep and cattle populations in the areas involved. Target populations and sampling design varied in the various regions that were monitored, depending on behaviour of the epidemic and existing knowledge (9). The objective of the monitoring, target population, sampling criteria and features of the infection are summarised in Table I. Sardinia was the first region affected and recorded the greatest number of outbreaks (91% of the outbreaks occurred in Sardinia). In the winter of , the geographic distribution of BTV infection appeared to be well defined, involving virtually the entire region, except the highest mountains. Therefore, the objectives of the monitoring programme, were to define the following: the prevalence of sub-clinical infection in infected flocks the involvement of cattle. These data were considered important for two main reasons, namely: to predict the likely outcome of the next epidemic, based on the level of immunity in the animal populations to evaluate the possible epidemiological role of cattle in the Sardinian environment. Table I Ad hoc monitoring performed during the winter of Epidemiological behaviour of infection, objectives of monitoring, target population and sampling criteria Region No. of clinical outbreaks Date of first outbreak Objective of monitoring Target population Sampling criteria Sardinia August 2000 Prevalence of infection in the cattle population Cattle Cluster sampling, stratified by cattle population density and by date of first detection of infection in the municipality Prevalence of subclinical infection in the sheep population Sheep Serological testing of animals in flocks clinically affected during the epidemic Sicily October 2000 Geographic distribution of infection in the cattle population Cattle Random sampling of the entire region, with 3 animals/km 2, so as to detect one infected animal per 100 km 2 Geographic distribution of infection in the sheep population Sheep Random sampling of the entire region, with 3 animals/km 2, so as to detect one infected animal per 100 km 2 Calabria October 2000 Geographic distribution of infection in the cattle population Cattle Random sampling of the entire region, with 3 animals/km 2, so as to detect one infected animal per 100 km 2 Prevalence of subclinical infection in the sheep population Sheep Serological testing of all animals in flocks within a radius of 20 km around any clinically affected flock during the epidemic Basilicata 0 Possible presence and geographic distribution of infection (region bordering infected areas) Cattle Random sampling of the entire region, with 3 animals/km 2, so as to detect one infected animal per 100 km 2 Salerno Province 0 Possible presence and geographic distribution of infection (province bordering infected areas) Cattle Random sampling of the entire province, with 3 animals/km 2, so as to detect one infected animal per 100 km Veterinaria Italiana, 40 (3), 2004

156 In Sicily, the infection commenced on 10 October and only a few scattered outbreaks were observed. The objectives of the monitoring programme, were to determine the following: the actual geographic distribution of the infection the involvement of the cattle population. In Calabria the infection started on 10 October as in Sicily, but the number of outbreaks observed was higher (589) than in Sicily (16) and the infection was clustered along the Ionian coast. The objectives of the monitoring programme, therefore, were to determine the following: whether or not the Tyrrhenian coast was affected the prevalence of sub-clinical infection in ruminants the involvement of cattle. In Basilicata and in the Salerno Province, the disease was not detected during the summer and autumn of 2000, but as these regions border Calabria, Basilicata and Salerno were monitored to determine if subclincal infections had occurred. Other ad hoc surveys were subsequently performed ( ) to verify the following: antibody coverage of vaccinated populations the risk linked to seasonal grazing practices the extent of virus circulation in certain zones in which individual animals seroconverted or where unexpected positive results were found. The results obtained from monitoring activities performed in the winter of were also used to plan the sentinel network adopted for Italy. Sentinel network Networks of sentinel animals have been implemented in several countries to monitor the presence and spread of vector-borne diseases. None of them, however, were designed to delineate the areas where virus circulation occurs with a precision significantly lower than 100 km. A sentinel system has been used in Australia since late 1975, but the number of sentinel sites was very limited. From 1988 to 1990, there were between 27 and 28 sites uniformly scattered within the known range of the vector (C. brevitarsis), as well as along the margins and outside its range (28). The total number of sites in 2001 was 94 (Fig. 7) (1), with the maximum density in New South Wales (one sentinel site every km 2 ). In Canada, a sentinel programme has been in place in the Okanagan Valley since 1988 (22). In 1988, the system was composed of five sentinel sites with up to 10 animals at each. The aim was to monitor activity of BTV and related orbiviruses (epizootic haemorrhagic disease viruses) in locations where BTV infection had previously been documented (since 1975). Sentinel monitoring locations Desert Figure 7 Location of sentinel sites in Australia, 2001 In the United States of America (USA), periodical surveys are performed, usually serological surveys of slaughter cattle for antibody against BTV. These surveys are conducted on sera from animals from low prevalence states or geographic areas (anticipated less than 2% antibody prevalence) (2, 32). The serological surveillance networks in place in other countries were not consistent with the requirements for serological surveillance of BTV infection in Italy. The Italian system was based on the following: the OIE Terrestrial animal health code, which states in Chapter (31) that random and targeted serological surveillance should provide at least a 95% level of confidence of detecting an annual seroconversion incidence of 2% in cattle (or other ruminant species if sufficient cattle are not available) a serological survey undertaken in the state of Queensland (Australia) in 1989 showed that the prevalence of serological positivity in cattle from locations with low prevalence of infection was on average 6.45% (34). The value of about 5% prevalence in areas of low prevalence of BTV infection was confirmed by the ad hoc monitoring conducted in Sardinia to evaluate the involvement of the cattle population in the BT epidemic. Therefore, it was decided to divide Italy into two main zones, based on the risk of infection (Fig. 8). The lower risk zone was subdivided into a Veterinaria Italiana, 40 (3),

157 grid of square cells of 400 km per side (1 600 km 2 per cell) and 148 sentinel animals were monitored in each cell (in compliance with the OIE requirements for free countries or zones to provide a confidence level of at least 95% to detect annual seroconversion incidence of 2% in cattle). In the higher risk area, a finer grid was designed, with squares of 20 km per side (400 km 2 ), to have a more precise definition of the distribution of infection. A total of 58 sentinel animals were included in each cell to confirm that the prevalence of infection was less than that observed in low prevalence areas (i.e. to provide at least a 95% level of confidence of detecting an annual seroconversion incidence of 5% in cattle). Since the minimal movement restricted zone is a circle of 20 km radius around the infected holding, the geographic density of sentinels is able, for any area of km 2 (equivalent to a circle with a radius of 20 km), to provide at least a 95% level of confidence of detecting a seroconversion incidence of 1.6% in cattle. traps were moved around the study areas to define the distribution of C. imicola, and permanent traps were operated at different sites in Italy from June to October 2001 to evaluate the effect of soil type on the presence of C. imicola. Entomological surveillance has been extended nationwide since October Blacklight traps were positioned in fixed locations in each province (Fig. 9) and operated weekly to monitor the population dynamics of Culicoides spp. Blacklight traps were also operated on a temporary basis in suspected or confirmed cases of virus circulation and whenever a more specific understanding of vector distribution was required. Geo-referenced traps Non-geo-referenced traps Figure 9 Location of permanent blacklight traps in Italy, 2001 Discussion Figure 8 Subdivision of Italy into high- and low-risk zones, September 2001 Area A: high risk zone (squares of 400 km 2 ) Area B: low risk zone (squares of km 2 ) Entomological surveillance Given the lack of knowledge on the distribution of vectors in Italy, an entomological surveillance programme was implemented in infected and adjoining areas at the beginning of the BT epidemic in October 2001 to map the distribution of vectors (with particular reference to C. imicola). Blacklight Decision 2001/138/EC (13) instituted protection and surveillance zones according to the criteria established by Directive 2000/75/EC (23) and a third of Italy was subjected to movement restrictions: 26% in the protection zone and 9% in the surveillance zone. Animal movement to diseasefree regions was prohibited from the regions of Sicily, Sardinia and southern Italy (protection and surveillance zones). Therefore, it was clearly essential from the beginning of the outbreak that the epidemiology of BTV infection in Italy needed to be defined, and that a comprehensive surveillance system be developed to accurately define the status of infection in the various regions. The implementation of a viable surveillance system that 194 Veterinaria Italiana, 40 (3), 2004

158 addressed the prescribed requirements was logistically challenging, and required the gathering of vast quantities of data and very intensive field activities. It involved the regular clinical evaluation (in most cases fortnightly) of more than sentinel animals and the placing of about 250 permanent insect traps throughout Italy. Information and data produced by the surveillance system constitutes the information base of the early warning system for BT in Italy. The system also has accurately established the epidemiology of BTV infection and the distribution and dynamics of BT vectors (11, 27). In addition, monitoring of the spread of BTV infection was facilitated (9, 24), and risk factors linked to the spread of the vectors and to animal movement were evaluated. The surveillance systems implemented in the countries of southern Europe, and particularly in Italy where BTV had spread and persisted more than in any other EU country, produced information that was critical to the development of the flexibility that now characterises the European Union legislation on BT. Decision 2001/783/EC of 11 September 2001 (14), has included two procedures foreseen by the OIE Terrestrial animal health code (31) in the European provisions, namely: Article for animal movements from infected zones Article to define seasonally free areas. Furthermore, the decision has reduced the radius of the zone from which slaughter animals cannot be sent to free zones from 100 km to 20 km, provided that a surveillance system is in place. Subsequently, from 16 January 2002, a series of Decisions (2002/35/EC, 2002/189/EC, 2002/543/EC (15, 16, 17) have excluded some Italian provinces from those surveillance zones in which the surveillance system had documented the absence of virus circulation. From 10 January 2003 (Decision 2003/14/EC) (18), the despatch of slaughter animals from infected to free areas was permitted, provided that the vaccination coverage is at least 80% in the province of origin and a risk assessment has been made. Decision 2003/218/EC (19) of 27 March 2003 introduced the concept of risk in the European provisions and has subdivided the territories in areas of higher and lower epidemiological risk. The decision, therefore, allows the following: the despatch of live animals from the lower risk areas, where viral circulation has not been detected, to the remainder of the European Union the despatch of slaughter animals from lower risk areas even with active infection and from higher risk areas where viral circulation has not been detected to free areas in the national territory. The latter is allowed only if the animals have been vaccinated at least 30 days previously, if they belong to a herd where all the animals have been vaccinated and if their transport occurs during daylight hours only. According to Decision 2003/218/EC, the Member State can demarcate epidemiologically relevant areas of origin ; in other words, on the basis of surveillance results, it can reduce or increase the protection zones to a lower or higher radius than 20 km and it can evaluate the possibility of demarcating lower risk areas in higher risk territories. The data and knowledge obtained in these studies has also facilitated risk assessments to: define the optimal national BT control strategy define the risk arising from movement of slaughter and live animals from restricted zones as related to the presence or absence of viral circulation and to population immunity from vaccination (24) define the minimum level of serological surveillance able to detect ongoing BTV infection with comparable sensitivity to the existing surveillance programme (8). References 1. Animal Health Australia (2002). Animal health in Australia Canberra, 106 pp. 2. Animal and Plant Health Inspection Service (APHIS) (2003). Bluetongue surveillance. The 2000 serological survey of slaughter cattle for antibody against bluetongue virus. Diagnostic Virology Laboratory, National Veterinary Services Laboratories, Ames (aphis.usda.gov/vs/nahps/blue tongue/serological_survey.html accessed on 23 June 2004). 3. Anon. (2000). Bluetongue in France: in the island of Corsica. Dis. Info., 13 (43), Anon. (2000). Bluetongue in Spain: in the Balearic islands. Dis. Info., 13 (40), Boorman J. (1986). Presence of bluetongue virus vectors on Rhodes. Vet. Rec., 118, 21. Veterinaria Italiana, 40 (3),

159 6. Boorman J.P.T. & Wilkinson P.J. (1983). Potential vectors of bluetongue in Lesbos, Greece. Vet. Rec., 113, Boorman J., Jennings M., Mellor P. & Wilkinson P. (1985). Further data on the distribution of midges in southern Europe and the Mediterranean area, with special reference to Culicoides imicola. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, New York, Progr. Clin. Biol. Res., 178, Calistri P., Giovannini A., Conte A. & Caporale V. (2004). Use of a Montecarlo simulation model for the re-planning of bluetongue surveillance in Italy. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Calistri P., Giovannini A., Conte A., Nannini D., Santucci U., Patta C., Rolesu S. & Caporale V. (2004). Bluetongue in Italy: Part I. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Calistri P., Goffredo M., Caporale V. & Meiswinkel R. (2003). The distribution of Culicoides imicola in Italy: application and evaluation of current Mediterranean models based on climate. J. Vet. Med. B, 50, Conte A., Giovannini A., Savini L., Goffredo M., Calistri P. & Meiswinkel R. (2003). The effect of climate on the presence of Culicoides imicola in Italy. J. Vet. Med. B, 50, Dulac G.C., Sterrit W.G., Dubus C., Afshar A., Myers D.J., Taylor E.A., Jamieson B.R. & Martin M.W. (1991). Incursions of orbiviruses in Canada and their serological monitoring in the native animal population between 1962 and In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, European Commission (2001). Decision of 9 February 2001 establishing protection and surveillance zones in the Community in relation with bluetongue (2001/138/EC). Off. J., L 050, 21/02/2001, European Commission (2001). Decision of 9 November 2001 on protection and surveillance zones in relation to bluetongue, and on rules applicable to movements of animals in and from those zones (2001/783/EC). Off. J., L 293, 10/11/2001, European Commission (2002). Decision of 16 January 2002 amending Decision 2001/783/EC as regards the protection and surveillance zones in relation to bluetongue in Italy (2002/35/EC). Off. J., L 015, 17/01/2002, European Commission (2002). Decision of 5 March 2002 amending Decision 2001/783/EC as regards the protection and surveillance zones in relation to bluetongue in Italy (2002/189/EC). Off. J., L 063, 06/03/2002, European Commission (2002). Decision of 4 July 2002 amending Decision 2001/783/EC as regards the protection and surveillance zones in relation to bluetongue in Italy (2002/543/EC). Off. J., L 176, 05/07/2002, European Commission (2003). Decision of 10 January 2003 amending Decision 2001/783/EC as regards the bluetongue protection and surveillance zones and conditions for movements of animals for immediate slaughter (2003/14/EC). Off. J., L 007, 11/01/2003, European Commission (2003). Decision of 27 March 2003 on protection and surveillance zones in relation to bluetongue, and on rules applicable to movements of animals in and from those zones and repealing Decision 2001/783/EC (2003/218/EC). Off. J., L 082, 29/03/2003, European Commission (EC) (2003). Aiuto di Stato No. 662/2001 Italia (Sardegna). Aiuti a favore degli allevatori di bovini per i danni causati dall epizoozia Blue Tongue, C(2003) 2519fin, 27 July. EC, Brussels, 8 pp. (europa.eu.int/comm/secretariat_general/sgb/state_aids /agriculture/n pdf accessed on 28 July 2004). 21. European Council (1992). Directive 92/35/EEC of 29 April 1992 laying down control rules and measures to combat African horse sickness. Off. J., L 157, 10/06/1992, European Council (1992). Directive 92/119/EEC of 17 December 1992 introducing general Community measures for the control of certain animal diseases and specific measures relating to swine vesicular disease. Off. J., L 062, 15/03/1993, European Council (2000). Directive 2000/75/EC of 20 November 2000 laying down specific provisions for the control and eradication of bluetongue. Off. J., L 327, 22/12/2000, Giovannini A., Calistri P., Nannini D., Paladini C., Santucci U. & Caporale V. (2004). Bluetongue in Italy: Part II. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Giovannini A., Conte A., Calistri P., Di Francesco C. & Caporale V. (2004). Risk analysis on the introduction into free territories of vaccinated animals from restricted zones. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Goffredo M., Satta G., Torina A., Federico G., Scaramozzino M., Cafiero M.A., Lelli R. & Meiswinkel R. (2001). The 2000 bluetongue virus (BTV) outbreak in Italy: distribution and abundance of the principal vector Culicoides imicola Kieffer. 196 Veterinaria Italiana, 40 (3), 2004

160 In Proc. 10th International Symposium of the American Association of Veterinary Laboratory Diagnosticians (AAVLD) and OIE Seminar on biotechnology, Salsomaggiore, Parma, 4-7 July. AAVLD, Ames, Goffredo M., Conte A.M., Cocciolito R. & Meiswinkel R. (2003). Distribuzione ed abbondanza di Culicoides imicola in italia. Vet. Ital., 39, Kirkland P.D., Kennedy D.J., Williams C.F., Hornitzky C.L., Gleeson A. & Batty E.M. (1991). Distribution of vector-free areas by monitoring of sentinel cattle for multiple arbovirus infections. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Mellor P.S., Boorman J.P.T., Wilkinson P.J. & Martinez-Gomez F. (1983). Potential vectors of bluetongue and African horse sickness viruses in Spain. Vet. Rec., 112, Mellor P.S., Jennings D.M., Wilkinson P.J. & Boorman J.P.T. (1985). Culicoides imicola: a bluetongue virus vector in Spain and Portugal. Vet. Rec., 116, Office International des Épizooties (OIE) (2003). Terrestrial animal health code, Chapter OIE, Paris (oie.int/eng/normes/mcode/a_00038.htm accessed on 28 July 2004). 32. Pearson J.E., Gustafson G.A., Shafer A.L. & Alstad A.D. (1991). Distribution of bluetongue in the United States In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Scaramozzino P., Boorman J., Semproni G., Vitale F., Mellor P. & Caracappa S. (1996). Entomological survey on Ceratopogonidae in central-southern Italy. In European Multicolloquial on Parasitology. Parassitologia, 38, Ward M.P., Carpenter T.E. & Osburn B.I. (1994). Host factors affecting seroprevalence of bluetongue virus infections of cattle. Am. J. Vet. Res., 55 (7), Veterinaria Italiana, 40 (3),

161 Vet. Ital., 40 (3), Transmission potential of South African Culicoides species for live-attenuated bluetongue virus G.J. Venter (1), G.H. Gerdes (1), P.S. Mellor (2) & J.T. Paweska (3) (1) Agricultural Research Council- Onderstepoort Veterinary Institute, Private Bag X05, Onderstepoort, South Africa (2) Institute for Animal Health, Pirbright Laboratory, Ash Road, Pirbright, Woking, Surrey GU24 ONF, United Kingdom (3) Special Pathogens Unit, National Institute for Communicable Diseases, Sandringham, South Africa Summary Field-collected Culicoides were fed on sheep blood-virus mixtures, each containing one of four liveattenuated vaccine strains of bluetongue virus (BTV), namely: BTV-1, BTV-4, BTV-9, and BTV-16. A South African field isolate of BTV-1 was used as the non-attenuated control virus. Titres of vaccine strains in blood meals ranged from 5.1 to 6.1 log 10 TCID 50 /ml; the titre of the field isolate of BTV-1 was 7.1 log 10 TCID 50 /ml. Recovery rates of vaccine viruses from Culicoides assayed immediately after feeding varied from 0% to 10.6%. This indicates that virus concentrations in blood meals were too low to ensure that all individuals ingested detectable amounts of virus. Thus, the oral susceptibility of Culicoides to infection with BTV vaccine strains determined in this study might be an underestimation. Of a total of Culicoides that survived a 10-day extrinsic incubation period at 23.5 C, 124 tested positive for BTV; 65 individuals yielded vaccine strains, and the remaining 59, the field isolate of BTV-1. Infection prevalences with the vaccine viruses ranged from 11.0% in C. bolitinos fed on blood containing 6.1 log 10 TCID/ml of BTV-1 down to 0.3% in C. imicola fed on a blood containing 5.3 log 10 TCID/ml of BTV-4. The infection rate for C. imicola and C. bolitinos fed on the field isolate of BTV-1 was 9.5% and 36.0%, respectively. In most infected midges the replication levels of vaccine strains were below the postulated threshold for a systemic infection with an orbivirus as previously calculated in the larger American vector, C. sonorensis (>2.5 log 10 TCID 50 /midge) but some individuals replicated BTV vaccine strains to high titres. This carries an implication that if ruminants become viraemic after vaccination with live-attenuated BTV vaccines, they might act as a source for the infection of Culicoides vectors. Keywords Artificial infection Bluetongue virus Culicoides bolitinos Culicoides imicola Infection Oral susceptibility Prevalence South Africa Vaccine. Introduction A potential problem with current live-attenuated bluetongue virus (BTV) vaccines is that they can induce a low level of viraemia in vaccinated animals. Thus, there is the fear that vaccine viruses may reverse to virulence through passage in vectors, and then be transmitted in the field (16). Culicoides sonorensis infected through feeding on vaccinated sheep have been shown to transmit vaccine virus to susceptible animals (3). Another concern is that reassortment between vaccine and wild-type viruses may occur in the field and generate strains of different virulences (12). The fact that Theiler s original BTV strain 4 has been used as a vaccine strain for 80 years without detectable antigenic drift and the absence of any evidence indicating that new serotypes have recently arisen suggests, however, that such events, should they occur at all, are likely to be very rare (15). Culicoides imicola has been implicated as the major field vector of BTV in southern Africa (6, 7, 10) and southern Europe (2, 8). However, evidence is growing that other Culicoides species may also act as competent field vectors (7, 9, 14). In South Africa, it has been found that C. bolitinos supported replication of BTV-1, 198 Veterinaria Italiana, 40 (3), 2004

162 BTV-3 and BTV-4 to higher levels (14) and that it had significantly higher vector competence for BTV-1 over a range of different incubation periods and temperatures than C. imicola (11). In addition, non-avaritia South African Culicoides species such as C. bedfordi, C. leucostictus, C. pycnostictus and C. milnei were shown to be susceptible to BTV infection (11). In addition, at least three European Culicoides species namely: C. nubeculosus, C. pulicaris and C. obsoletus can become infected with BTV after feeding on viraemic sheep (7). Thus, data on vector competence of Culicoides midges for BTV liveattenuated vaccine strains are of importance for risk assessment analyses and particularly when intervention with live-attenuated vaccines is being considered in regions vulnerable to virus incursions. The oral susceptibility and vector competence of South African livestock-associated Culicoides species for vaccine viruses that are included in the polyvalent BTV vaccine in current use has not yet been evaluated. The aim of this study was to determine the oral susceptibility of C. imicola, C. bolitinos and other livestock-associated Culicoides species to cell culture-attenuated BTV vaccine strains that are currently used for the annual vaccination of sheep in southern Africa. Materials and methods Viruses The viruses used and the titres/ml of blood meal are listed in Table I. Stocks of vaccine viruses for oral infection studies were grown in BHK-21 cells, titrated and stored in 10% foetal bovine serum in 1 ml aliquots at 70 C. Aliquots of virus stocks for oral infection were titrated during the experiments using a previously described procedure (14). Insects Culicoides were collected near farm animals at the Onderstepoort Veterinary Institute (25 39 S, E; m above sea level) and on Koeberg Farm near Clarens (28º32 S, E, m above sea level) in the eastern Free State from January to April 2002 as described previously (14). Feeding technique Before feeding, the field-collected Culicoides were held without access to nutrients or water for 24 h at 23.5 C and a relative humidity of 50-70%. Lighting in the room was dimmed to ~1% daylight (~65 lux). Surviving flies were subsequently fed in batches of for min on defibrinated sheep blood Table I Virus recovery and virus titres in field-collected Culicoides fed on blood/bluetongue virus mixtures assayed immediately after feeding, January-April 2002 Culicoides species C. imicola Mean virus titre/midge (a) C. bolitinos Mean virus titre/midge C. magnus Mean virus titre/midge C. huambensis Mean virus titre/midge C. leucostictus Mean virus titre/midge C. zuluensis Mean virus titre/midge C. pycnostictus Mean virus titre/midge Bluetongue virus strain and mean virus titre (log10tcid50/ml) of blood meal Field strain Vaccine strain (virus titre) (virus titre) 1 (7.1) 1 (6.1) 1 (5.1) 4 (6.3) 4 (5.3) 9 (5.3) 16 (6.1) 16 (5.1) 1/7 (b) 1.9 1/ /6 1/ / / / / /6 0/11 0/29 0/17 0/21 0/5 0/2 0/7 1/ /1 0/2 0/8 0/8 0/6 0/4 1/4 1.7 C. bedfordi 0/1 C. coarctatus 0/1 C. onderstepoortensis 0/1 C. gulbenkiani 0/2 a) log10tcid50/midge b) number positive/number tested Veterinaria Italiana, 40 (3),

163 containing one of the vaccine viruses through a oneday-old chicken-skin membrane using techniques described previously (14). A blood-virus mixture was freshly prepared immediately before feeding as described previously (11). The blood-virus mixture was maintained at 35.5 C and stirred constantly during feeding. Immediately after feeding, midges were immobilised by holding them at 4 C until movement ceased. Blood-engorged females were separated out on a chill-table and handled as described previously (14). To improve the survival of the Culicoides after blood feeding, midges were maintained on a 5% (w/v) sucrose solution containing 500 IU penicillin, 500 µg streptomycin and 1.25 µg per ml of sucrose solution (1). All female Culicoides surviving the incubation period were sorted into species on a chill-table and stored individually in 1.5 ml microfuge tubes at 70 C until assayed. Processing of Culicoides and virological assays Flies were assayed for virus immediately after feeding on blood-virus mixtures (the day 0 value) and after 10 days extrinsic incubation at 23.5 C. Processing of samples and virus microtitration assays in BHK-21 cells were performed as described previously (11). The identity of virus isolates was determined by a micro-titre virus-neutralisation procedure (4) using type-specific antisera produced in guinea-pigs. Results The feeding rate in the laboratory for field-collected Culicoides varied between 10% and 70% regardless of site and time of collection. Of a total of Culicoides fed, (66.3%) survived the 10-day incubation period at 23.5 C. Due to high bacterial and fungal infections in the field-collected Culicoides, results could not be obtained from all the Culicoides that survived incubation. The numbers of midges examined, from which results could be obtained, are shown in Tables I and II. None of 57 Culicoides that fed on blood containing less than 6 log 10 TCID 50 /ml tested positive for virus immediately after feeding. Of a total of 141 midges fed on blood containing more than 6 log 10 TCID 50 /ml of virus, 15 (10.6%) were positive immediately after feeding. In positive midges, virus titres ranged between 1.7 and 1.9 log 10 TCID 50 /midge (Table I). Of the 124 Culicoides that assayed positive for virus after incubation, 59 were infected with the field isolate of BTV-1. Although there were significant differences between the different serotypes of the virus used, all of the BTV vaccine strains were recovered from at least four of the 18 Culicoides species assayed, namely: C. bolitinos, C. imicola, C. huambensis and C. magnus (Table II). In infected C. bolitinos and C. imicola, the maximum virus titres of BTV vaccine strains, after incubation, ranged from 0.7 to 4.4 log 10 TCID 50 /midge and of the field isolates of BTV-1, from 3.9 to 4.4 log 10 TCID 50 /midge (Table II). Discussion The level of viraemia that is necessary to infect Old World Culicoides vectors with BTV is not known. Based on the blood-meal size of C. imicola and C. bolitinos (see below) it can be calculated that, theoretically, 5 log 10 infectious doses per ml of blood are needed to expose vectors to approximately one TCID 50 of virus. Consequently, none of the Culicoides which fed on blood/btv mixtures with concentrations lower than 6 log 10 TCID 50 /ml were found infected and only 10.6% of midges, which fed on blood/btv mixtures with concentrations higher than 6 log 10 TCID 50 /ml assayed positive for virus immediately after feeding (Table I). The low recovery rate in this study was not due to the inactivation of BTV during the 3-4 hour feeding period as blood samples assayed before and after each feed showed no drop in virus titres. This low rate of virus detection, in day 0 midges is not surprising as the average blood-meal size of C. imicola and C. bolitinos is between 0.01 µl and 0.06 µl, and in this study only 75 µl of the 200 µl of original midge homogenate was tested (minimum detection level of the infectivity assay = 0.63 TCID 50 /midge). Thus, these results indicate that the virus titre in the infecting blood-meal in relation to the volume of the blood-meal ingested by the midges was too low to allow all engorging midges to take up virus. Prevalence of infection, as determined in this study, could therefore be much lower than the true susceptibility rate of the Culicoides species tested. Despite low virus titres in the blood-meals, this study also shows that the two major BTV vectors in South Africa, C. imicola and C. bolitinos, are susceptible to oral infection with the current BTV vaccine strains. Moreover, BTV has been recovered after the 10 days extrinsic incubation period from non-avaritia livestockassociated species (C. magnus and C. huambensis) (Table II). The possible involvement of non-avaritia species in vectoring orbiviruses in South Africa is supported by the successful isolation of BTV after 10- day incubation period from orally infected C. bedfordi, C. magnus and C. pycnostictus (11). Concentration of live virus in the head or body of infected midges, higher than 2.5 log 10 TCID 50 /insect has been postulated as an indicator of fully disseminated infection (5). Since this threshold titre distinguishes between potentially transmissive and non-transmissive 200 Veterinaria Italiana, 40 (3), 2004

164 Table II Virus recovery and virus titres in field-collected Culicoides fed on blood/bluetongue virus mixtures assayed after 10 days extrinsic incubation at 23.5ºC, January-April 2002 Culicoides species C. imicola infection prevalence (%) Max. virus titre (a) Mean virus titre/ midge (a) (STD) C. bolitinos infection prevalence (%) Max. virus titre Mean virus titre/ midge (STD) C. magnus infection prevalence (%) Max. virus titre Mean virus titre/ midge (STD) C. huambensis infection prevalence (%) Max. virus titre Mean virus titre/ midge Bluetongue virus strain and mean virus titre (log10tcid50/ml) of blood meal Field strain Vaccine strain (virus titre) (virus titre) 1 (7.1) 1 (6.1) 1 (5.1) 4 (6.3) 4 (5.3) 9 (5.3) 16 (6.1) 16 (5.1) 27/285 (b) (9.5) (1.04) 32/89 (36.0) (0.88) 22/712 (3.1) (0.37) 13/118 (11.0) (0.69) 2/30 (6.7) (0.42) 1/4 (25.0) /476 10/886 (1.1) 1.3 1/5 (20.0) (0.26) 1/715 (0.1) 0.7 0/169 0/13 4/166 (2.4) 1.7 0/349 3/728 (0.4) (0.39) 1.7 7/150 (4.7) (1.04) 0/1 0/26 0/7 0/48 0/9 0/1 0/1 1/371 (0.3) 0.7 C. bedfordi 0/4 0/13 0/22 0/4 C. coarctatus 0/1 0/3 0/10 0/11 0/5 C. enderleini 0/3 0/1 0/7 0/5 C. engubandei 0/4 0/5 0/15 0/8 C. gulbenkiani 0/2 0/17 0/42 0/15 0/15 C. leucostictus 0/1 0/29 0/1 0/59 0/121 0/31 C. milnei 0/2 0/6 0/4 C. neavei 0/1 C nevilli 0/1 0/2 0/1 C. nivosus 0/1 0/1 0/1 0/1 C. onderstepoortensis 0/1 0/3 0/1 C. pycnostictus 0/3 0/4 0/1 0/27 0/32 0/8 0/1 C. sp35 0/1 C. zuluensis 0/9 0/23 0/1 0/7 0/4 0/34 Total (%) 59/393 (15.0) 38/954 (4.0) 1/485 (0.2) 10/1264 (0.8) 1/728 (0.1) 4/758 (0.5) 10/1039 (1.0) 1/383 (0.3) a) log10tcid50/midge STD standard deviation in virus titre b) number positive/number tested 0.7 0/10 individuals, the epidemiological significance of laboratory oral infection studies can be assessed. At present however, it is unknown if this precise threshold value, derived from results with the much larger C. sonorensis-african horse sickness virus (AHSV)/BTV model applies directly to other, differently sized Culicoides species, without extensive experimental validation. In contrast to the highly standardised AHSV/BTV model, vector competence studies on field-collected Culicoides cannot be fully controlled. Culicoides are a biologically highly diverse genus and thus extrapolation of vector competence data from one species to others is not advisable. C. sonorensis is at least four times bigger than either C. imicola or C. bolitinos. In the absence of experimental data this suggests that the infectivity threshold for transmission potential in smaller Culicoides species may be significantly lower than in C. sonorensis. As has been found previously (11, 14), C. bolitinos was shown to be more susceptible to infection with the field strain of BTV-1 than C. imicola (Table II). Furthermore, the prevalence of infection in both species, as well as the maximum virus titre in midges Veterinaria Italiana, 40 (3),

165 assayed after 10 days incubation, was 2 to 3 times higher in midges fed on the field strain of BTV than in those fed on the vaccine strains (Table II). However, the virus titres of the blood-meals containing the field strain of BTV were 10 times higher than those containing the vaccine strains (Table II) so direct comparisons on susceptibility to field and vaccine strains of BTV are not possible at present. In view of the very high abundance of the two highly competent Orbivirus vectors, C. imicola and C. bolitinos, in South Africa (13), the level of viraemia in vaccinated animals, transmission of BTV vaccine strains from vaccinated to unvaccinated animals may well occur. However, such an event has not yet been demonstrated in South Africa and neither has vaccine virus reversion to virulence been demonstrated on vector insect passage so the epidemiological significance of these results remains uncertain. Further studies, e.g. including sequencing analysis of vaccine strains and field isolates of BTV and reversion studies on insect-passaged vaccine virus, are urgently required. It is also important to develop experimental data to estimate the threshold level of viraemia in the vertebrate host above which C. imicola and C. bolitinos become orally infected. Acknowledgements Grateful thanks are extended to Pamela Hunter for supplying vaccine strains of BTV, Ina Hermanides and Sandra Prinsloo for technical assistance and the OIE Reference Laboratory for Bluetongue at ARC- OVI for supporting this work. Ralph Burls is also sincerely thanked for making his farm available as a collecting site. This study was undertaken as part of a European Union funded project (Contract Number: QLK ). References 1. Bellis G.A., Gibson D.S., Polkinghorne I.G., Johnson S.J. & Flanagan M. (1994). Infection of Culicoides brevitarsis (Diptera: Ceratopogonidae) with four Australian serotypes of bluetongue virus. J. Med. Entomol., 31, Boorman J. (1986). Presence of bluetongue virus vectors on Rhodes. Vet. Rec., 118, Foster N.M., Jones R.H. & Luedke A.J. (1968). Transmission of attenuated and virulent bluetongue virus with Culicoides variipennis orally via sheep. Am. J. Vet. Res., 29 (2), House C., Mikiciuk P.E. & Berninger M.L. (1990). Laboratory diagnosis of African horse sickness: comparison of serological techniques and evaluation of storage methods of samples for virus isolation. J.Vet. Diagn. Invest., 2, Jennings D.M. & Mellor P.S. (1987). Variation in the responses of Culicoides variipennis (Diptera: Ceratopogonidae) to oral infection with bluetongue virus. Arch. Virol., 95, Meiswinkel R., Nevill E.M. & Venter G.J. (1994). Vectors: Culicoides spp. In Infectious diseases of livestock with special reference to southern Africa, Vol. I (J.A.W. Coetzer, G.R. Thomson & R.C. Tustin, eds). Oxford University Press, Cape Town, Mellor P.S. (1992). Culicoides as potential orbivirus vectors in Europe. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Mellor P.S., Jennings D.M., Wilkinson P.J. & Boorman J.P.T. (1985). Culicoides imicola: a bluetongue virus vector in Spain and Portugal. Vet. Rec., 116, Mellor P.S., Boorman J. & Baylis M. (2000). Culicoides biting midges: their role as arbovirus vectors. Ann. Rev. Entomol., 45, Nevill E.M., Erasmus B.J. & Venter G.J. (1992). A sixyear study of viruses associated with Culicoides biting midges throughout South Africa (Diptera: Ceratopogonidae). In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Paweska J.T., Venter G.J. & Mellor P.S. (2002). Vector competence of South African Culicoides species for bluetongue virus serotype 1 (BTV-1) with special reference to the effect of temperature on the rate of virus replication in C. imicola and C. bolitinos. Med. Vet. Entomol., 16, Samal S.K., El Hussein A., Holbrook F.R., Beatty B.J. & Ramig R.F. (1987). Mixed infections of Culicoides variipennis with bluetongue virus serotypes 10 and 17: evidence for high frequency reassortment in the vector. J. Gen. Virol., 68, Venter G.J., Nevill E.M. & Van der Linde T.C. de K. (1996). Geographical distribution and relative abundance of stock-associated Culicoides species (Diptera: Ceratopogonidae) in southern Africa, in relation to their potential as viral vectors. Onderstepoort J. Vet. Res., 63, Venter G.J., Paweska J.T., Van Dijk A.A., Mellor P.S. & Tabachnick W.J. (1998). Vector competence of Culicoides bolitinos and C. imicola (Diptera: Ceratopogonidae) for South African bluetongue virus serotypes 1, 3 and 4. Med. Vet. Entomol., 12, Verwoerd D.W. & Erasmus B.J. (1994). Bluetongue. In Infectious diseases of livestock with special reference to southern Africa, Vol. I (J.A.W. Coetzer, G.R. Thomson & R.C. Tustin, eds). Oxford University Press, Cape Town, Walton T.E. (1992). Attenuated and inactivated orbiviral vaccines. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Veterinaria Italiana, 40 (3), 2004

166 Vet. Ital., 40 (3), Epidemiology and vectors Epidemiology of bluetongue and epizootic haemorrhagic disease in wildlife: surveillance methods D.E. Stallknecht (1, 2) (1, 3) & E.W. Howerth (1) Southeastern Cooperative Wildlife Disease Study, College of Veterinary Medicine, The University of Georgia, Athens, GA 30605, United States of America (2) Department of Medical Microbiology and Parasitology, College of Veterinary Medicine, The University of Georgia, Athens, GA 30605, United States of America (3) Department of Parasitology, College of Veterinary Medicine, The University of Georgia, Athens, GA 30605, United States of America Summary Bluetongue (BT) and epizootic haemorrhagic disease (EHD) are the most important viral diseases that affect wild ungulates, especially white-tailed deer, in the United States of America (USA). For this reason, considerable surveillance has been conducted. Surveillance has relied upon standard serological and virus detection methods, and both passive and active surveillance strategies have been employed effectively. These efforts have led to an improved understanding of the epidemiology of these diseases in wild ungulate populations, specifically the recognition and understanding of geographically predictable disease patterns ranging from enzootic stability to sporadic epizootics. The utilisation of wildlife in BT and EHD surveillance may be unique to the USA where these diseases are important to both wildlife and livestock interests. Keywords Bluetongue Epidemiology Epizootic haemorrhagic disease White-tailed deer Wildlife United States of America. The bluetongue (BT) viruses (BTV) and epizootic haemorrhagic disease (EHD) viruses (EHDV) can infect ruminant species of wildlife as well as domestic animals. Clinical disease in wildlife, however, is common only in North America where mortality and morbidity have been documented in white-tailed deer (Odocoileus virginianus), mule deer (O.hemionus), pronghorn (Antilocapra americana), elk (Cervus elaphus), mountain goat (Oreamnos americanus) and bighorn sheep (Ovis canadensis) (14). Of all of these species, white-tailed deer have been the most affected, and mortality has been associated with all of the serotypes of BTV and EHDV that have been isolated in North America (BTV-10, BTV-11, BTV-13 and BTV-17 and EHDV-1 and EHDV-2) except BTV-2 (1, 13, 21, 26). Both BT and EHD are clinically indistinguishable in these wildlife species and BT and EHD in wild ungulates are often collectively referred to as haemorrhagic disease (HD). The first confirmed case of HD was documented in white-tailed deer in New Jersey in 1955 with the isolation of EHDV-1 (21). Mortality events consistent with HD have been reported as early as 1901 and, since 1955, have occurred consistently within the United States of America (18). Initial interest in understanding the epidemiology of HD in North American wildlife started with the observation of extensive mortality in white-tailed deer in the USA during the 1950s and 1960s (17). At that time, significant efforts were being made by state conservation agencies to rebuild remnant white-tailed deer populations and re-establish them in areas of their former range. The detection of HDrelated mortality in these growing populations led to immediate concerns that BT, EHD, or both, would have an impact on these conservation efforts. At present, there are no indications that HD will eliminate, regulate, or limit a wild ungulate population, but severe population reductions capable of affecting short-term management goals can and have occurred (25). HD currently is considered as the most important viral disease affecting white- Veterinaria Italiana, 40 (3),

167 tailed deer and for this reason significant surveillance has been directed at wildlife populations and continues to take place. Surveillance components and methods Surveillance methods directed at wildlife in the USA have relied on standard serological and virus detection methods, and both passive and active surveillance strategies have been employed effectively. Surveillance strategies successfully used to date include the following: virus isolation and polymerase chain reaction (PCR)-based diagnostic support for clinical submissions questionnaire-based surveillance cross-sectional serologically based studies of selected wild ungulate populations outbreak investigations. White-tailed deer and other wild ungulates routinely enter animal disease diagnostic channels within the USA. In addition, there are increasing numbers of wildlife disease oriented laboratories with diagnostic capabilities. The isolation of BTV or EHDV, especially from clinical submissions from domestic animals, has relied on a combination of egg inoculation and tissue culture (19). Although these same techniques have application to wild ungulate samples, the virus isolation protocols used by the authors rely exclusively on tissue culture, specifically the inoculation of CPAE (cattle pulmonary artery endothelial) and BHK-21 (baby hamster kidney) cells. Of the two cell lines, CPAE cells are most sensitive. This system works in the absence of egg inoculation for several reasons. First, the reported difference in sensitivity of embryonated chicken eggs and CPAE cells for BTV is less than 1 log of virus (27). Secondly, in experimentally infected deer viraemia has been detected with BTV-10 and EHDV-2 by virus isolation in these cell lines for up to 12 and 56 days, respectively (20). This provides an extended window of opportunity to isolate these viruses. Thirdly, because case submissions consist almost entirely of white-tailed deer dying of acute BT or EHD, viral titres in blood and tissue are close to their peak. In experimentally infected deer, peak titres exceeding 10 4 and 10 6 TCID 50 /ml of blood have been observed for BTV-10 and the EHDV (both EHDV-1 and EHDV-2), respectively (9, 20). Finally, these viruses are very durable and viral titres are not greatly reduced due to minor delays in clinical submissions. PCR protocols are readily available for both BTV and EHDV and their use in diagnostics is increasing. Although quicker and potentially more sensitive than virus isolation, results are usually limited to identification of viruses at the serogroup level. The primary disadvantages associated with clinical submissions relate to the need for detected mortality or morbidity. Because infections with both BTV and EHDV in wild ungulates often do not result in clinical disease, infection rates cannot be estimated from such data. Wildlife diseases also have an inherent problem with case detection resulting in under-reporting. In a recent outbreak of HD in deer in Missouri that was detected in radio-monitored animals, a mortality rate of 8% was estimated (2). However, not a single report of deer mortality or morbidity was received from public sources during this period. Questionnaire-based surveillance for HD in wild ungulate populations has been used effectively in the USA since 1981 (16). Information relating to HD in free-living wildlife is requested annually from State wildlife management agencies by the Southeastern Cooperative Wildlife Disease Study. This survey is based on four criteria as follows: 1) sudden and unexplained deer mortality that occurs during late summer and early autumn 2) necropsy-based diagnosis of HD based on clinical signs 3) isolation of EHDV or BTV 4) detection of deer with sloughing hooves. This system has been used effectively to map the distribution of HD in the USA. Although most data (criteria 1, 2 and 3) relate to detected mortality, criteria 4 provides an estimate of morbidity, which is applicable to geographic areas where mortality seldom occurs. The major advantages of this system relate to simplicity, continuity and a national scope. The major disadvantages relate to reporting bias and a lack of confirmatory diagnostics associated with some of the criteria. This problem is improving, however, as diagnostic submissions and laboratory confirmation have become more available. Cross-sectional serologically based studies have primarily relied on agar gel immunodiffusion (AGID)-based serology and serum neutralisation (22, 23). Competitive enzyme-linked immunosorbent assay (c-elisa)-based serological tests also have potential application to such studies especially with BTV. The primary advantage of serologically based surveillance lies in the capacity to detect evidence of previous infection. This is especially important for areas where infections are subclinical or result in mild disease. An additional advantage to serologically based surveillance includes the ability to use hunterkilled animals, greatly reducing cost and time associated with sample collection. The primary disadvantage relates to generating reliable prevalence 204 Veterinaria Italiana, 40 (3), 2004

168 rates for specific viruses, as problems with specificity are common with both the AGID (between serogroups) and serum neutralisation (between serotypes) (19). Although not a problem in the USA, increased serotype diversity or incomplete knowledge of serotype diversity would also make serum neutralisation cumbersome and possibly inaccurate. Outbreak investigations are not performed on a routine basis during HD outbreaks in the USA, but can be extremely valuable in attempts to understand impacts on populations. Difficulties associated with outbreak investigations are reflected in the paucity of information available in the scientific literature. To date, there have been few studies that have attempted to measure or even estimate population impacts in free-living white-tailed deer populations (2, 5, 6, 7, 15, 21). The restraints of such work relate to reliability and availability of both pre- and postoutbreak epidemiological and population data. Wildlife-based surveillance: understanding epidemiology Work related to BT and EHD in wildlife has not only led to a better understanding of potential disease impacts but also has served to better define the epidemiology of these diseases within the USA. Based on long-term questionnaire-based data, it is clear that patterns of clinical disease in these wild ungulate populations (especially white-tailed deer) are spatially and temporally predictable (25). The distribution of HD extends throughout the southeastern USA, extending as far north as New Jersey and as far west as eastern Texas. From this area, the range extends in a north-westerly direction through the central USA to eastern Montana. Reports also are relatively common from California, Oregon and Washington, primarily from black-tailed deer (O.hemionus). There are few reports from states in the north-east and south-west and from those states bordering the Great Lakes. Except for the southwestern USA, this distribution of HD follows the known distribution of Culicoides sonorensis (12); this explains the lack of disease in the north-eastern USA. Clinical disease with HD is extremely variable ranging from death to subclinical infection. Expected clinical patterns, however, can be predicted based on the occurrence of endemic or epidemic disease patterns (4, 17). Endemic areas include the coastal plains of the south-east, and in this area, most reported cases of HD represent the chronic form of the disease. These chronic cases are characterised by hoof and rumen lesions; the disease may affect condition but most infected animals survive. In contrast, in certain areas of the central USA, and in the Piedmont and Appalachian Mountain physiographical regions of the south-east, a pattern of epidemic HD occurs where high levels of mortality are common. A third pattern exists in Texas and possibly other areas of the south-west and mid-west. In these areas, infections do not result in clinical disease. In Texas, for example, there are very few reported cases of HD even though infection rates, as determined by the presence of antibodies to these viruses, approach 100% (24). Based on few clinical reports of HD (16) and a high antibody prevalence for both the EHDV and BTV (22), a similar situation of enzootic stability may occur in southern Florida. This clinical variation relates to variation in herd immunity, specifically the combined effects of maternal antibody transfer (10), acquired immunity through previous challenge (9, 11, 20), and innate resistance within specific host populations (8). Currently, there is no evidence to suggest that this observed regional variation is related to variation in EHDV or BTV virulence, either associated with individual EHDV or BTV serotypes or between strains within these serotypes. Virulent strains of EHDV and BTV do occur in areas of enzootic stability as indicated by the high mortality rates observed in naive penned deer that have been moved to these areas. In fact, all of the virus isolations collected for this study from clinically infected deer in Texas have been associated with mortality in penned deer that have been moved into this state or have originated from such animals. In wild ungulates, HD is seasonal, occurring from mid-summer through to late autumn, and usually peaks in September (3). From 1990 to 2002, over 220 isolations of EHDV and BTV were made from deer throughout the south-east and mid-west and all have come from clinical submissions within this same seasonal period. Seasonal distribution is most likely related to seasonal patterns in vector abundance. Annual variation is more difficult to understand. In endemic areas, HD appears to occur in a two- to three-year cycle (3). In epidemic areas, disease occurs in a longer eight- to ten-year cycle (3, 17). These cycles cannot be explained at this time but probably relate to combined effects of herd immunity and natural or weather-induced fluctuations in vector populations. This is further complicated by the possibility that these short- and long-term cycles may occur concurrently. Although surveillance directed at wildlife has provided much insight into the epidemiology of these diseases, it is important to emphasise that these findings would not be possible with any one Veterinaria Italiana, 40 (3),

169 surveillance system or without supportive experimental work. Application of wildlife surveillance to other geographic areas Wildlife-based surveillance for BT and EHD is effective in the USA for three reasons. First, the presence of disease and mortality has resulted in interest from wildlife management organisations resulting in both funding and co-operation in obtaining research and diagnostic samples and mortality and morbidity reports on a national level. Secondly, the availability of samples from hunterkilled deer provides for very cost and time-efficient sampling of these populations. Finally, the broad distribution and abundance of these species, especially white-tailed deer, provides national and regional coverage for such surveillance activities. Wildlife-based surveillance of BT and EHD has led to a better understanding of the epidemiology of these diseases, especially EHD, which has little significance to livestock production. This situation, however, may be unique to the USA where these diseases have relevance to both wildlife and livestock interests. References 1. Barber T.L. & Jochim M.M. (1975). Serotyping bluetongue and epizootic hemorrhagic disease virus strains. Proc. Am. Assoc. Lab. Diag., 18, Beringer J., Hansen L.P. & Stallknecht D.E. (2000). An epizootic of hemorrhagic disease in white-tailed deer in Missouri. J. Wildl. Dis., 36, Couvillion C.E., Nettles V.F., Davidson W.R., Pearson J.E. & Gustafson G.A. (1981). Hemorrhagic disease among white-tailed deer in the Southeast from 1971 through Proc. US Anim. Hlth Assoc., 85, Davidson W.R. & Doster G.L. (1997). Health characteristics and white-tailed deer population density in the southeastern United States. In The science of overabundance: deer ecology and population management (W.J. NcShea, H.B. Underwood & J.H. Rappole, eds). Smithsonian Institution Press, Washington DC, Fay L.D., Boyce A.P. & Youatt W.G. (1956). An epizootic in deer in Michigan. Trans. North Am. Nat. Res. Conf., 21, Fischer J.R., Hanson L.P., Turk J.R., Miller M.A., Fales W.H. & Gosser H.S. (1995). An epizootic of hemorrhagic disease in white-tailed deer (Odocoileus virginianus) in Missouri: necropsy findings and population impact. J. Wildl. Dis., 31, Gaydos J.K. (2001). Evaluation of white-tailed deer host resistance factors to epizootic hemorrhagic disease viruses. PhD thesis. Department of Medical Microbiology and Parasitology, University of Georgia, Athens, 146 pp. 8. Gaydos J.K., Davidson W.R., Elvinger F., Mead D.G., Howerth E.W. & Stallknecht D.E. (2002). Innate resistance to epizootic hemorrhagic diseases in white-tailed deer. J. Wildl. Dis., 38, Gaydos J.K., Davidson W.R., Howerth E.W., Murphy M., Elvinger F. & Stallknecht D.E. (2002). Cross-protection between epizootic hemorrhagic disease virus serotypes 1 and 2 in white-tailed deer. J. Wildl. Dis., 38, Gaydos J.K., Stallknecht D.E., Kavanaugh D., Olson R.J. & Fuchs E.R. (2002). The dynamics of maternal antibodies to hemorrhagic disease viruses (Reoviridae: Orbivirus) in white-tailed deer. J. Wildl. Dis., 38, Hoff G.L. & Trainer D.O. (1974). Observations on bluetongue and epizootic hemorrhagic diseases in white-tailed deer: (1) distribution of virus in blood (2) cross challenge. J. Wildl. Dis., 10, Holbrook F.R. (1996). Biting midges and the agents they transmit. In The biology of disease vectors (B.J.Beaty & W.C. Marquardt, eds). University Press of Colorado, Niwot, Howerth E.W., Greene C.E. & Prestwood A.K. (1988). Experimentally induced bluetongue virus infection in white-tailed deer. Coagulation, clinical pathologic, and gross pathologic changes. Am. J. Vet. Res., 49, Howerth E.W., Stallknecht D.E. & Kirkland P.D. (2001). Bluetongue, epizootic haemorrhagic disease, and other orbivirus-related diseases. In Infectious diseases of wild mammals (E.S. Williams & I.K. Barker eds). Iowa State University Press, Ames, Karstad L., Winter A. & Trainer D.O. (1961). Pathology of epizootic hemorrhagic disease of deer. Am. J. Vet. Res., 22, Nettles V.F., Davidson W.R. & Stallknecht D.E. (1992). Surveillance for hemorrhagic disease in white-tailed deer and other wild ruminants. Proc. Southeast. Assoc. Fish Wildl. Agencies, 46, Nettles V.F., Hylton S.A., Stallknecht D.E. & Davidson W.R. (1992). Epidemiology of epizootic hemorrhagic disease viruses in wildlife in the USA. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Press, Nettles V.F. & Stallknecht D.E. (1992). History and progress in the study of hemorrhagic disease of deer. Proc. North Am. Wildl. Nat. Res. Conf., 57, Pearson J.E., Gustafson G.A., Shafer A.L. & Alstad A.D. (1991). Diagnosis of bluetongue and epizootic hemorrhagic disease. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Veterinaria Italiana, 40 (3), 2004

170 20. Quist C.F., Howerth E.W., Stallknecht D.E., Brown J., Pisell T. & Nettles V.F. (1997). Host defense responses associated with experimental hemorrhagic disease in white-tailed deer. J. Wild. Dis., 33, Shope R.E., MacNamara L.G. & Mangold R. (1960). A virus-induced epizootic hemorrhagic disease of Virginia white-tailed deer (Odocoileus virginianus). J. Exp. Med., 111, Stallknecht D.E., Blue J.L., Rollor E.A, Nettles V.F., Davidson W.R & Pearson J.E. (1991). Precipitating antibodies to epizootic hemorrhagic disease and bluetongue viruses in white-tailed deer in the southeastern United States. J. Wildl. Dis., 27, Stallknecht D.E., Nettles V.F., Rollor E.A. & Howerth E.W. (1995). Epizootic hemorrhagic disease virus and bluetongue virus serotype distribution in white-tailed deer in Georgia. J. Wildl. Dis., 31, Stallknecht D.E., Luttrell M.P., Smith K.E. & Nettles V.F. (1996). Hemorrhagic disease in whitetailed deer in Texas: a case for enzootic stability. J. Wildl. Dis., 32, Stallknecht D.E., Howerth E.W. & Gaydos J.K. (2002). Hemorrhagic disease in white-tailed deer: our current understanding of risk. Trans. North Am. Wildl. Nat. Res. Conf., 67, Thomas F.C., Willis N. & Ruckerbauer G. (1974). Identification of viruses involved in the 1971 outbreak of hemorrhagic disease in southeastern United States white-tailed deer. J. Wild. Dis., 10, Wechsler S.J. & McHolland L.E. (1988). Susceptibilities of 14 cell lines to bluetongue virus infection. J.Clin. Microbiol., 26, Veterinaria Italiana, 40 (3),

171 Vet. Ital., 40 (3), 208 The role of the Food and Agriculture Organization in bluetongue disease P. Roeder EMPRES Animal Health Service Animal Production and Health Division FAO, viale delle Terme di Caracalla, 00100, Rome, Italy Summary Recent years have witnessed a dramatic northwards extension of the bluetongue virus (BTV)- infected and disease-affected area within the Mediterranean Basin and the Balkans. Serious losses have been experienced for the first time in some areas never before known to have been infected. Associated with this are also changes in vector distribution and possible changes in vector competence. Similarly, in East Asia changes are occurring in virus and vector distribution which might be matched by northerly extension of BTV infection in Central Asia and southerly extension in South America. These events are neither fully defined nor are the determinants of their occurrence understood. Whether or not the events witnessed are due to climate change associated with global warming needs to be determined. The United Nations Food and Agriculture Organization (FAO) is concerned to ensure that the evolution of BTVs is monitored on a global scale to ensure that all countries, especially developing countries and those in transition, can be prepared in advance for future problems, if necessary. Concern over BTV does not stop there for several other vector-borne viruses and their associated diseases exist which could threaten livestock agriculture in vulnerable areas for which bluetongue could serve as a useful model. The FAO has a responsibility to assist member countries to establish epidemiological systems and specific investigational expertise and capacity to fill what could otherwise be lacunae in monitoring networks. Tactical and strategic responses to virus spread also merit united international action by the FAO, OIE and partners in animal disease control. Keywords Bluetongue Control International organisation. 208 Veterinaria Italiana, 40 (3), 2004

172 Vet. Ital., 40 (3), Epidemiology and vectors Culicoides (Diptera: Ceratopogonidae) in Albania: results of the 2002 entomological survey for bluetongue M. Goffredo (1), J.-C. Delécolle (2), G. Semproni (1) & A. Lika (3) (1) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy (2) Musée Zoologique de l Université Louis Pasteur et de la Ville de Strasbourg (MZS), 29 bd de la Victoire, Strasbourg, France (3) Institute of Veterinary Research Bilal Golemi, Department of Virology, Tirana, Albania Summary A survey for Culicoides Latreille, 1809, was made in Albania in 2002 to establish whether Culicoides imicola Kieffer, 1913, the main vector of bluetongue virus in the Mediterranean Basin, or any other suspected vector species, was present. The collections and analyses were performed in accordance with the protocols of the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche) in Teramo, Italy. A total of 43 catches were made in October and November in 15 districts (Bulqise, Devoll, Dibre, Durres, Fier, Gjirokaster, Has, Kolonje, Korce, Librazhd, Permet, Pogradet, Shkoder, Tirane and Tropoje). Twenty species of Culicoides were identified in the collections; the most abundant species belonged to the Obsoletus Complex (98% of total Culicoides in some catches). Culicoides imicola was never captured during the survey. However, a larger number of Culicoides collections and collection sites are needed to exclude the presence of this species at low abundance levels. Keywords Albania Bluetongue Culicoides Culicoides obsoletus Entomological surveillance Obsoletus Complex. Introduction The circulation of bluetongue (BT) virus (BTV) was confirmed serologically in Albania in 2002 (1) during a joint project implemented jointly by the Albanian Veterinary Services, the Veterinary Research Institute of Tirana and the Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Teramo, Italy. In the same season, a survey for Culicoides Latreille, 1809, was also conducted to establish whether Culicoides imicola Kieffer, 1913, the main vector of BTV in the Mediterranean Basin, or any other suspected vector species, was present. This paper presents the results of this first entomological survey conducted between October and November Materials and methods A total of 43 catches were made in October and November in 15 districts of Albania (Bulqise, Devoll, Dibre, Durres, Fier, Gjirokaster, Has, Kolonje, Korce, Librazhd, Permet, Pogradet, Shkoder, Tirane and Tropoje) (Table I). The collections and the analyses were performed in accordance with the protocols of the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche in Teramo, Italy (3). Results The distribution and abundance map of Culicoides spp. is presented in Figure 1. With the single exception of Kolonje, the number of midges per catch was always less than 400 (Table I). Culicoides imicola was never captured during the survey. The species of the Obsoletus Complex were the most abundant (Fig. 2) and at least two species of the complex, C. obsoletus and C. scoticus, were present. In the Kolonje district where three catches were performed, the number of midges per catch ranged between and and 98% of the midges belonged to the Obsoletus Complex. Veterinaria Italiana, 40 (3),

173 Table I Culicoides collections made in Albania, October- November 2002 District No. of catches Maximum number of Culicoides per catch Bulgise Devoll 3 2 Dibre 3 2 Durres 3 20 Fier Gjirokaster 3 91 Has 1 0 Kolonje Korce 3 2 Librazhd Permet Pogradet 3 2 Shkoder 3 45 Tirane 3 29 Tropoje Figure 1 Abundance of Culicoides spp. in Albania October-November 2002 Twenty species of Culicoides were identified in collections, as follows: C. alazanicus Dzhafarov, 1961 C. cataneii Clastrier, 1957 C. circumscriptus Kieffer, 1918 C. festivipennis Kieffer, 1914 C. gejgelensis Dzhafarov, 1964 C. kibunensis Tokunaga, 1937 C. maritimus Kieffer, 1924 C. newsteadi Austen, 1921 C. nubeculosus Meigen, 1830, C. obsoletus Meigen, 1818 C. odiatus Austen, 1921 C. pulicaris Linnaeus, 1758 C. punctatus Meigen, 1804 C. puncticollis Becker, 1903 C. riethi Kieffer, 1914 C. saevus Kieffer, 1922 C. scoticus Downes and Kettle, 1952 C. sejfadinei Dzhafarov, 1958 C. submaritimus Dzhafarov, 1962 C. univittatus Vimmer, Saut de colonne Log midges (max./catch Districts Total Culicoides Obsoletus Complex Figure 2 Abundance of total Culicoides and Obsoletus Complex in Albania, October-November 2002 Discussion Bulgise Devoll Dibre Durres Fier Gjirokaster Has Kolonje Korce Librazhd Permet Pogradet Shkoder Tirane Tropoje In areas where the abundance level of C. imicola is low (less than 100 per trap), high trapping pressure is necessary to reveal its presence (2). Consequently, a larger number of Culicoides collections and collection sites are required to exclude the presence of C. imicola at low abundance levels in Albania. However, the evidence of BTV circulation in Albania (1) and the absence of the main vector C. imicola, suggest that other Culicoides species could be implicated in virus transmission. The high 210 Veterinaria Italiana, 40 (3), 2004

174 abundance level of the Obsoletus Complex, from which BTV was recently isolated in outbreaks where no specimens of C. imicola were captured (4), also suggests that the vectors in Albania probably belong to this species complex. References 1. Di Ventura M., Tittarelli M., Semproni G., Bonfini B., Savini G., Conte A. & Lika A. (2004). Serological surveillance of bluetongue virus in cattle, sheep and goats in Albania. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Goffredo M., Conte A.M., Cocciolito R. & Meiswinkel R. (2003). The distribution and abundance of Culicoides imicola in Italy. Vet. Ital., 39 (47), Goffredo M. & Meiswinkel R. (2004). Entomological surveillance of bluetongue in Italy: methods of capture, catch analysis and identification of Culicoides biting midges. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Savini G., Goffredo M., Monaco F., de Santis P. & Meiswinkel R. (2003). Transmission of bluetongue virus in Italy. Vet. Rec., 152 (4), 119. Veterinaria Italiana, 40 (3),

175 Vet. Ital., 40 (3), Epizootiological control of bluetongue disease in Bulgaria in 2002 G. Georgiev (1), N. Nedelchev (1), Y. Ivanov (2) & E. Veleva (1) (1) National Diagnostic and Research Veterinary Medical Institute, P. Slabeikov Blvd 15 A, Sofia 1606, Bulgaria (2) National Veterinary Service, P. Slabeikov Blvd 15 A, Sofia 1606, Bulgaria Summary In accordance with the National Surveillance Programme for the control of bluetongue (BT) disease in 2002, serum surveillance was performed in 22 sentinel, seronegative animal herds located in western Bulgaria. These herds were at least 40 km outside the settlements affected by the 2001 epidemic. Another 42 sentinel villages (herds) were established in southern Bulgaria in a 10 km border strip zone in the Bourgas, Yambol, Haskovo, Kardjali, Smolyan and Blagoevgrad Districts. The implementation of the programme in 2002 commenced on 15 April and continued until 15 November. More than serum samples were tested prior to 26 August and no evidence of active BT virus (BTV) infection detected. This was confirmed by further viral, serological, epidemiological and clinico-pathological observations. In addition, there was no evidence of transborder penetration of BTV into Bulgaria by infected livestock or by infected Culicoides. However, on 26 August 2002, BTV seropositive sentinel animals were detected close to the southern Bulgarian border. Subsequently, animals were detected in more than 20 villages, but clinical disease was not observed. Bulgaria was divided into 58 quadrants (50 km 50 km) and a Culicoides surveillance programme established in 23 of these. A total of 92 Culicoides light-trap collections were made. During three years of Culicoides surveillance, not a single specimen of the principal BT vector C. imicola was captured. The dominant Culicoides species was C. obsoletus, followed by C. pulicaris and C. punctatus; in August 2001, C. puncticollis was recorded for the first time. Studies on the seasonal phenology of Culicoides were conducted in two villages (Vacsevo and Bersin in the District of Kiustendil) affected in the 2001 outbreak of BT. Here trapping of Culicoides commenced on 1 March 2002 and continued until 15 November 2002; midges became active during the third week of April to almost cease in the second half of November. There appeared to be three peaks of activity: one during the second half of May, another in August and a third at the beginning of October. Keywords Bluetongue Bulgaria Culicoides Sentinel animals Surveillance. Introduction Bluetongue (BT) virus (BTV) the aetiological agent of BT disease in ruminants, belongs to the genus Orbivirus of the family Reoviridae (5). BTV infects and replicates in the Culicoides vector and in domestic and wild ruminants, including cattle, buffalo, deer, antelope, sheep, goat and elk (6). The clinical and pathological symptoms vary among species and range from subclinical or mild to acute and often fatal disease with a high percentage of deaths in some sheep breeds. In 1943, BT was first recognised in Cyprus (4) and in , BTV infected Spain and Portugal. Approximately animals were killed to eradicate the disease from Spain and Portugal (1). Between 1977 and 1979, BTV infected Turkey and some of the islands of Greece (2). On the Greek islands of the Dodecanese and Rhodes, BTV infection was recognised again in late Sheep located on the island of Lesbos developed antibody against BTV but clinical disease was not observed. In the summer of 1999, BTV affected the European part of Turkey, continental Greece and for the first time Bulgaria (3). BT affected Bulgaria again in September Antibody was detected without any clinical appearance of the disease. Samples collected in previous outbreaks in the region were identified as serotypes 4, 9 and 16. This observation of various types from the past outbreaks is indicative of the high risk of entry of multiple virus types into 212 Veterinaria Italiana, 40 (3), 2004

176 the surrounding countries and possible persistence of the virus in these regions. Objectives The objectives of the national programme in Bulgaria was to perform serum surveillance in sentinel zones, to implement Culicoides surveillance by trapping midges, and to perform phenology investigations by trapping Culicoides from the beginning to the end of the season in two villages in the Kiustendil District, which had been affected in 2001 outbreak. Materials and methods To detect antibodies against the group antigen of BTV, the competitive enzyme-linked immunosorbent assay (ELISA) (VMRD, USA) was used. Samples were considered positive if they had an average optical density ( OD) lower than or equal to the OD of the positive control of the kit; otherwise, results were calculated according to a formula provided by the manufacturer. In accordance with the National Surveillance Programme for the control of BT, serum surveillance was performed in 22 sentinel serumnegative animal herds in 2002 (Fig. 1) located in western Bulgaria at least 40 km inland from the areas affected during the epidemic in The sentinel animals were tested every 30 days for the presence of BTV antibodies. Each sentinel herd consisted of 10 cattle and 10 goats; each animal was identified with an ear-tag. The settlements with sentinel herds and villages (by district) were as follows: Vidin Archar and Drenovet Montana Brusartzi, Dolno Tzerovene and Lehchevo Vratza Furen, Krivodol, Vratza and Zli dol Sofia District Milanovo, Tzerovo, Novatchene and Skravena Sofia town Trebich, Buchovo and Lozen Pernik G. Butchino, Bosnek and Dren Kiustendil Kraynitzi, Ovchartzi and Stob. Another 42 sentinel villages (herds) located in southern Bulgaria in a 10 km border-strip zone in the Bourgas, Yambol, Haskovo, Kardjali, Smolyan and Blagoevgrad Districts were established (Fig. 1). Each sentinel herd consisted of 10 calves over six months of age and 10 lambs over four months of age. These sentinels were tested every 30 days for the presence of BTV antibodies. The settlements with sentinel herds and villages (by district) were as follows: Bourgas Rezovo, Slivarovo, Brashlyan, Granichar, Belevren and Gorno Yabalkovo Yambol Strandja, Kraynovo, Goliam Haskovo Dervent and Lesovo Radovetz, Prisadetz, Matochina, Kap, Andreevo, Mezek, Lambuch, Slaveevo, Mandritza and Dolno Lukovo Kardjali Tchernichevo, Avren, Egrek, Tichomir, Lozengradtzi, Gorno Kapinovo and Drangovo Smolyan Zlatograd, Kushla, Mochura, Arda, Buinovo and Barutin Blagoevgrad Teplen, Katuntsi, Marikostinovo, Ilinden, Dzijevo, Topolnitza, Kolarovo, Elenovo, Krupnik and Mikrevo. Sentinels located 40 km inside the 2001 outbreaks Villages with sentinel animals Villages with sentinel animals born in 2002 Figure 1 Sentinel animals in western and southern Bulgaria, 2002 The implementation of the 2002 programme commenced on 15 April and continued until 15 November The blood samples taken once every 30 days from the sentinel (indicator) large and small ruminants were tested for presence of BTV antibodies at the National Reference Laboratory. All blood samples tested in 2002 (until 26 August) were negative for the presence of antibodies to BTV. Onderstepoort light traps were used to perform Culicoides surveillance. Midges were trapped from the beginning to the end of the season in two villages in the Kiustendil District that had been affected by the 2001 outbreak. Veterinaria Italiana, 40 (3),

177 Results The blood samples collected from sentinels prior to 26 August were all BT-negative. On 26 August 2002, BTV seropositive sentinel animals were detected close to the southern border of Bulgaria (Fig. 2). Serum samples positive in the c-elisa were found in more than 20 villages but clinical disease was not observed in local animals. The phenology of Culicoides was studied in two villages (Vacsevo and Bersin in the District of Kiustendil) that had been affected by BT during the 2001 epizootic. Culicoides were trapped from 1 March 2002 to 15 November Results from trapping and Culicoides spp. activity during the summer of 2002 are shown in Tables II and III. Table I Trapped quadrants and sites: the Culicoides surveillance programme, 2002 No. Region Village/farm Sample code Villages with BTV-positive animals after 26 August 2002 Figure 2 Villages in southern Bulgaria with seropositive animals in 2002 The Culicoides trapping programme was undertaken in 23 of the 58 quadrants of km at the beginning of July 2002 in conjunction with an international BTV project in collaboration with the Institute for Animal Health (IAH) in Pirbright (Fig. 3). A total of 92 Culicoides catches were performed (Table I) principal surveillance 2002 samples for virus isolation 2001 principal surveillance Figure 3 Culicoides trapping sites and principal surveillance points in Bulgaria, Vidin Bregovo BG 1 A S1, S2 2 Vidin Bregovo BG 1 B S1, S2 3 Vidin Vidin BG 2 A S1, S2 4 Vidin Vidin BG 2 B S1, S2 5 Montana Lom BG 3 A S1, S2 6 Montana Brusartzi BG 3 B S1, S2 7 Montana Chiprovtsi BG 13 A S1, S2 8 Montana Govejda BG 13 B S1, S2 9 Montana Berkovitza BG 14 A S1, S2 10 Montana Montana BG 14 B S1, S2 11 Vratza Tichevitza BG 15 A S1, S2 12 Vratza Borovan BG 15 B S1, S2 13 Vratza Mezdra BG 26 A S1, S2 14 Vratza Roman BG 26 B S1, S2 15 Pleven Pleven BG 16 A S1, S2 16 Pleven Pleven BG 16 B S1, S2 17 Pleven Kozar BG 17 A S1, S2 18 Pleven Levski BG 17 B S1, S2 19 Lovetch Beli Osam BG 27 A S1, S2 20 Lovetch Oreshak BG 27 B S1, S2 21 Gabrovo Bogatovo BG 28 A S1, S2 22 Gabrovo Kormiansko BG 28 B S1, S2 23 Pazardjik Malo Konare BG 46 A S1, S2 24 Pazardjik Glavinitza BG 46 B S1, S2 25 Rouse Viatovo BG 8 A S1, S2 26 Rouse Chervena voda BG 8 B S1, S2 27 Razgrad Isperich BG 9 A S1, S2 28 Razgrad Isperich BG 9 B S1, S2 29 Silistra Kalipetrovo BG 10 A S1, S2 30 Silistra Alftar BG 10 B S1, S2 31 V. Tarnovo Pavlikeni BG 17 C S1, S2 32 V. Tarnovo P. Karavelovo BG 18 C S1, S2 33 Rouse G. Ablanovo BG 18 A S1, S2 34 Rouse Drianovetz BG 18 B S1, S2 35 Targoviste Popovo BG 19 A S1, S2 36 Targoviste Targoviste BG 19 B S1, S2 37 Shoumen Tzarev brod BG 20 A S1, S2 38 Shoumen Gradiste BG 20 B S1, S2 39 V. Tarnovo G. Oriachovitza BG 29 A S1, S2 40 Targoviste Antonovo BG 30 A S1, S2 41 Targoviste Krasnoseltzi BG 30 B S1, S2 42 Shumen Yangovo BG 31 A S1, S2 43 Shumen Biala reka BG 31 B S1, S2 44 St Zagora Zetevo BG 38 A S1, S2 45 St Zagora Tzenovo BG 38 B S1, S2 46 St Zagora Kran BG 39 A S1, S2 47 St Zagora Dunavtzi BG 39 B S1, S2 It is clear from Tables II and III that the seasonal activity of Culicoides in 2002 started in the third week of April and continued until 15 November. It is possible to estimate two or three peaks of activity (second half of May, August and probably the beginning of October). 214 Veterinaria Italiana, 40 (3), 2004

178 Table II Phenology investigations and seasonal dynamics of Culicoides spp. trapped in 2002 (Vaksevo Village, Kiustendil District) No. Sample code Trap Date Result 1 BG 43 А S BG 43 D S BG 43 A S BG 43 A S BG 43 A S BG 43 A S BG 43 A S BG 43 A S BG 43 A S /82* 10 BG 43 A S BG 43 A S / BG 43 A S / BG 43 A S / BG 43 A S / BG 43 A S / BG 43 A S /93 17 BG 43 A S / BG 43 A S / BG 43 A S / BG 43 A S / BG 43 A S / BG 43 A S /73 23 BG 43 A S /11 24 BG 43 A S /22 * Culicoides spp./other insects Table III Phenology investigations and seasonal dynamics of Culicoides spp. trapped in 2002 (Bersin Village, Kiustendil District) No. Sample code Trap Date Result 1 BG 43 B S BG 43 B S BG 43 B S BG 43 B S BG 43 B S BG 43 B S BG 43 B S BG 43 B S BG 43 B S BG 43 B S BG 43 B S BG 43 B S /95* 13 BG 43 B S BG 43 B S /25 15 BG 43 B S / BG 43 B S /59 17 BG 43 B S / BG 43 B S / BG 43 B S / BG 43 B S /93 21 BG 43 B S /43 22 BG 43 B S /36 23 BG 43 B S /40 24 BG 43 B S /7 25 BG 43 B S /39 26 BG 43 B S /2 27 BG 43 B S /5 28 BG 43 B S /3 29 BG 43 B S /3 * Culicoides spp./other insects Discussions, hypotheses and conclusions The absence of clinical symptoms in affected sheep, and the existence of seropositive sentinel animals only after the long summer rainfall period, suggests that insufficient ultra-violet exposure may have played a role in the absence of clinical signs of the disease. Presence of BTV seropositive sentinel animals without evidence of cross-border penetration of the infection supports the hypothesis for the special γ-δ-t-lymphocyte role of virus surviving between epidemic periods. In the past two to three years, only these regions of southern Bulgaria remained free from BTV. No BTV vaccine was used and no evidence of any virus circulation was found during this period. The entire population of sentinel ruminants remained BTV-negative but are susceptible to infection. Firstly, the hypothesis of BTV remaining in a γ-δ-t-lymphocyte in a population of sensitive animals outside Bulgaria is also possible. Secondly, it is possible BTV penetrated the region from areas located close to the 10-km border strip. This hypothesis should be confirmed by detecting BTV by c-elisa or by virus isolation from trapped samples of Culicoides spp. from these regions or after estimating a dramatic change in the vectors with C. imicola in the trapped Culicoides samples. The active Culicoides seasonal dynamics in 2002 started in the third week of April and two or three peaks of activity are estimated to have occurred: one during the second half of May, another in August and probably a third at the beginning of October. These peaks are equal to five or six generations throughout the season. The following conclusions can be drawn: 1) Based on complex viral, serological, epidemiological and clinico-pathological investigations, evidence of the presence of active BTV infection in Bulgaria was not been found. 2) Prior to 26 August 2002, there was no transborder incursion of BTV into Bulgaria by persistently infected animals or by infected Culicoides spp. 3) In three years of Culicoides surveillance, C. imicola has not been detected in Bulgaria. 4) The dominant Culicoides species recorded were C. obsoletus, C. pulicaris and C. punctatus. 5) Amongst Culicoides midges trapped in August 2001 in south-east Bulgaria, C. puncticollis was recorded for the first time. Veterinaria Italiana, 40 (3),

179 6) Culicoides phenology studies indicated two or three main peaks of Culicoides activity representing the occurrence of five or six generations between the beginning and end of the 2002 season. Acknowledgements Grateful thanks are extended to P.S. Mellor & B. Purse from the Institute for Animal Health in Pirbright for the estimation of the Culicoides catches and for their valuable collaboration. 3. Office International des Épizooties (OIE) (1999). Bluetongue in Bulgaria. Dis. Info., 12, Sellers R.F. (1975). Bluetongue in Cyprus. Aust. Vet. J., 51 (4), Urbano P. & Urbano F.G. (1994). The Reoviridae family. Comp. Immunol. Microbiol. Infect. Dis., 17 (3-4), Verwoerd D.W. & Erasmus B.J. (1994). Bluetongue. In Infectious diseases of livestock with special reference to southern Africa (J.A.W. Coetzer, G.R. Thomson & R.C. Tustin, eds). Oxford University Press, Cape Town, References 1. Mellor P.S., Jennings D.M., Wilkinson P.J. & Boorman J.P. (1985). Culicoides imicola: a bluetongue vector in Spain and Portugal. Vet. Rec., 116, Mostroyannic M., Axiotis I. & Strofovos E. (1981). Study of the first bluetongue disease in Greece. Bull. Hell. Vet. Med. Cos., 32, Veterinaria Italiana, 40 (3), 2004

180 Vet. Ital., 40 (3), Epidemiology and vectors Spatial distribution of bluetongue in cattle in southern Croatia in the last quarter of 2002 A. Labrović (1), Z. Poljak (2), S. Šeparović (1), B. Jukić (2), D. Lukman (1), E. Listeš (3) & S. Bosnić (3) (1) Ministry of Agriculture and Forestry, Veterinary Administration, Ulica grada Vukovara 78, Zagreb, Croatia (2) Veterinary Faculty of the Zagreb University, Heinzelova 55, Zagreb, Croatia (3) Croatian Veterinary Institute, Savska cesta 143, Zagreb, Croatia Summary The domestic ruminant population of southern Croatia was affected by bluetongue (BT) in late A sentinel cattle scheme was developed to detect the presence of bluetongue virus (BTV) activity in the domestic cattle population in the protection zone (based on the distribution of BT in 2001: Dubrovacko-Neretvanska County and the southern area of the Splitsko-Dalmatinska County) as well as in the surveillance zone (the northern area of the Splitsko-Dalmatinska County). Twentyfive villages were selected to serve as sentinel locations during the observation period which lasted from 15 September to 15 December Seroconversion was not detected in cattle in sentinel locations in the surveillance zone. However, in the protection zone, serum antibodies to BTV serotype 9 were detected in eight cattle in five of the ten sentinel locations. Although no clinical case of BT disease was detected in sheep on mainland Croatia in late 2002, BTV activity was present in sentinel cattle in the protection zone. When compared with 2001, spatial distribution of the locations in which cattle seroconverted to BTV-9 in the last quarter of the 2002 suggests a northward trend to the spread of BTV in the cattle of southern Croatia. Keywords Bluetongue virus Cattle Croatia Sentinel herd Spatial data. Introduction Over the last several years a significant change has been recorded in the epizootiological situation of bluetongue (BT) in Europe. In Croatia, there is no documented evidence of BT virus (BTV) infection in the ruminant population prior to While BTV can infect most ruminants, BT disease primarily affects sheep (8), with the incidence of clinical disease highly variable (5). While natural and experimental BTV infection of cattle is asymptomatic in the vast majority of cases, rare and authentic instances of disease are likely to occur (7). BTV infection of cattle often results in prolonged viraemia, hence cattle serve as a reservoir from which the virus may be recovered by the haematophagous insect vectors and then transmited to other ruminants (6). This could explain why bovines are often used as early warning sentinel animals to determine the prevalence of BTV activity in an area in which clinical BT disease has not been endemic. Sentinel surveillance can provide clearly defined data (usually incidence data) on a regular basis. Although not all outbreaks are likely to be detected at an early stage if surveillance is limited to sentinel sites, a sentinel herd scheme may enable the detection of BTV circulation between domestic ruminants and vectors and can assist in the identification of risk factors associated with the occurrence of BT disease. Furthermore, when sentinel sites are randomly selected, generalisation of the results can be applied to a wider population. Sentinel herd schemes are particularly effective surveillance and monitoring tools in regions where clinical BT disease is uncommon (10). In Queensland, Australia, 47 sentinel herds (with cattle in each herd), most of which were bled monthly over the period from 1990 to 1992 (10), gave valuable information on BTV serotypes present in Queensland. In 1988, the official BT policy adopted by Canada included a surveillance mechanism for determining BTV seroprevalence by Veterinaria Italiana, 40 (3),

181 using a sentinel herd programme (2). The sentinel herd programme consisted of six herds of cattle (with 7 animals in each herd), which were distributed along or within 48 km of the Canada-United States border (2). The identification of two infected animals illustrated the effectiveness of the sentinel herd programme. Greece initiated a sentinel cattle programme in 1999 (4). The programme commenced on 30 June 1999 and animals were sampled approximately every 15 days. The sentinels were posted in four locations: Rhodes, Kos, Leros and Samos, with 50, 51, 18 and 18 animals per location, respectively, in two to five villages per location. This programme also proved to be effective in detecting animals that seroconverted. The main objective of the 2002 sentinel cattle scheme implemented in the two southernmost counties of Croatia (Dubrovacko-Neretvanska and Splitsko-Dalmatinska) was to investigate whether there was BTV activity present in cattle populations. In addition, one of the aims of the scheme was to obtain baseline data required for the planning of the 2003 sentinel cattle programme to more accurately inform decision-makers on ruminant movement control. Materials and methods A sentinel cattle pilot scheme was conducted from 15 September to 15 December Surveillance of sentinel cattle in the protection zone (defined as having a 100-km radius around the infected holdings and so included Dubrovacko-Neretvanska County and the southern part of Splitsko-Dalmatinska County) was designed to detect seroconversion in cattle at a prevalence of 5%. For this purpose, 60 cattle in 10 sentinel locations (6 cattle per location) were sampled approximately every 15 days. The criterion for inclusion of cattle in a sentinel study was that the animals had to be serologically negative and over six months of age. Those cattle which became seropositive to BT were replaced with seronegative cattle. To determine the number of cattle (sample size) that would serve as sentinel animals in the surveillance and protection zones, the formula of Cannon and Roe was used to detect the presence of the disease (based on the assumption of a perfect test) (1). This formula may be used either for a herd or for a welldefined geographic zone (9). The number of animals required for the purpose of estimating the prevalence of BTV antibodies at levels of 2% and 5% were 147 and 59, respectively. This enables detection of at least one seropositive animal if BT infection is present in the cattle population at and above the specified prevalence level. Serum samples were screened with competitiveenzyme-linked immunosorbent assay (c-elisa,) (VMRD Inc, USA). To declare a bovine animal positive, two positive reactions to the c-elisa were necessary. Visualisation of spatial data was performed using ArcView 8.2 (3). Results The spatial distribution of sheep flocks showing clinical signs of BT, as well as cattle herds found to have BTV antibodies in late 2001, are presented in Figure 1. Sentinel cattle in the surveillance zone (defined as having a 100-km radius around the protection zone and so included the northern area of Splitsko- Dalmatinska County) were monitored to detect seroconversion in cattle at a prevalence of 2%. For this purpose, 150 cattle in 15 sentinel locations (10 cattle per location) were sampled approximately every 30 days. The three criteria for selecting locations where sentinel animals should be placed were ruminant population density (e.g. the higher the number of cattle in the village, the higher the probability that the village would be selected), locations which the local Veterinary Service knows, or suspects, to be suitable for vectors, and to have owners willing to participate in the sentinel scheme. County Location of sheep and cattle farms km Figure 1 Spatial distribution of flocks in which sheep showed clinical signs of bluetongue as well as herds in which cattle had antibodies to BTV-9 Last quarter of Veterinaria Italiana, 40 (3), 2004

182 The spatial distribution of sentinel cattle in the protection and the surveillance zones from September to December 2002 is shown in Figure 2. detected in one sentinel cow per location). On 16 November, BTV antibodies were detected in a cow in Mihanici. No clinical sign of BT disease was reported in any ruminant animal in the locations where BT-seropositive cattle were detected. The spatial distribution of locations in the protection zone in which BT seroconversion was detected in sentinel cattle during the 2002 sentinel programme is shown in Figure 3. Antibodies to BTV were not detected in sentinel cattle in the surveillance zone. Discussion County Sentinel surveillance zone Sentinel protection zone km Figure 2 Spatial distribution of sentinel villages in the protection and surveillance zones 15 September-15 December 2002 During the observation period in the protection zone, serum antibodies to BTV were detected in eight cattle in 5 of the 10 sentinel locations. More specifically, from 14 to 17 October 2002, BTV antibodies were detected in one sentinel cow in Krstatice, Runovici and Krvavac and in two sentinel cows in Podgorje. Further BTV activity in sentinel cattle in the protection zone was recorded between 2 and 4 November 2002 in the sentinel locations of Podgorje and Krvavac (BTV antibodies being The finding that BTV antibodies were present in sentinel cattle in southern Croatia in late 2002 suggests that viral activity continues in the absence of the disease in the areas of mainland Croatia that were affected by BT in However, in December 2002, only one isolated outbreak of BT disease in one sheep flock was detected on the Island of Hvar. This could be explained by the fact that BTV is probably maintained by a cycle of infection in the insect vector and cattle, and only when the vector population is very high does the virus spill over into other species such as sheep (6). The results of the 2002 sentinel cattle scheme in southern Croatia also suggest that although the cattle densities in the Counties of Dubrovacko- Neretvanska Splitsko-Dalmatinska are very low, they still seem high enough to sustain the virus. However, since a detailed analysis of the factors contributing to the presence of BTV activity in 2002 was not performed, we can only speculate as to whether BTV had been present because it overwintered in vectors and/or in hosts, or because of new BTV incursions that occurred in BTV-positive cattle, mid-october 2002 Sentinel surveillance zone, 2002 Sentinel protection zone, 2002 BTV-positive cattle, early November 2002 Sentinel surveillance zone, 2002 Sentinel protection zone, 2002 BTV-positive cattle, mid-november 2002 Sentinel surveillance zone, 2002 Sentinel protection zone, 2002 Figure 3 Spatial distribution of sentinel cattle that seroconverted to BTV-9 14 October-16 November 2002 Veterinaria Italiana, 40 (3),

183 Better utilisation of data collected during the 2002 sentinel programme as well as the sentinel programme that commenced on 15 July 2003, might provide more comprehensive information on the epidemiology of BT in southern Croatia. Although the nature of the disease and the clustering level might require a different approach as well as adjustments of the sample size (a perfect test in detection of BTV antibodies was not used), the results of the 2002 sentinel cattle scheme provided valuable information which served to aid in the decision to not only continue but to extend further the BT surveillance programme to the north-western areas of Croatia. This information also proved to be valuable for animal movement control measures. Acknowledgements Grateful thanks are extended to Ivana Lampek and GISDATA Croatia for providing the authors with the electronic map of Croatia, and to participating farmers and veterinarians. References 1. Cannon R.M. & Roe R.T. (1982). Livestock disease surveys: a field manual for veterinarians. Australian Government Publishing Service, Canberra. 2. Clavijo A., Munroe F., Zhou E.-M., Booth T.F. & Roblesky K. (2000). Incursion of bluetongue virus into the Okanagan Valley, British Columbia. Can. Vet. J., 41, Environmental Systems Research Institute, Inc. (ESRI) (2003). ArcViewGis TM spatial analyst. ESRI Redlands, California (esri.com/software/arcgis/ index.html/ accessed on 15 August 2004). 4. European Commission (EC) (1999). Final report of a Mission carried out in Greece from 18 October to 22 October 1999 in relation to bluetongue. European Commission Health and Consumer Protection Directorate General. DG(SANCO)/1157/1999 MR final. EC, Brussels. 5. Gibbs E.P.J. & Greiner E.C. (1994). The epidemiology of bluetongue. Comp. Immunol. Microbiol. Infect. Dis., 17, MacLachlan N.J. (1994). The pathogenesis and immunology of bluetongue virus infection of ruminants. Comp. Immunol. Microbiol. Infect. Dis., 17, MacLachlan N.J., Barratt-Boyes S.M., Brewer A.W. & Scott J.L. (1992). Bluetongue virus infection of cattle. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Pearsons I.M. (1992). Overview of bluetongue virus infection in sheep. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Putt S.N.H., Show A.P.M., Woods A.J., Tyler L. & James A.D. (1988). Veterinary epidemiology and epidemics in Africa: a manual for use in the design and appraisal of livestock health policy, 2nd Ed. Food and Agriculture Organization, Rome, H.CA Manual No. 3 (fao.org/wairdocs/ilri/x5436e/x5436e00.htm accessed on 11 September 2004). 10. Ward M.P., Flanagan M., Carpenter T.E., Hird D.W., Thurmond M.C., Johnson S.J. & Dashort M.E. (1995). Infection of cattle with bluetongue viruses in Queensland, Australia: results of a sentinel herd study, Vet. Microbiol., 45, Veterinaria Italiana, 40 (3), 2004

184 Vet. Ital., 40 (3), Epidemiology and vectors Serological evidence of bluetongue and a preliminary entomological study in southern Croatia E. Listeš (1), S. Bosnić (2), M. Benić (2), M. Lojkić (2), Ž. Čač (2), Ž. Cvetnić (2), J. Madić (3), S. Šeparović (4), A. Labrović (4), G. Savini (5) & M. Goffredo (5) (1) Croatian Veterinary Institute, Institute Split, Poljička cesta 33, HR Split, Croatia (2) Croatian Veterinary Institute, Savska Cesta 143, HR Zagreb, Croatia (3) School of Veterinary Medicine, Heinzelova 55, HR Zagreb, Croatia (4) Ministry of Forestry and Agriculture, Department of Veterinary Service, Ulica grada Vukovara 78, HR Zagreb, Croatia (5) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy Summary In December 2001, bluetongue (BT) was confirmed serologically by the Croatian Veterinary Institute using the competitive enzyme-linked immunosorbent assay (c-elisa). Results of the serological testing of blood samples from ruminants in the Dubrovacko-Neretvanska County are presented (3 318 sera of ruminants from 53 herds were examined). In total, 357 bovine sera (178 or 49.9% positive), ovine sera (174 or 13.7% positive) and caprine sera (270 or 15.9% postive) were tested. Antibodies to BT virus serotype 9 were detected in 212 of the positive sera by serum neutralisation. A preliminary light-trap survey for midges of the Culicoides genus was also performed in the Dubrovacko-Neretvanska County. Fourteen light-trap collections from seven locations were examined and yielded a total of Culicoides of which (92%) belonged to the Obsoletus Complex (including C. obsoletus and C. scoticus). Keywords Bluetongue Competitive enzyme-linked immunosorbent assay Croatia Culicoides Light trap Ruminant Serum neutralisation test. Introduction Bluetongue (BT) is a viral, non-contagious, arthropod-borne, infectious disease of domestic and wild ruminants caused by the BT virus (BTV) of the family Reoviridae, genus Orbivirus (5). To date, twentyfour virus serotypes have been identified worldwide (7). Biting midges of the genus Culicoides (Diptera: Ceratopogonidae), the biological vectors of BTV, play the principal role in disease spread. Culicoides imicola, subgenus Avaritia, which is the most widely spread species of bloodsucking Culicoides, acting as vectors of BT and African horse sickness, occurs in Africa and most countries of the Mediterranean Basin (8, 14). In central and southern Europe, C. obsoletus is one of the most common species encountered (11). More than 50 virus types have been isolated from Culicoides worldwide (8). For bluetongue, clinical symptoms vary from being subclinical to mild and acute, and can terminate in the death of the animal. Symptoms are most pronounced in sheep. The disease can be seen in the acute or subacute form, depending on the pathogenicity of the viral strain and the sensitivity of the breed. In sheep, the predominant symptoms are elevated body temperature, oral lesions, lameness, catarrhal rhinitis, oedema of the tongue, lips, intermaxillary space and neck (4). In cattle and goats, the disease usually assumes a subclinical course without severe symptoms (4). The disease was first identified and described in South Africa; however, with time, it has also been recorded from other regions of the world, i.e. in the Caribbean, Asia, Australia, North and South America, and Europe. BT is found only in regions where competent vectors occur, between latitudes of 35 S and 40 N. However, in the western areas of North America and in the People s Republic of China, the disease has spread to 50 N (11). In Veterinaria Italiana, 40 (3),

185 Europe, the disease has previously been recorded in Cyprus (1924, 1943 and 1977), Portugal and Spain ( ), and on the Greek islands of Rhodes and Lesbos ( ). Between 1998 and 2004, an epizootic of BT occurred in the Mediterranean Basin, also affecting countries where BT had never previously been recorded. At least four virus serotypes circulated during the epizootic, which progressed in two directions, with eastern and western fronts and with two sources of infection (11). The eastern front of the BT epizootic originated on the Greek islands (Rhodes, Leros, Kos and Samos) in October 1998 where serotype 9 was identified. During 1999, serotype 9 was confirmed in southern Bulgaria, the European part of Turkey and on mainland Greece, along the borders with Bulgaria and Turkey (the epizootic covering one third of Greece). In addition to serotype 9, serotypes 4 and 16 were also identified (11). During the years that followed, the disease was recorded for the first time in Kosovo, Serbia and Montenegro, the Republic of Macedonia and Croatia (11). In 2002, BT was recorded in Albania and Bosnia and Herzegovina for the first time. The western front of the epizootic originated in Tunisia in December 1999 where the virus was serotyped as BTV-2 which, in 2000, spread to Italy, Corsica (France) and the Balearic Islands (Spain) (11). Results of serological studies to determine the prevalence and distribution of BT antibodies, and the conclusions of a preliminary study of the Culicoides species present in the Dubrovacko- Neretvanska County, are presented. The study was conducted after clinical symptoms of the disease were detected in September and November Material and methods Competitive enzyme-linked immunosorbent assay The testing of sera was performed by the competitive-enzyme-linked immunosorbent assay (c- ELISA) for detection of BT antibodies in ruminant sera using a commercial kit (VMRD Inc., USA) (1). Tests were conducted in accordance with the instructions of the manufacturer, with the known positive and negative controls tested on each plate. All serum samples were tested in duplicate. The optical density of each well was read at a wavelength of 620 nm on a microplate reader (Anthos II, Austria). Microtitre serum neutralisation The serum neutralisation (SN) test was used to type BTV in ruminant sera (6). Positive and negative controls for SN were kindly provided by the Onderstepoort Veterinary Institute (OVI) (OIE Reference Laboratory) in South Africa. The microtitre neutralisation method was used in this study. Fifty µl of serum dilutions, from 1:10 to 1:1 280, were added to each well of flat-bottomed microtitre plates and mixed with an equal volume of OIE standard reference sera for BTV serotypes 2, 4, 9 and 16 (100 TCID 50 ). These were incubated at 37 C in 5% CO 2. After 1 h incubation, approximately 10 4 Vero cells were added to each well at a volume of 100 µl of minimum essential medium (MEM) containing antibiotics. After incubation for 4-6 days, the test was read using an inverted microscope. The wells were rated according to the degree of cytopathic effect (CPE) observed. A sample was considered positive when CPE neutralisation exceeded 50% at the lowest dilution (1:10). The serum titre represented the highest serum dilution capable of neutralising more than 50% of CPE in tissue culture. Culicoides collections Insect collections were made using three light traps (220V, 8W) placed close to the sheep, goats and cattle, and operated before dusk until after dawn in September 2002 at seven locations in the Dubrovacko-Neretvanska County. The insects were kept in 70% ethyl alcohol. The fourteen insect collections were made in accordance with the protocols of the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche) in Teramo, Italy. Results In October and November 2001, clinical symptoms indicative of BT were observed in sheep herds in Konavle (in the southernmost area of Croatia along the borders with Montenegro and Bosnia and Herzegovina) (Fig. 1). Symptoms included abundant salivation, oedema of the tongue, buccal and intermaxillary regions, erosive stomatitis, elevated body temperature (40.5 C-40.7 C), coronitis with resulting lameness and sporadic abortion during late gestation. Symptoms of the disease were not observed in cattle, whereas some sporadic cases were recorded in goats. 222 Veterinaria Italiana, 40 (3), 2004

186 Dubrovačka and along the Dubrovnik coast, and from one goat herd on the Pelješac Peninsula. Due to the small number of animals in individual herds, all ruminants from a village were considered a herd. All animals are usually kept indoors. A total of serum samples from 53 herds were tested. These included 357 cattle sera, 178 (49.9%) of which gave positive results; of sheep sera, 174 (13.7%) were positive; and of goat sera, 270 (15.9%) were positive. The overall seroprevalence in all ruminants studied was 18.7%. The SN test revealed antibodies to serotype 9 in 212 of the positive sera. The geographic distribution of the animals tested is presented in Table I. Figure 1 Clinical symptoms of bluetongue in a sheep from Konavle, southern Croatia, November 2001 A team from the Croatian Veterinary Institute visited the area to take serum samples from sheep herds that presented clinical symptoms of the disease. A sheep with very pronounced symptoms was sacrificed, and a gross pathological examination performed (Fig. 2); the liver, spleen and lymph nodes were despatched for virological testing. Table I Distribution of animals tested for bluetongue in Croatia, January-June 2002 Area Cattle Goats Sheep Total Dubrovnik coast Dubrovnik Konavle Pelješac Total Culicoides were identified on the basis of their wing pattern and findings were verified by the Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale in Teramo, Italy. A total of Culicoides were collected; of these, (92%) belonged to the Obsoletus Complex and included C. obsoletus and C. scoticus. The other Culicoides species identified were C. circumscriptus, C. paolae, C. pulicaris, C. punctatus, C. seavanicus, C. fascipennis, C. haranti and C. fagineus. The Obsoletus Complex was represented in 13 of the 14 catches. The mean air temperature at the time of the study was 20 C. Discussion Figure 2 Haemorrhage in the tunica media at the base of the pulmonary artery of a sheep from Konavle, southern Croatia, November 2001 After the presence of BTV antibodies in the blood samples of the sheep was confirmed, serological testing was performed to determine the spread of the disease in Croatia. From the end of January to June 2002, blood samples were obtained from the majority of ruminants in the Konavle community, and randomly from the communities of Župa Serological testing and the preliminary survey for the presence of Culicoides species, following the occurrence of clinical symptoms characteristic of BT in the Dubrovacko-Neretvanska County (situated in the southernmost area of Croatia), are presented. The area is 52 m above sea level and lies at 42 N, corresponding to the latitude at which BTV and BTV vectors have been recorded in other parts of the world (11). Between 1993 and 2002, the mean daily air temperature in the area in September was 21.3 C, mean air humidity 61.4%, and mean monthly precipitation 90.8 mm. Veterinaria Italiana, 40 (3),

187 Croatia had never previously recorded the presence of BT and the present study revealed an overall seroprevalence of 18.7%. Of the ruminants examined, the highest total seroprevalence was recorded in cattle (49.9%); this is consistent with the opinion that due to protracted viraemia, cattle enable viral replication and thus become virus reservoirs (2). In sheep, seroprevalence was 13.7%, whereas in goats the percentage was 15.9%. Sheep presented very pronounced symptoms of the disease, whereas in goats the infection developed without notice showing only sporadic symptoms of excessive salivation and intermaxillary oedema. Highest seroprevalences occurred in cattle (54.87%), goats (18.92%) and sheep (17.74%) in Konavle (Table II). Seroprevalence was much lower in other parts of the region, decreasing from the south-east (Konavle) to the north-west including Dubrovnik where the seroprevalence rate in cattle reached 75% (but only 20 animals were tested), in goats 7.34% and in sheep 7.14%. Along the Dubrovnik coast, a seroprevalence rate of 18.33% was recorded in cattle, 6.4% in goats and 5.03% in sheep. In the Pelješac Peninsula, seroprevalence in goats was 4.65% (Fig. 3). The differences observed between geographic areas are highly significant (χ 2 =89.4, P<0.0001). Conclusions The predominant species of Culicoides captured in Croatia were those belonging to the Obsoletus Complex, and included C. obsoletus and C. scoticus (92%). However, the following species were also Table II Percentage prevalence of bluetongue antibodies in livestock in the bluetongue-affected areas of Croatia, January-June 2002 Area Cattle Goats Sheep Total Dubrovnik coast Dubrovnik Konavle Pelješac Total identified: C. circumscriptus, C. paolae, C. pulicaris, C. punctatus, C. seavanicus, C. fascipennis, C. haranti and C. fagineus. BTV has been isolated from C. obsoletus in Cyprus (9). African horse sickness virus was found in mixed pools of C. obsoletus and C. pulicaris in Spain; both species are widespread in Europe (14). Culicoides obsoletus comprises a complex of at least four species with almost identical wing patterns (12). The remaining species identified (C. circumscriptus, C. pulicaris, C. seavanicus, C. fascipennis and C. fagineus) are also found in Greece (10), whereas C. paolae is a new species recently discovered in southern Italy (3). Climatic factors play an important role in the occurrence of BTV infection in animals and also influence the size of vector populations and periods of their seasonal activity. Culicoides activity has not been recorded below approximately 13 C and above 35 C (13). An analysis of climatic data was used to model the potential distribution of C. imicola in Bosnia and Herzegovina Pelješac 4.65 Seroprevalence (%) Cattle Goats Sheep Dubrovnik coast Dubrovnik Konavle Montenegro Figure 3 The areas of Croatia affected by bluetongue, showing the seroprevalence rates (%) in cattle, goats and sheep, January- June Veterinaria Italiana, 40 (3), 2004

188 Europe, predicting that C. imicola might have spread from Spain, Greece and Italy to some areas along the Croatian coast as well as to the coastal areas of Albania, Serbia and Montenegro, and Bosnia and Herzegovina (15). However, C. imicola has not yet been reported from these latter Balkan states. It appears quite conceivable that BT was spread by infected Culicoides or by infected animals from neighbouring Serbia and Montengro. This presumption is supported by disease reports from Kosovo in October and from Serbia and Montenegro in November 2002 (11). The Istituto Zooprofilattico Sperimentale in Teramo screened a total of 212 sera using the SN test. The presence of BT antibodies in these sera had previously been demonstrated by the c-elisa. Antibodies to BTV serotype 9 were identified by the SN test, confirming the presumption that BT had spread to Croatia along the eastern front of the epizootic in the Mediterranean Basin, since the infection with BTV serotype 9 had previously been confirmed in Turkey, Greece and Bulgaria (11). References 1. Afshar A., Thomas F.C., Wright P.F., Shapiro J.L., Shettigara P.T. & Anderson J. (1987). Comparison of competitive and indirect enzyme-linked immunosorbent assays for detection of bluetongue virus antibodies in serum and whole blood. J. Clin. Microbiol., 25, Barratt-Boyes S.M. & MacLachlan N.J. (1995). Pathogenesis of bluetongue virus infection of cattle. JAVMA, 9, Boorman J., Mellor P.S. & Scaramozzino P. (1996). A new species of Culicoides (Diptera, Ceratopogonidae) from southern Italy. Parassitologia, 38, Erasmus B.J. (1975). Bluetongue in sheep and goats. Aust. Vet. J., 51, Erasmus B.J. (1990). Bluetongue virus. In Virus infections of ruminants, Vol. 3 (Z. Dinter & B. Morein, eds.) Elsevier, Amsterdam and New York, Gard G.P. & Kirkland P.D. (1993). Bluetongue virology and serology. In Australian standard diagnostic techniques for animal diseases (L.A. Corner & T.J. Bagust, eds). CSIRO Information Services, Melbourne, Gibbs E.P.J. & Greiner E.C. (1994). The epidemiology of bluetongue. Comp. Immunol. Microbiol. Infect. Dis., 17, Meiswinkel R., Nevill E.M. & Venter G.J. (1994). Vectors: Culicoides spp. In Infectious diseases of livestock with special reference to southern Africa, Vol. 1 (J.A.W. Coetzer, G.R. Thomson & R.C. Tustin, eds.) Oxford University Press, Cape Town, Mellor P.S. & Pitzolis G. (1979). Observations on breeding sites and light-trap collections of Culicoides during an outbreak of bluetongue in Cyprus. Bull. Entomol. Res., 69, Mellor P.S., Jennings M. & Boorman J.P.T. (1984). Culicoides from Greece in relation to the spread of bluetongue virus. Rev. Elev. Med. Vet. Pays Trop., 37 (3), Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean Basin, Vet. J., 164, Rawlings P. (1996). A key, based on wing patterns of biting midges (genus Culicoides Latreille, Diptera: Ceratopogonidae) in the Iberian Peninsula, for use in epidemiological studies. Graellsia, 52, Ward M.P. & Thurmond M.C. (1995). Climatic factors associated with risk of seroconversion of cattle to bluetongue viruses in Queensland. Prev. Vet. Med., 24, Wittmann E.J. & Baylis M. (2000). Climate change: effects on Culicoides-transmitted viruses and implications for the UK. Vet. J., 160, Wittmann E.J., Mellor P.S. & Baylis M. (2001). Using climate data to map the potential distribution of Culicoides imicola (Diptera: Ceratopogonidae) in Europe. Rev. Sci. Tech. Off. Int. Épiz., 20 (3), Veterinaria Italiana, 40 (3),

189 Vet. Ital., 40 (3), Entomological surveillance of bluetongue in France in 2002 T. Baldet (1), J.-C. Delécolle (2), B. Mathieu (3), S. de La Rocque (1) & F. Roger (1) (1) CIRAD-EMVT, TA 30 E, Campus International de Baillarguet, Montpellier, France (2) Musée Zoologique de l Université Louis Pasteur et de la Ville de Strasbourg (MZS), 29 bd de la Victoire, Strasbourg, France (3) Entente interdépartementale pour la démoustication du littoral méditerranéen (EID-Méditerranée), 165 avenue Paul Rimbaud, Montpellier Cedex 4, France Summary Bluetongue (BT) is an arboviral disease that appeared in the Mediterranean in In France, the principal vector, Culicoides imicola, was detected for the first time in Corsica in October 2000, a few weeks before outbreaks of BT virus serotype 2 (BTV-2). Entomological surveillance was implemented in Corsica and on mainland France in In Corsica, the aim was to study the population dynamics of C. imicola and other potential vectors. On the coastal mainland, the objective was to survey the introduction of C. imicola. One-night catches per site were performed every three weeks at 12 sites on Corsica and every month at 19 sites on the mainland. In Corsica, Culicoides belonging to 50 species were collected over 180 nights. C. imicola accounted for 18.3% of the total captured. On the mainland, Culicoides (44 species over 109 nights) were collected, none of which were C. imicola. The geographic and seasonal distribution of C. imicola and other species of interest are discussed in relation with their bio-ecology and environment. These datasets are essential for a better understanding of the epidemiology of BT, and to create and validate predictive models based on remote sensing in order to identify areas at risk for BT. Keywords Bluetongue Corsica Culicoides imicola Culicoides newsteadi Culicoides obsoletus Culicoides pulicaris France Surveillance. Introduction Bluetongue (BT) is an infectious arthropod-borne viral disease that affects ruminants, mainly sheep. BT virus (BTV) is transmitted between its vertebrate hosts by certain species of Culicoides biting midges (Diptera: Ceratopogonidae). BT occurs pantropically between 44 N and 35 S depending mainly upon the distribution and seasonal presence of Culicoides midges. More than Culicoides species have been identified in the world (1), but only 17 have been connected with BTV transmission. The major vector species are C. imicola and C. bolitinos in Africa; C. imicola and C. fulvus in Asia; C. brevitarsis and C. fulvus in Australia; C. sonorensis in North America, C. insignis and C. pusillus in South and Central America (9). Worldwide, BTV causes losses of approximately US$3 billion each year (13). Indeed, BT is considered to be of such major international concern that it has been given List A status by the Office International des Épizooties. Since its appearance in the Mediterranean Basin in 1998, BT disease has had a devastating effect on the sheep industry, resulting in the loss of over sheep until 2001 (10). The spread of BT into areas of Europe never previously affected, is linked to the northern spread of its main Afro-Asian vector C. imicola (2) by air streams. C. imicola sensu lato is a complex of at least 10 sibling species but, at present, only C. imicola sensu stricto is present in Europe (8). In the eastern Mediterranean Basin, BTV outbreaks have occurred in regions of the Balkans up to N, and in places where C. imicola has not been been detected during previous insect surveys (10). Recently, BTV-2 has been isolated in one or more species of the C. obsoletus complex on mainland Italy (15) and C. pulicaris on the island of Sicily (5), suggesting that BTV could be transmitted by these Palaearctic Culicoides species associated with livestock. As these species or species complexes are widespread and abundant in the Mediterranean and across most of northern Europe, it would be of 226 Veterinaria Italiana, 40 (3), 2004

190 interest to confirm whether they are able to transmit BTV in the field. In France, the main vector, C. imicola, was detected for the first time on the island of Corsica in October 2000 (7). Subsequently, important outbreaks caused by BTV-2 occurred in autumn 2000 and 2001 (18, 19, 20). In combination with the vaccination campaigns conducted between 2001 and 2003, an entomological surveillance network was established in 2002, as follows: a) in Corsica, to study the population dynamics of C. imicola and other potential vector species (C. obsoletus and C. pulicaris) b) on the coastal mainland of France, to survey the introduction of C. imicola and study the population dynamics of other Culicoides species. Materials and methods Twelve representative sites were selected in Corsica (farms affected by BTV-2 outbreaks in 2000 and/or 2001) and 19 farms at risk on the coastal mainland at intervals of 50 km (Fig. 1). One-night catches per site using UV-light traps (7) were performed every three weeks in Corsica and every month on the mainland. Results To facilitate rapid access to entomological surveillance data, results are available on the website of the Centre de coopération internationale en recherche agronomique pour le développement (CIRAD) at bluetongue.cirad.fr/resultats_entomologiques/consultati on.php. Corsica From February to December 2002, Culicoides belonging to at least 50 species were collected in a total of 180 night catches. C. imicola accounted for 18.3% of the specimens captured, with a maximum and average catch size of and 86, respectively (Table I). C. imicola was predominant at two sites located in the extreme south of the island (Porto- Vecchio and Figari) and less prevalent in Moltifao, the only site located inland at an altitude of 250 m (Fig. 2). Adult densities reached a peak in September-October, at the end of the summer (Fig. 3). The other most abundant and widespread species were C. newsteadi (36.8%, and 174), C. scoticus (18%, and 85), C. obsoletus (9.2%, and 43), C. circumscriptus (4.2%, 507 and 20) and C. pulicaris (3.8%, 626 and 18). The 43 remaining species accounted for less than 2% of the total. C. newsteadi, a halophilic species, is predominant in the western and southern coastal plains and is rare on the eastern and northern rocky coasts. C. pulicaris, a species found more often in low and midland Table I Adults of Culicoides collected in Corsica, February-December 2002 (12 sites/11 months/180 night catches) Figure 1 Location of sites surveyed for Culicoides in France, 2002 Traps were located outdoors, within 25 m of livestock premises and suspended from the walls of buildings m above ground level. Traps were set 1 h before sunset and collection was made at about 8 am the next morning. The insects were transported to the laboratory in a water-filled beaker and then covered and preserved in 90% ethanol. Ceratopogonidae were first separated from all other insects. Culicoides were identified based on wing patterns, and subsequently confirmed by mounting specimens on microscope slides (6, 16). Culicoides species Number collected (%) Size of maximum catch Size of average catch C. newsteadi (36.8) C. imicola (18.3) C. scoticus (18.0) C. obsoletus (9.2) C. circumscriptus (4.2) C. pulicaris (3.8) C. griseidorsum (1.8) C. subfagineus (1.3) C. lupicaris (1.3) Culicoides spp. (40 species) (5.1) 24.2 Total Culicoides (50 species) (100) * The 40 remaining species accounted for less than 2% of the total Veterinaria Italiana, 40 (3),

191 with less adults for each species. As in Corsica, C. newsteadi is predominant along the eastern coastal plains backed by marshland (from the Camargue, Rhône delta, to the Spanish border). C. pulicaris is more abundant on the western rocky coast (from the Rhône delta to the foothills of the Alps along the Italian border) and along the Spanish border at the foothills of the Pyrenees (Fig. 4). Similar to the findings in Corsica, a bimodal pattern of seasonal adult dynamics for C. obsoletus and C. pulicaris was observed with two peaks: one in spring and the other in autumn (Fig. 5). Table II Adults of Culicoides captured on the coastal mainland France, April-November 2002 (19 sites/8 months/109 night catches) No specimens of C. imicola were found Culicoides species Number collected Size of maximum catch Size of average catch C. newsteadi (73.5) C. obsoletus (8.0) C. scoticus 877 (5.4) Culicoides imicola Culicoides newsteadi Culicoides obsoletus Culicoides pulicaris Culicoides spp. Figures: average catch of Culicoides per site (C. imicola) Figure 2 Spatial distribution of Culicoides collected in Corsica, 2002 (12 sites/11 months/180 night catches) areas, predominated on these rocky coasts and in inland areas. For C. obsoletus and C. pulicaris, the seasonal dynamics of adult density show a bimodal pattern with two distinct peaks: one in spring and the other in autumn. During the summer, especially August, the elevated temperatures and weak hygrometry could be unfavourable to the larval development and/or adult active flight of these European species. Mainland France From April to November 2002, Culicoides belonging to 44 species were collected in a total of 109 night catches. No specimens of C. imicola were found (Table II). The more abundant and widespread species were C. newsteadi (73.5%, and 109), C. obsoletus (8%, 201 and 12), C. scoticus (5.4%, 417 and 8), C. circumscriptus (3.2%, 177 and 5) and C. griseidorsum (2.7%, 337 and 4). The remaining 39 species accounted for less than 2% of the total. Results are similar to those observed in Corsica but C. circumscriptus 526 (3.2) C. griseidorsum 430 (2.7) C. pulicaris 188 (1.2) C. lupicaris 178 (1.1) C. submaritimus 173 (1.1) Culicoides spp. (36 species)* 637 (3.9) 5.8 Total Culicoides (44 species) (100) * The 36 remaining species accounted for less than 2% of the total Discussion During this entomological surveillance programme, C. imicola proved to be widely represented in Corsica. Its north/south gradient is likely to be the result of recent colonisation from Sardinia. Nevertheless, the relative abundance of C. imicola in the north (Balagne, Cap) compared to higher areas inland and to the west coast, suggests that several factors (climate, soil, topography, host presence) specific to each area play a crucial role in its local establishment and development. In general, in the Mediterranean region, C. imicola is restricted to plain coastal habitats up to an altitude of 800 m (12). Adult densities generally reach a peak in late summer and early autumn. These findings correlate with BT epidemiology in the temperate regions of the Mediterranean (10). The presence at most sites of C. imicola for eight months, from May to December, confirms that C. imicola over-wintered and is now permanently established in Corsica with several generations of adults during the active season. Thus, despite the fact that no specimens were found on 228 Veterinaria Italiana, 40 (3), 2004

192 Average catch size Culicoides imicola Average catch size 1, Culicoides newsteadi Feb [2] Mar [18] Apr [18] May [18] Jun [18] Jul 18] Aug [17] [n] : number of catches per month Sep [17] Oct [20] Nov [12] Dec [22] Feb [2] Mar [18] Apr [18] May [18] Jun [18] Jul 18] Aug [17] [n] : number of catches per month Sep [17] Oct [20] Nov [12] Dec [22] Average catch size Culicoides obsoletus 38 Feb [2] 0 Mar [18] 12 Apr [18] May [18] Jun [18] 34 Jul 18] 17 Aug [17] 25 Sep [17] 114 Oct [20] 77 Nov [12] [n] : number of catches per month [n] : number of catches per month Figure 3 Seasonal distribution of Culicoides imicola, C. newsteadi, C. obsoletus and C. pulicaris in Corsica, 2002 NB: scales differ, depending on species 39 Dec [22] Average catch size Culicoides pulicaris 0 2 Feb [2] Mar [18] Apr [18] May [18] 25 Jun [18] 15 Jul 18] 28 Aug [17] 57 Sep [17] 20 Oct [20] 15 Nov [12] 1 Dec [22] France. Spatial distribution is linked to specific bioecology, such as humid coastal lowlands for C. newsteadi and hilly rocky areas for C. pulicaris. Peak catches occurred in the spring and autumn; for C. obsoletus and C. pulicaris densities were higher in the autumn, and for C. newsteadi in the early summer. These spatial and seasonal patterns are similar to those observed in other countries of the Mediterranean Basin (3, 4, 11, 14). Culicoides newsteadi Culicoides obsoletus Culicoides pulicaris Culicoides spp. Figures: average catch of Culicoides per site Figure 4 Spatial distribution of Culicoides collected on mainland France, 2002 (19 sites/8 months/109 night catches) mainland France in 2002, the risk of invasion through air streams does exist. Other species of interest that are widely distributed in Corsica are found in lower densities on mainland C. obsoletus, the potential vector of BT in Europe, is close to C. imicola in terms of systematics (both belong to the subgenus Avaritia) and bio-ecology (both are commonly found in livestock-rearing environments). The abrupt decrease of C. obsoletus in summer may be due to the hot and dry conditions that would, in turn, favour the development of C. imicola, a tropical species more adapted to this kind of environment. Global warming could facilitate the colonisation of northern territories by C. imicola. Models based on an increase of 2 C in temperature showed that most of southern Europe and mainland France were susceptible to colonisation by C. imicola (17). Veterinaria Italiana, 40 (3),

193 Average catch size Culicoides imicola Average catch size Culicoides newsteadi Apr [2] May [17] Jun [15] Jul [17] Aug [10] Sep [21] [n] : number of catches per month Oct [25] Nov [2] Apr [2] May [17] Jun [15] Jul [17] Aug [10] [n] : number of catches per month Sep [21] Oct [25] Nov [2] Average catch size Culicoides obsoletus Average catch size Culicoides pulicaris Apr [2] May [17] Jun [15] Jul [17] Aug [10] Sep [21] [n] : number of catches per month Oct [25] Nov [2] Apr [2] May [17] Jun [15] Jul [17] Aug [10] Sep [21] [n] : number of catches per month Oct [25] Nov [2] Figure 5 Seasonal distribution of Culicoides imicola, C. newsteadi, C. obsoletus and C. pulicaris on mainland France, 2002 NB: scales differ, depending on species Entomological surveillance of BT in France was pursued after This long-term follow-up appears essential to ensure a better understanding of the epidemiology of BT, and the creation of models based on environmental and bio-ecology patterns that might help predict and identify areas at risk to introduction of the disease. Acknowledgements This work was funded by National Food Directorate, France (Direction générale de l Alimentation)/French Ministry of Agriculture and conducted jointly by the CIRAD-EMVT, the Université Louis Pasteur de Strasbourg, Interdepartmental group for mosquito control from the Mediterranean coast (Entente interdépartementale pour la démoustication du littoral méditerranéen (EID- Méditerranée), Departmental Directorate of Veterinary Services, France (Direction départementale des services veterinaires) (DDSV) Haute-Corse and DDSV Corse du Sud. References 1. Borkent A. & Wirth W.W. (1997). World species of biting midges (Diptera: Ceratopogonidae). Bull. Am. Mus. Nat. Hist., 233, 257 pp. 2. Braverman Y. & Chechik F. (1996). Air streams and the introduction of animal diseases borne on Culicoides (Diptera: Ceratopogonidae) into Israel. Rev. Sci. Tech. Off. Int. Épiz., 15, Calistri P., Goffredo M., Caporale V. & Meiswinkel R. (2003). The distribution of Culicoides imicola in Italy: application and evaluation of current Mediterranean models based on climate. J. Vet. Med., 50, Capela R., Purse B.V., Pena I., Wittmann E.J., Margarita Y., Capela M., Romao L., Mellor P.S. & Baylis M. (2003). Spatial distribution of Culicoides species in Portugal in relation to the transmission of African horse sickness and bluetongue viruses. Med. Vet. Entomol., 17, Caracappa S., Torina A., Guercio A., Vitale F., Calabro A., Purpari G., Ferrantelli V., Vitale M. & Mellor P.S. (2003). Identification of a novel bluetongue virus vector species of Culicoides in Sicily. Vet. Rec., 153, Veterinaria Italiana, 40 (3), 2004

194 6. Delécolle J.-C. (1985). Nouvelle contribution à l étude systematique et iconographique des espèces du genre Culicoides (Diptera: Ceratopogonidae) du nord-est de la France. Thesis, Université Louis Pasteur de Strasbourg, UER Sciences, Vie et Terre, 238 pp. 7. Delécolle J.-C. & de La Rocque S. (2002). Contribution à l étude des Culicoides de Corse. Liste des espèces recensées en 2000/2001 et redescription du principal vecteur de la fièvre catarrhale ovine C. imicola Kieffer, 1913 (Diptera: Ceratopogonidae). Bull. Soc. Entomol. Fr., 107, Linton Y.M., Mordue Luntz A.J., Cruickshank R.H., Meiswinkel R., Mellor P.S. & Dallas J.F. (2002). Phylogenetic analysis of the mitochondrial cytochrome oxidase subunit I gene of five species of the Culicoides imicola species complex. Med. Vet. Entomol., 16, Mellor P.S., Boorman J. & Baylis M. (2000). Culicoides biting midges: their role as arbovirus vectors. Ann. Rev. Entomol., 45, Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean Basin Vet. J., 164, Miranda M.A., Borras D., Rincon C. & Alemany A. (2003). Presence in the Balearic islands (Spain) of the midges Culicoides imicola and Culicoides obsoletus group. Med. Vet. Entomol., 17 (1), Rawlings P., Pro M.J., Pena I., Ortega M.D. & Capela R. (1997). Spatial and seasonal distribution of Culicoides imicola in Iberia in relation to the transmission of African horse sickness virus. Med. Vet. Entomol., 11 (1), Tabachnick W.J., Robertson M.A. & Murphy K.E. (1996). Culicoides variipennis and bluetongue disease. Ann. NY Acad. Sci., 791, Sarto i Monteys V. & Saiz-Ardanaz M. (2003). Culicoides midges in Catalonia (Spain), with special reference to likely bluetongue virus vectors. Med. Vet. Entomol., 17, Savini G., Goffredo M., Monaco F., de Santis P. & Meiswinkel R. (2003). Transmission of bluetongue virus in Italy. Vet. Rec., 152, Wirth W.W. & Marston N. (1968). A method for mounting small insects on microscope slides in Canada balsam. Ann. Entomol. Soc. Am., 61, Wittmann E.J., Mellor P.S. & Baylis M. (2001). Using climate data to map the potential distribution of Culicoides imicola (Diptera: Ceratopogonidae) in Europe. Rev. Sci. Tech. Off. Int. Épiz., 20, Zientara S., de La Rocque S., Gourreau J.M., Gregory M., Diallo A., Hendricks P., Libeau G., Sailleau C. & Delécolle J.-C. (2000). La fièvre catarrhale ovine en Corse en Epidémiol. Santé Anim., 38, Zientara S., Grillet C., de La Rocque S., Gourreau J.M., Gregory M., Diallo A., Hendricks P., Libeau G., Sailleau C., Albina E., Bréard E. & Delécolle J.-C. (2001). La fièvre catarrhale ovine en Corse en Epidémiol. Santé Anim., 40, Zientara S., Sailleau C., Dauphin G., Roquier C., Remond E.M., Lebreton F., Hammami S., Dubois E., Agier C., Merle G. & Bréard E. (2002). Identification of bluetongue virus serotype 2 (Corsican strain) by reverse-transcriptase PCR reaction analysis of segment 2 of the genome. Vet. Rec., 150, Veterinaria Italiana, 40 (3),

195 Vet. Ital., 40 (3), Culicoides imicola in Greece M.J. Patakakis Department of Parasitology, Centre of Athens Veterinary Institutes, 25 Neapoleos Street, Athens, Greece Summary Culicoides imicola, the major vector of bluetongue virus in Africa and the Middle East, was recorded in Greece for the first time in 1982 following an outbreak of the disease on the island of Lesbos (October 1979). Since then, many hundreds of Culicoides trappings have been made and thousands of Culicoides have been collected from the islands and from mainland Greece. Culicoides imicola is now present on most of the eastern Aegean islands and in northern, central and south-eastern mainland Greece. Keywords Bluetongue Culicoides imicola Greece Vector. Introduction In October 1979, an outbreak of bluetongue (BT) caused by BT virus (BTV) serotype 4 was reported on the Greek island of Lesbos. The disease affected mainly the eastern part of the island and caused considerable economic prejudice to the farmers. Furthermore, the incident created a new field of investigation for many scientists within the Greek Veterinary Service (3, 5). John Boorman was the first scientist who studied Culicoides collected from Lesbos and identified at least 17 species, among which was C. imicola (2) (Fig. 1). This was the first record of C. imicola in Greece. identified. (Fig. 3). Since 1985, many hundreds of Culicoides trappings have been made and thousands of insects have been sorted in the Parasitology Department in Athens. A comprehensive account of this work is presented here, with emphasis on the distribution of C. imicola in the country. Figure 2 Two Culicoides imicola trappings on the island of Rhodes in 1983 (1) 5 sites, 4 nights 11 catches Figure 1 Culicoides imicola trappings on the island of Lesbos in 1982 (2) In October 1984, C. imicola was collected on the island of Rhodes (1) (Fig. 2). At the same time, Mellor et al. collected Culicoides on mainland Greece (4) but did not find C. imicola among the 20 species Materials and methods Insects were initially collected with Monks Wood traps and with Pirbright traps of similar design. Since 1999, more durable and efficient traps of similar design made in South Africa have been used. Insects were collected in a weak solution of detergent and preserved in a 5% formalin solution and/or 70% alcohol. 232 Veterinaria Italiana, 40 (3), 2004

196 Light traps were operated from September 1991 to December 1992 on a weekly basis and more than 260 catches were made, sorted and identified in the Athens Laboratory. Culicoides imicola was found only on the island of Chios, near Lesbos and Rhodes, where this species was recorded previously (eastern Aegean). Culicoides imicola was observed in April, May, July, August, September, October and November, with a peak population in autumn. No C. imicola was found on mainland Greece. Figure 3 Culicoides trappings on mainland Greece in 1983 (4) Identification was made by comparing specimens with collections of the Parasitology Department in Athens or with those provided by the Institute for Animal Health in Pirbright or those from early catches by J.P.T. Boorman and P.S. Mellor. Results and discussion From 1985 to 1990, light traps operated in four locations, namely: Mantamados, Pelopi, Kalloni and Agia Paraskevi, with sorting intervals throughout the year. A new record for Greece was C. parroti, and C. imicola was present in winter catches, although in very low numbers (1, 2 or 3 specimens). In 1990, a survey was conducted to detect BT vectors in Greece. This was sponsored by the European Commission (VET/AH/4), with the assistance of P.S. Mellor and co-ordinated by O. Papadopoulos of the Faculty of Veterinary Medicine, Aristotle University in Thessaloniki. Collection sites are shown in Figure 4. Between 1993 and 1998, Culicoides were collected occasionally in many parts of the country to monitor the population. In September 1998, an outbreak of BT was recorded in the north-western part of the island of Rhodes and rapidly spread across the island. Trappings were made immediately and large numbers of Culicoides imicola were caught. Traps were operated twice a week through the winter into 1999 and C. imicola was found to be present continuously. In 1999, BTV was reported from the islands of Simi, Kos, Samos and Lesbos and finally an outbreak occurred in northern Greece (Halkidiki) in late September and in central Greece (Larisa, Omolio) in early October. Trappings of Culicoides commenced and C. imicola was present on the islands and, for the first time, on mainland Greece (Halkidiki, Omolio). At the same time, a large research project (Contract No. QLK ), was approved by the European Union; many Mediterranean countries participated in this project co-ordinated by P.S. Mellor and M. Baylis from Pirbright. The project commenced in 2000 and is due to end in As part of the project, more than 100 collection sites across Greece were established and additional sites are planned (Fig. 5). Figure 4 Culicoides trappings on mainland Greece from 1991 to 1992 (Contract VET/AH/4) Figure 5 Sites sampled for Culicoides in Greece during the summers of As shown in Figure 6, many other C. imicola sites were discovered on mainland Greece and the project Veterinaria Italiana, 40 (3),

197 is progressing. In 1992, no C. imicola was detected on mainland Greece or on the island of Crete. As the situation appears today, the future does not seem promising. References 1. Boorman J.P.T. (1986). Presence of bluetongue virus vectors on Rhodes. Vet. Rec., 118, Boorman J.P.T. & Wilkinson P.J. (1983). Potential vectors of bluetongue in Lesbos, Greece. Vet. Rec., 113, Mastroyanni M., Axiotis I. & Stoforos E. (1981). Study of the first outbreak of bluetongue disease in sheep in Greece. Bull. Hell. Vet. Med. Soc., 32, Mellor P.S., Jennings M. & Boorman J.P.T. (1984). Culicoides from Greece in relation to the spread of bluetongue virus. Rev. Elev. Med. Vet. Pays Trop., 37 (3), Vassalos M. (1980). A case of bluetongue on the Island of Lesbos (Greece). Bull. Off. Int. Épiz., 92, Figure 6 Distribution of Culicoides imicola in Greece, July 2003 Acknowledgements The author would like to thank his colleagues and veterinary surgeons across Greece for their valued and constant help during all these years of Culicoides trapping. Finally, grateful thanks are extended to J.P.T. Boorman and especially P.S. Mellor who gave me more than knowledge; they provided the ability to enjoy working on Culicoides and I feel deeply obliged to them. 234 Veterinaria Italiana, 40 (3), 2004

198 Vet. Ital., 40 (3), Epidemiology and vectors What factors determine when epidemics occur in the Mediterranean? Prediction of disease risk through time by climate-driven models of the temporal distribution of outbreaks in Israel Y. Braverman (1), M. Baylis (2), A.J. Tatem (3), D.J. Rogers (3), P.S. Mellor (2) & B.V. Purse (2) (1) Kimron Veterinary Institute, PO Box 12, Beit Dagan 50250, Israel (2) Institute for Animal Health, Ash Road, Pirbright, Surrey GU24 0NF, United Kingdom (3) TALA Research Group, Department of Zoology, University of Oxford, South Parks Road, Oxford OX1 3PS, United Kingdom Summary Determination of the temporal relationships between climate and epidemics of Culicoides-borne viral disease may lead to control measures and surveillance being implemented earlier and more efficiently. Although Israel has reported few cases of bluetongue (BT) during the recent Mediterranean epidemic, outbreaks have occurred almost annually since the disease was first confirmed there (1950) with severe episodes occurring periodically. The south Mediterranean location and intensive farming of BT-susceptible European sheep breeds make the area ideal for investigation of the effect or role of climatic factors versus other potential host or virus factors in governing the timing of severe BT episodes. The authors present regression analyses of 20-year time-series of BT outbreaks versus four remotely sensed climatic variables. Low temperatures and high moisture levels (relative to average levels) in the preceding autumn coincident with the seasonal peak of vector abundance and outbreaks had a positive effect on the number of outbreaks the following year. The positive effects of high moisture levels are postulated to increase breeding site availability and refugia for adult C. imicola vectors (from desiccation) in autumn whilst low temperatures may increase fecundity, offspring size and survival through adulthood in winter by increasing initial vector population size the following year. The proportion of variance in the annual BT outbreak time series accounted for by climate factors was relatively low (approximately 20%), probably because most BT virus (BTV) circulation occurs silently, due to the circulation of non-virulent BTV strains, combined with the prevalence of relatively resistant local sheep breeds. Thus, the level of BTV transmission is poorly correlated with the rate of outbreak notification. Keywords Bluetongue Climate Culicoides imicola Epidemic Israel Mediterranean Remote-sensing Time-series analysis. Introduction The recent unprecedented epidemic of bluetongue (BT) in the Mediterranean Basin has led to a spate of studies on the spatial distribution of vectors and virus in this region (5, 15, 17, 18, 42). However, studies of the temporal distribution of Culicoides and BT have been limited to restricted areas or short time-scales (46, 48). Climatic factors may influence the temporal distribution of BT outbreaks via their effects on the life-history parameters and breeding sites of Culicoides vectors (29), the main Mediterranean vector species being C. imicola Kieffer, In regard to African horse sickness, another Culicoides-borne disease, a strong association was noted between the timing of epidemics during the last century and the warm phase of the El Niño/Southern Oscillation (ENSO) in South Africa due to the combination of rainfall Veterinaria Italiana, 40 (3),

199 and drought brought by this phase (4). In both Israel and Cyprus, severe BT outbreaks were preceded by higher than average rainfall in winter or autumn (33, 34, 39, 40) that increases the availability of suitable breeding sites for C. imicola. This species breeds in rich mixtures of organic matter and wet soil (without surface water) (8). Other potential factors that could influence the temporal distribution of BT outbreaks include the occurrence of viral incursion in surrounding countries (Fig. 1), the introduction of new viral strains (20) and the variation in presence of susceptible hosts (and levels of herd immunity). Sellers (37) was the first to note the simultaneous appearance of BT outbreaks in southern Mediterranean countries in 1943 to 1944 (Cyprus, Turkey, Syria and Israel), in 1950 to 1951 (Cyprus and Israel) and in 1964 to 1965 (Cyprus, Israel and Egypt) (Fig. 1). This observation has since been developed further by Hassan (24) and Taylor (44, 45) for later outbreaks. The rate at which outbreaks occur or are notified may depend on the turnover, annual vaccination coverage (40) or importation of susceptible sheep breeds. In the 1960s, outbreaks were preceded in both Israel (39) and Egypt (1, 22) by the importation of German Merino breeds, to which cases were largely restricted, whilst local African sheep breeds (such as Awassi) were relatively unaffected (26). Figure 1 Countries of the south-eastern Mediterranean Basin, showing the location of bluetongue outbreaks in recent severe episodes in Israel, 1987, 1993, 1994, 1996 (inset) The authors present an investigation of whether the timing of severe BT outbreaks is attributable to climatic factors in Israel compared to other potential host or virus factors. The authors analyse relationships between a continuous 20-year monthly bluetongue incidence data-set and monthly climatic variables derived from remotely-sensed advanced very high resolution radiometer (AVHRR) data (8 km spatial resolution). Material and methods Study area Israel includes areas of relatively moist temperate climate in the north (coastal plains), cool central mountain ranges and an arid desert area in the south (Negev). The annual climate can be divided into a rainy season between October and April (peak rainfall and minimum temperatures are December to February) and a dry season from May to August (usually to October). The livestock population of Israel usually comprises approximately bovines, goats and ovines. In the past, approximately 25% of the sheep population were exotic breeds and their crosses (40). In the Jewish sector, from which much of the BT outbreak data is derived, the sheep population is currently made up of Awassi, Merino and Assaf (East Friesian Awassi). Since 1964, the latter are vaccinated annually using a polyvalent vaccine from the Onderstepoort Veterinary Institute (containing live attenuated BTV types 2, 4, 6, 10 and 16) changing in 1974 to a quadrivalent vaccine (types 2, 4, 6 and 10). The spatial distribution of 116 villages affected by BT during recent severe outbreaks is shown in the inset of Figure 1 (M. Van Ham, unpublished data). This village distribution (point data) was overlaid on the 8 km 8 km pixels of Israel to identify pixels within 4 km of an outbreak (outbreak pixels). Incidence data, vector and remotely-sensed climatic time-series data and data analysis Sources, time period covered and form of entomological, epidemiological and climate time series data available for analyses are summarised in Table I. The environmental significance of the remotely-sensed climate variables is as follows: normalised difference vegetation index (NDVI) specifically measures chlorophyll abundance and light absorption, but is correlated with soil moisture, rainfall and vegetation biomass, coverage and productivity (16). Middle infa-red reflectance (MIR) is correlated with the water content, surface temperature and structure of vegetation canopies (6). Land surface temperature (LST) is a general index of the apparent environmental surface temperature (whether soil or vegetation) and air temperature 236 Veterinaria Italiana, 40 (3), 2004

200 Table I Sources, period and form of entomological, epidemiological and climate time-series data available for analysis Data type Period and continuity Variable form Bluetongue incidence Monthly numbers of BT outbreaks in sheep flocks in Israel Annual total of outbreaks Annual duration of the outbreaks (months) Proportion of annual total of outbreaks contained in each month Vector numbers July-December Climate (remotelysensed data) Climate (weather station data) Daily maximum trap catch of C. imicola across all traps Monthly index of abundance i.e. monthly average across daily maximum trap catches Mean of proportion of annual vector population contained in each month Monthly middle infra-red reflectance (MIR) Monthly land surface temperature (LST) Monthly air temperature (TAIR) Monthly were normalised difference vegetation index (NDVI) All variables were averaged across outbreak pixels Monthly minimum and maximum temperature (Jan-Dec) Average daily minimum and maximum temperatures (Jan-Dec) Monthly rainfall quantities (Sept-May) Source Israeli Veterinary Services Du Toit light-trap collections from Beit Dagan (32º05 N, 34º50 E) (9, 11, 12) Pathfinder advanced very high resolution radiometer (AVHRR) imagery at 8 by 8 km resolution (36) Kefar Blum (33º09 N, 35º38 E) weather station (TAIR) is an estimate of the air temperature a few metres above the land surface (21). Data analysis Strong seasonal variation was removed from the monthly climate time series by seasonal decomposition (19) in MINITAB release An additive model of the type X t = mt + S t + ε t was used for seasonal decomposition (where mt is the deseasonalised mean level at time t, S t is the seasonal effect at time t, and ε t is the random error), to make the seasonal effect constant from year to year. Since seasonal decomposition was unsuccessful for the monthly BT outbreak time series, analysis was based on raw monthly and annual totals of BT outbreaks. Relationship between outbreaks and climate variables Cross-correlation functions (CCF) were calculated between the monthly totals of BT outbreaks and the deseasonalised climate variables (one CCF was calculated from and one from ). On an annual time-scale, a linear regression was calculated between the annual total of BT outbreaks (log-transformed), the year (to consider the linear temporal trend) and 40 independent climatic variables. These were, for each of the deseasonalised TAIR, NDVI, LST and MIR: annual mean, annual minimum, annual maximum, annual amplitude, mean of the variable across months within each quarter of the same year and within the last two quarters of the previous year (the quarters being January to March, April to June, July to September and October to December). Variables significant in univariate regressions were then included in a global model and the best one (or two) variable model was chosen by best subsets regression. This process was repeated with the duration of outbreaks as the dependent variable. CCFs were calculated between deseasonalised satellite-derived variables and weather station-derived variables for Kefar Blum. A Spearman s rank correlation was calculated between monthly rainfall amounts (only available for September to May for each year) and satellite-derived variables. Results Temporal patterns in an outbreak, vector and climatic time series A total of 386 outbreaks of BT were recorded between 1968 and 2001 in Israel. They occurred almost annually (Fig. 2) (range: 0 to 60 outbreaks a year) and only 6 of 34 years (18%) had no outbreaks. However, 50% of years had five outbreaks or fewer. More than 20 outbreaks occurred in 1969, 1975, 1987, 1988, 1991 and 1994 (18% of years). The annual duration of the outbreaks ranged from one to six months (mean duration ± s.e = 3.14 ± 2.84) and there was a significant positive correlation between the annual total of outbreaks and their duration (Spearman s rank correlation r s = 0.76, p<0.001, n = 28). Veterinaria Italiana, 40 (3),

201 No. of bluetongue cases Year Figure 2 Annual time-series of bluetongue cases and livestock numbers (1 000s per decade) Outbreaks only occurred between July and January in the year but were concentrated in October and November when 71% of all outbreaks were detected with 21% detected in August and September (Fig. 3). The seasonal distribution of outbreaks is mirrored by that of the vector C. imicola populations. Mean (%) of annual total contained each month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month Sum of outbreaks per month Figure 3 Seasonal distribution of bluetongue outbreaks and vector Culicoides imicola populations Closed bars: mean (± s.e.) proportion of the annual total of outbreaks contained in each month over six years with >20 outbreaks Open bars: mean (± s.e.) proportion of the annual total of C. imicola contained in each month in 2001 and 2001 at Beit Dagan Closed circles: sum of outbreaks contained in each month over all years ( ) No. of sheep and goats ('000) per decade Relationship between outbreaks and climatic variables Between 1981 and 1994, monthly BT outbreak numbers were negatively correlated with MIR (r = 0.26), TAIR (r = 0.26) and LST (r = 0.28) at a lag of 1 month. Between 1995 and 1999, monthly BT outbreak numbers were negatively correlated with MIR (r = 0.13), TAIR (r = 0.35) and LST (r = 0.23) at a lag of five months and positively correlated with NDVI in the same month (r = 0.31) and at a lag of four months (r = 0.26). These results suggest that outbreaks are less likely in a month when preceded by periods of high temperature (indicated by high LST, MIR or TAIR) one to five months earlier. Outbreaks are more likely to occur when moist conditions prevail (indicated by high NDVI) in a particular month or up to four months previously. There was no significant relationship between the annual total of outbreaks and year (F 1, 17 = 0.05, p = 0.82, adjusted R 2 = 0.0) or duration of outbreaks and year (F 1, 17 = 0.03, p = 0.88, adjusted R 2 = 0.0) indicating that there was no linear temporal trend. The annual total of outbreaks was negatively related to mean LST (equation; y = *mean LST; F 1, 17 = 4.7, p = 0.047, adjusted R 2 = 18.7) and mean MIR (equation; y = *mean MIR: F 1, 17 = 5.6, p = 0.032, adjusted R 2 = 22.2) in the last quarter of the previous year (October to December). It was unrelated to all other independent variables. A regression model including both variables did not sufficiently increase the amount of variance explained by the predictor variables to justify the inclusion of LST (equation; y = *mean MIR *mean LST: F 1, 17 = 2.6, p = 0.109, adjusted R 2 = 16.8). Thus, mean MIR in the last quarter of the previous year was the predictor of the annual total number of outbreaks (i.e. the most parsimonious model). Figure 4 shows observed time series of outbreaks compared to that derived from the fitted values from this relationship. The observation for the 1994 outbreak year had a large influence in these regressions (equation omitting 1994; y = mean MIR: F 1, 15 = 2.1, p = 0.169, adjusted R 2 = 6.5). The duration of outbreaks was negatively related to mean LST in the last quarter of the previous year (equation; y = *mean LST; F 1, 15 = 6.5, p = 0.023, adjusted R 2 = 25.4) and 1994 again had a large influence on this relationship (equation omitting 1994; y = mean LST: F 1, 14 = 3.1, p = 0.10, adjusted R 2 = 12.3). 238 Veterinaria Italiana, 40 (3), 2004

202 Log (annual outbreak total) Year Annual total of outbreaks of bluetongue from 1982 to 1999 Fitted regression model Corresponding upper and low er limits of the 95% confidence interval around the fitted regression model Figure 4 Results of the regression model (y = *mean MIR) Relationship between satellite-derived and weather station derived climate variables for Kefar Blum Monthly minimum temperatures were uncorrelated with satellite-derived variables whist monthly maximum temperatures were positively correlated with TAIR (r = 0.32), LST (r = 0.26) and MIR (r = 0.30) in the same month. Since three images are maximum-composited to produce the monthly satellite variables, one would not expect them to be correlated with monthly minimum temperatures. Average daily minimum temperatures were negatively correlated with the NDVI in the same month (r = 0.26) and positively correlated with MIR (r = 0.29) and LST (r = 0.28) in the same month. Average daily maximum temperatures were negatively correlated with the NDVI one month later (r = 0.24) and positively correlated with TAIR (r = 0.39), MIR (r = 0.33) and LST (r = 0.36) in the same month. Monthly rainfall amount was correlated with MIR (Spearman s rank r s = 0.17, p = 0.03, n = 162) and LST (Spearman s rank r s = 0.18, p = 0.02, n = 162) in the same month. Discussion Bluetongue outbreaks are significantly related to satellite-derived climate variables in Israel with some degree of delay. This suggests potential for a climatebased early warning system for BT, contingent on further detailed analyses of these relationships. On a monthly basis, BT outbreak numbers decreased with high temperatures (TAIR and LST) and MIR in the preceding months (one to five months before) and increased with high NDVI in the same and preceding months (four months before). On an annual basis, total outbreaks decreased with high MIR in the quarter at the end of the previous year (October to December). The duration of outbreaks the following year was reduced by warm temperatures (LST) in the same period. Thus low temperatures (indicated by low LST, TAIR and MIR) and high moisture levels (indicated by low MIR and high NDVI) have a positive effect on the number of outbreaks in Israel. Conditions in the preceding year, between October and December, appeared to be more important than spring or early summer conditions in the same year. Previous spatial satellite-driven models of abundance of the vector C. imicola have found positive effects of high moisture levels, i.e. high or early peaks in NDVI (2, 3, 5, 42). Short-term temporal studies found increases in this species abundance following rainfall (31, 46). In addition, severe BT outbreaks were observed to be preceded by higher than average rainfall in autumn or winter (33, 34, 39, 40). Due to its location in the southern Mediterranean, Israel has a mixture of warm, temperate and arid desert climates. Temperature conditions will generally be far from the lower limits required for development of C. imicola vectors whilst moist conditions may generally be close to the lower limits of the requirements for this species. This is consistent with the restriction of BT outbreaks to northern lowland areas and river valleys in Israel (that probably broadly reflect the distribution of C. imicola abundance). The number of outbreaks the following year may depend on the initial size of the vector population after the winter period and the transmission intensity both during the previous autumn peak of vector and virus populations and over winter. High moisture levels during the peak of vector abundance will increase the availability and quality of breeding sites (wet soil and organic matter) (8) and provide refugia where adults can resist desiccation (30). Low temperatures during autumn and winter may have positive effects on fecundity, offspring size (27) and survival through adulthood (25, 47) and so increase initial adult population size Veterinaria Italiana, 40 (3),

203 the following year. These mechanisms may produce the positive annual relationship between outbreaks and low temperature and high moisture in late autumn and winter observed in Israel. The opposite relationship may arise further north in Europe, at the northern range limit of C. imicola. Here, proximity to the lower temperature limit for development within moist climates produces positive effects of high temperature and relatively dry summer conditions on C. imicola abundance (32, 35). Braverman et al. (14) point out that no BT outbreaks occur after very cold winters such as that of Indeed, for 1992, our satellite-driven model over-estimates the likelihood of outbreaks (Fig. 4). Since satellite variables are maximum-composited, extreme climatic minima may not be represented in their time-series. Thus the effect of very low temperatures, for example, on numbers of outbreaks may not be testable. The proportion of variance in the epidemiological time-series accounted for by climate factors is relatively low (approximately 20%) (Fig. 4). The rate at which BT outbreaks are notified is unlikely to be proportional to the actual level of virus circulation for two reasons. Firstly, a large proportion of the sheep population has consisted of relatively resistant breeds in the past (40) in which BTV circulation occurs without any detectable clinical signs. The sheep population currently consists mainly of Assaf in which less severe BT cases (relative to those in Merinos) are usually observed (41). Thus, long-term variation in the importation, movements and vaccination of such sheep within the country will have affected the rate of BT outbreak notification. In addition, there is evidence of high rates of seroconversion to BTV in sentinel cattle in several years without high numbers of outbreaks, suggesting that many strains of BTV are non-virulent (41). For example, in 1983 and 1984, where the satellite-driven model over-estimates the number of BT outbreaks, there was a high rate of seroconversion to serotypes 2, 4 and 6 across Israel. In the absence of annual, standardised, sentinel surveillance data, a strong relationship between climate and virus circulation is not easily detectable with such a data set. Other factors, such as the occurrence of a viral incursion from surrounding countries, may affect the temporal distribution of outbreaks. Most authors in surrounding countries assert that BTV circulates continually undetected in resistant livestock (23, 37, 44) such that sources of viral incursion should frequently be available. However, only five (1, 2, 4, 13, 16) of the ten different BTV serotypes recorded in adjacent countries (1, 2, 3, 4, 6, 9, 11, 13, 14, 16) have been detected in Israel (39). This, together with the considerable temporal variation in serotype prevalence in Israel, suggests that occurrence of incursions from surrounding countries, permitted probably by wind dispersal of infected midges, plays some role in the dynamics of BTV in Israel. As such, temporal variation in wind speed and direction should be considered in future models (7, 10, 13, 28, 33, 38, 44), although these factors are also inherently difficult to quantify. An examination of BT case data from the current epidemic in relation to a range of moderate resolution imaging spectroradiometer (MODIS) imagery (43) at shorter time-scales may reveal stronger relationships between climate and the timing of outbreaks that are more useful for outbreak prediction. References 1. Barsoum G.W. (1992). Bluetongue and African horse sickness situation in Egypt. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Baylis M., Bouayoune H., Touti J. & El Hasnaoui H. (1998). Use of climatic data and satellite imagery to model the abundance of Culicoides imicola, the vector of African horse sickness virus, in Morocco. Med. Vet. Entomol., 12, Baylis M., Meiswinkel R. & Venter G.J. (1999). A preliminary attempt to use climate data and satellite imagery to model the abundance and distribution of Culicoides imicola (Diptera: Ceratopogonidae) in southern Africa. J. Sth Afr. Vet. Assoc.,70, Baylis M., Mellor P.S. & Meiswinkel R. (1999). Horse sickness and ENSO in South Africa. Nature, 397, Baylis M., Mellor P.S., Wittmann E.J. & Rogers D.J. (2001). Prediction of areas around the Mediterranean at risk of bluetongue by modelling the distribution of its vector using satellite imaging. Vet. Rec., 149, Boyd D.S. & Curran P.J. (1998). Using remote sensing to reduce uncertainties in the global carbon budget: the potential radiation acquired in the middle infrared wavelengths. Remote Sensing Rev., 16, Braverman Y. (1992). The possible introduction to Israel of Culicoides (Diptera, Ceratopogonidae) borne animal diseases by wind. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Braverman Y., Galun R. & Ziv M. (1974). Breeding sites of some Culicoides species (Diptera: Ceratopogonidae) in Israel. Mosquito News, 34, Braverman Y. & Linley J.R. (1988). Parity and voltinism of several Culicoides spp. (Diptera: 240 Veterinaria Italiana, 40 (3), 2004

204 Ceratopogonidae) in Israel, as determined by two trapping methods. J. Med. Entomol., 25, Braverman Y. & Chechik F. (1993). Air streams and their possible potential for the introduction of Culicoides (Diptera: Ceratopogonidae) borne animal diseases to Israel. Isr. J. Vet. Med., 48, Braverman Y. & Linley J.R. (1993). Effect of lighttrap height on catch of Culicoides (Diptera: Ceratopogonidae) in Israel. J. Med. Entomol., 30 (6), Braverman Y. & Linley J.R. (1994). Fecundity and proportions of gravid females in populations of the Bluetongue vector Culicoides imicola (Diptera: Ceratopogonidae) and several other species in Israel. J. Med. Entomol., 31, Braverman Y. & Chechik F. (1996). Air streams and the introduction of animal diseases borne on Culicoides (Diptera: Ceratopogonidae) into Israel. Rev. Sci. Tech. Off. Int. Épiz., 15 (3), Braverman Y., Chechik F. & Mullens B.A. (2001). The interaction between climatic factors and bluetongue outbreaks in Israel and the eastern Mediterranean and the feasibility of establishing bluetongue-free zones. Isr. J. Vet. Med., 56(3), Calistri, P., Goffredo M., Caporale V. & Meiswinkel R. (2003). The distribution of Culicoides imicola in Italy: application and evaluation of current Mediterranean models based on climate. J. Vet. Med., 50, Campbell J.B. (1996). Introduction to remote sensing. Taylor & Francis, London. 17. Capela R., Purse B.V., Pena I., Wittmann E.J., Margarita Y., Capela M., L. Romo, Mellor P.S. & Baylis M. (2003). Spatial distribution of Culicoides species in Portugal in relation to the transmission of African horse sickness and bluetongue viruses. Med. Vet. Entomol., 17 (2), Conte A., Giovannini A., Savini L., Goffredo M., Calistri P. & Meiswinkel R. (2003). The effect of climate on the presence of Culicoides imicola in Italy. J. Vet. Med., 50, Cryer J.D. (1986). Time series analysis. Boston, USA, Duxbury Press. 20. Gambles R.M. (1949). Bluetongue of sheep in Cyprus. J. Comp. Pathol., 59, Goetz S.J., Prince S.D. & Small J. (2000). Advances in satellite remote sensing of environmental variables for epidemiological applications. Remote sensing and geographical information systems in epidemiology (S.I. Hay, S.E. Randolph & D.J. Rogers, eds). Academic Press, San Diego, 47, Gouma I.S. & A. Stevens A.J. (1964). An outbreak of suspected BT in sheep in Egypt. J. Arab. Vet. Med. Assoc., 24, Hafez S.M., Pollis E.G. & Mustafa S.A. (1978). Serological evidence of the occurrence of bluetongue in Iraq. Trop. Anim. Hlth Prod., 10 (2), Hassan A. (1992). Status of bluetongue in the Middle East and Africa. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Hunt G.J., Tabachnick W.J. & McKinnon C.N. (1989). Environmental factors affecting mortality of adult Culicoides variipennis (Diptera: Ceratopogonidae) in the laboratory. J. Am. Mosquito Control Assoc., 5, Komarov A. & L. Goldsmit (1951). A disease, similar to bluetongie in cattle and sheep in Israel. Refuah Vet., 8, Linley J.R. & Hinds M.J. (1976). Seasonal changes in size, female fecundity and male potency in Culicoides melleus (Diptera: Ceratopogonidae). J. Med. Entomol., 13, Mastroyianni M. (1987). Bluetongue in Lesbos and Rhodes (Greece) in In Bluetongue in the Mediterranean. Proc. Commission of the European Communities Meeting in the community programme for co-ordination of agricultural research. Istituto Zooprofilatico Sperimentale dell Abruzzo e del Molise, Teramo, Italy. 29. Mellor P.S., Boorman J. & Baylis M. (2000). Culicoides biting midges: Their role as Arbovirus vectors. Ann. Rev. Entomol., 45, Murray M.D. (1991). The seasonal abundance of female biting-midges, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae), in coastal south-eastern Australia. Aust. J. Zool., 39, Nevill E.M. (1971). Cattle and Culicoides biting midges as possible overwintering hosts of bluetongue virus. Onderstepoort J. Vet. Res., 38, Ortega M.D., Mellor P.S., Rawlings P. & Pro M.J. (1998). The seasonal and geographical distribution of Culicoides imicola, C. pulicaris group and C. obsoletus biting midges in central and southern Spain. Arch. Virol. [Suppl.], 14, Pitzolis G. (1987). A review of bluetongue in Cyprus with evidence of the windborne spread of vectors. In Bluetongue in the Mediterranean. Proc. Commission of the European Communities Meeting in the community programme for co-ordination of agricultural research. Istituto Zooprofilatico Sperimentale dell Abruzzo e del Molise, Teramo, Italy. 34. Polydorou K. (1978). The 1977 outbreak of bluetongue in Cyprus. Trop. Anim. Hlth Prod., 10, Rawlings P., Capela R., Pro M.J., Ortega M.D., Pena I., Rubio C., Gasca A. & Mellor P.S. (1998). The relationship between climate and the distribution of Culicoides imicola in Iberia. Arch. Virol. [Suppl.], 14, Rogers D.J., Hay S.I., Packer M.J. & G.R.W. Wint M.J. (1997). Mapping landcover over large areas using multispectral data derived from NOAA- AVHRR: a case study of Nigeria. Int. J. Remote Sensing, 18 (15), Veterinaria Italiana, 40 (3),

205 37. Sellers R.F. (1975). Bluetongue in Cyprus. Aust. Vet. J., 51, Sellers R.F., Gibbs E.P.J., Herniman K.A.J., Pedgley D.E. & Tucker M.R. (1979). Possible origin of the bluetongue epidemic in Cyprus, August J. Hyg., Camb., 83, Shimshony A. (1964). An outbreak of bluetongue in sheep in the Ta anakh area of northern Israel in Refuah Vet., 21, Shimshony A. (1987). Bluetongue activity in Israel, The disease, virus prevalence, control methods. In Bluetongue in the Mediterranean. Proc. Commission of the European Communities Meeting in the community programme for co-ordination of agricultural research. Istituto Zooprofilatico Sperimentale dell Abruzzo e del Molise, Teramo, Italy. 41. Shimshony A., Barzilai E., Savir D. & Davidson M. (1988). Epidemiology and control of bluetongue disease in Israel. Rev. Sci. Tech. Off. Int. Épiz., 7 (2), Tatem A.J., Baylis M., Mellor P.S., Purse B.V., Capela R., Pena I. & Rogers D.J. (2003). Prediction of bluetongue vector distribution in Europe and North Africa using satellite imagery. Vet. Microbiol., 97 (1-2), Tatem A.J., Goetz S.J. & Hay S.I. (2004). Terra and aqua: new data for epidemiology and public health. Int. J. Appl. Earth Observ. Geoinfo. (submitted). 44. Taylor W.P. (1987). Bluetongue in Syria and Jordan. In Bluetongue in the Mediterranean. Proc. Commission of the European Communities Meeting in the community programme for co-ordination of agricultural research. Istituto Zooprofilatico Sperimentale dell Abruzzo e del Molise, Teramo, Italy. 45. Taylor W.P., Sellers R.F., Gumm I.D., Herniman K.A.J. & Owen L. (1985). Bluetongue epidemiology in the Middle East. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, Inc., New York, Walker A.R. & Davies F.G. (1971). A preliminary survey of the epidemiology of bluetongue in Kenya. J. Hyg., 69, Wellby M., Baylis M., Rawlings P. & Mellor P.S. (1996). Effect of temperature on survival and rate of virogenesis of African horse sickness virus in Culicoides variipennis sonorensis (Diptera: Ceratopogonidae) and its significance in relation to the epidemiology of the disease. Bull. Entomol. Res., 86, Wright J.C., Getz R.R., Powe T.A., Nusbaum K.E., Stringfellow D.A., Mullen G.R. & Lauerman L.H. (1993). Model based on weather variables to predict seroconversion to bluetongue virus in Alabama cattle. Prev. Vet. Med., 16, Veterinaria Italiana, 40 (3), 2004

206 Vet. Ital., 40 (3), Bluetongue in Italy: Part II A. Giovannini (1), P. Calistri (1), D. Nannini (1), C. Paladini (1), U. Santucci (2), C. Patta (3) & V. Caporale (1) (1) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy (2) Ministero della Salute, Direzione Generale della Sanità Pubblica Veterinaria, Alimenti e Nutrizione, Piazzale Marconi, Rome, Italy (3) Istituto Zooprofilattico Sperimentale della Sardegna G. Pegreffi, Via Duca degli Abruzzi N 8, Sassari, Italy Summary In summer 2000, bluetongue (BT) infection was reported in Italy and caused a widespread epidemic involving a total of ten southern and central regions and is still in progress after three years. From the date of the first case (18 August 2000) to 14 May 2001, when the lowpoint in the first epidemic curve was reached, a total of animals in flocks of three regions had been involved. From 15 May 2001 to 14 April 2002, when a second epidemic wave swept through central and southern Italy, a total of animals in flocks in seven regions were involved. During 2000 and 2001 virtually no susceptible ruminants were vaccinated. On 11 May 2001, the Italian Ministry of Health ordered the vaccination of all susceptible domestic ruminant species (i.e. sheep, goats, cattle and water buffalo) in the infected and surrounding areas. The vaccination strategy stemmed from a risk assessment that demonstrated the possibility of such a strategy preventing most of the direct economic losses and decreasing the level of virus circulation. Vaccination of the target populations commenced in January In July 2002, when the new epidemic peak was reached, the percentage of vaccinated populations varied between the regions with direct consequences on the spread of BT. The relationship between vaccination coverage of the target populations and animal losses due to disease and virus circulation, and as detected by the sentinel surveillance system, was analysed. The effectiveness of the vaccination campaign in limiting virus circulation and consequently indirect losses due to animal movement restrictions was analysed and evaluated. At the end of 2002, a second risk assessment led to the authorisation of the movement of vaccinated animals from infected areas (where at least 80% of the susceptible population was vaccinated) directly to slaughter in unvaccinated areas free from infection. This risk assessment also generated new criteria to define zones where animal movement restrictions should be applied. Following the second vaccination campaign (January to May 2003), a third risk assessment was performed and the results from vaccination trials performed in controlled and in field conditions studied. These studies indicated that procedures to move vaccinated breeding animals from zones where infection exists to unvaccinated infection free zones could be contemplated. Keywords Bluetongue Epidemiology Italy Surveillance Vaccination. Introduction In August 2000, bluetongue (BT) infection, due to serotype 2, was reported for the first time on mainland Italy, in the Balearic Islands (Spain), and in Corsica (France). By the autumn of 2000, a second serotype, BT virus (BTV) serotype 9, was observed in southern Italy. This incursion of BTV into the central Mediterranean region resulted in the largest epidemic of BT ever to affect Europe. Since 2000, BT has spread to a total of ten southern and central regions of Italy. In the first epidemic commencing on 18 August 2000 and ending on 14 May 2001, a total of animals in flocks in three regions were affected. From 15 May 2001 to 14 April 2002, a second epidemic wave swept through central and southern Italy, affecting a total of animals in flocks in seven regions (1). 252 Veterinaria Italiana, 40 (3), 2004

207 Following the first epidemic, a risk assessment indicated that it would be possible to prevent most direct economic losses by vaccinating sheep. Virus circulation could also be reduced if all domestic ruminant populations susceptible to BTV infection (sheep, goats and cattle) were vaccinated (4). Consequently, the Italian Ministry of Health ordered the vaccination of all domestic ruminants susceptible to BT infection (sheep, goats, cattle, buffaloes) in infected and in at-risk areas on 11 May 2001 (7). Despite the Order of the Ministry, virtually no ruminants were vaccinated during Vaccination of susceptible populations commenced in January 2002 and when the new seasonal epidemic started (July 2002), the level of vaccination in susceptible populations varied greatly in the different regions of Italy. Differences in vaccination coverage of susceptible domestic populations resulted in different levels of loss due to the disease and to different intensities of virus circulation and spread. The objectives of the study is to describe the third and fourth seasonal epidemics in Italy, after the implementation of a vaccination strategy, and to analyse the effectiveness of vaccination in reducing both the direct losses due to the disease and the indirect losses due to restrictions of movements of animals from infected areas. Methods Source of data The data on BT outbreaks, serological surveillance of sentinel animals, entomological surveillance, ad hoc surveys and vaccination activities recorded in the Bluetongue National Information System (6) are analysed here. Outbreaks When a case of BT was suspected to occur on a farm(s), all susceptible animals were examined for clinical signs by official veterinarians and blood samples collected and sent to the laboratory for confirmation of infection. Morbidity and mortality was monitored and recorded during weekly visits to affected farms. When BT was confirmed, official veterinarians extended visits to all ovine flocks within a radius of 20 km of the confirmed outbreak (when clinical disease was observed) or to all ovine flocks within a radius of 4 km (if there was subclinical infection). Serological surveillance of sentinel animals A nationwide sentinel network had already been implemented in January 2002 when the vaccination campaign commenced. Italy had been divided into a grid of square units of either 400 km 2 or km 2 depending on the occurrence, or the risk of occurrence, of infection. To detect at least 5% of positive animals with a 95% confidence level in each 400 km 2 unit, a sample of 58 bovine animals was selected from 5 to 8 farms. To detect at least 2% of positive animals with a 95% confidence level in each unit of km 2, a sample of 148 bovine animals was selected from 8 to 12 farms. If cattle were not present in the area, sheep were selected as sentinel animals. Blood samples from sentinels were collected regularly with a variable frequency dependent upon the season and infection occurrence in the area. Entomological surveillance Nationwide entomological surveillance was also implemented before vaccination started. Blacklight traps were positioned in fixed locations in each province and operated weekly to monitor Culicoides spp. population dynamics. Blacklight traps were also operated on a temporary basis in suspected or confirmed cases of virus circulation to identify the Culicoides spp. present. Vaccination To monitor the progress of the vaccination campaign, vaccine serotype(s), vaccine batch numbers, farm codes, total numbers of animals on the farm, numbers of eligible animals and numbers of vaccinated animals by species, and vaccination date were recorded by the Local Health Unit. Data were then sent to the Bluetongue National Information System where it was sorted by province and circulated on the internet. Sera from 35 randomly selected vaccinated animals per each grid cell were tested for antibody presence to monitor vaccination coverage. Antibody coverage in vaccinated populations was evaluated by serological examination of the sample collected. Possible undesirable vaccine side-effects (deaths, abortions and stillbirths), were monitored by: a) sampling animals in flocks where problems arose to assess the presence of the vaccine virus in dead animals and/or foetuses b) collecting information concerning the type and incidence of the loss observed, vaccine used, dates of vaccination, etc. Samples were submitted to the laboratory to perform differential diagnostic tests and for the identification of the BTV serotype involved (vaccine or wild-type virus). Veterinaria Italiana, 40 (3),

208 Analysis of data Efficacy of vaccination in reducing direct losses due to the disease in the regions affected was evaluated by comparing the percentage of the population vaccinated at the beginning of the new epidemic (July 2003) with the geographic distribution of the infection and the number of outbreaks during the 2002 epidemic. The correlation between the percentage of the population vaccinated by the end of July 2002 and the number of outbreaks in the epidemic was evaluated using Spearman s correlation coefficient (11). Massa (Tuscany) in September. The total number of outbreaks detected in the third epidemic was 432 in eight regions (Table I). The geographical distribution of the infection is presented in Figure 1. BTV-2 BTV-9 The effectiveness of vaccination in reducing indirect losses due to movement restrictions was evaluated in Sardinia. The number of animals exported from Sardinia to free areas in continental Italy in 2001 (before the commencement of the vaccination campaign) was compared to the number in 2002 (after vaccination had been completed). BTV-4 BTV-16 The decrease in the proportion of vaccinated animals in the population due to replacements was also estimated, assuming a regular yearly 20% replacement rate. Results and discussion Evolution of the epidemic The third BT epidemic started in Italy on 15 April 2002 and ended on 14 April During this outbreak, BTV-2 and BTV-9 infection spread to the Province of Avellino (Campania) in July and to the Provinces of Benevento and Caserta (Campania), Foggia and Bari (Apulia), l Aquila (Abruzzi) and Isernia (Molise) in September. The only spread of BTV-2 infection to a new zone was observed in Figure 1 Distribution of BTV infection in Italy during the third epidemic according to serotype, 15 April April 2003 Moreover, during the third epidemic, two new serotypes were detected in southern Italy, namely: BTV-4 and BTV-16. Both serotypes had been reported previously in the eastern Mediterranean and in Greece. The source of the spread of these serotypes to Italy is still unknown, but illegal trade in animals is suspected. Table I Clinical outbreaks of bluetongue in Italy during the third epidemic, 15 April April 2003 Region Number of outbreaks Total number of animals in infected flocks Number of diseased animals Number of dead animals Number of slaughtered animals Basilicata Calabria Campania Lazio Molise Puglia Sardinia Sicilia Total Veterinaria Italiana, 40 (3), 2004

209 During the summer of 2003, the fourth epidemic commenced and clinical disease was observed in Sardinia alone in August involving BTV-4, a serotype that had not been included in the vaccination programme. By 7 October 2003, 850 new outbreaks had occurred in southern and western Sardinia, following a pattern that closely resembled that recorded during the summer of 2000 (Table II). No disease was recorded in central and southern Italy, but seroconversions in sentinel animals did occur (Fig. 2). Table II Clinical outbreaks of bluetongue in Italy during the fourth epidemic, 15 April October 2003 Region Number of outbreaks Animals in infected flocks (total) Number of diseased animals Number of dead animals Number of slaughtered animals Sardinia disappeared (Tuscany: 158 outbreaks and 693 diseased animals in the epidemic, 0 outbreaks in ) (Fig. 6) or was reduced by a factor of 1/100 (Sardinia: outbreaks and diseased animals in the epidemic, 10 outbreaks and 28 diseased animals in 2002) (Fig. 5). In these regions, the spread of infection was also significantly reduced by the vaccination campaign (Fig. 12). A clear demonstration of the effectiveness of vaccination was observed in Sardinia where, in August 2003, a new epidemic due to BTV-4 started, causing 850 outbreaks in six weeks; this would seem to demonstrate that the decrease in BTV-2 circulation was due to vaccine-induced resistance and not to the disappearance of conditions favouring the spread of the virus. Vaccination Vaccination in Italy was performed using two different vaccines, depending on the BTV serotypes observed in the various zones, namely: a monovalent BTV-2 vaccine was used in Sardinia, Tuscany and Latium, and a bivalent vaccine (BTV-2 and BTV-9) was used in the southern regions. In 2002, zones in which vaccination had been practised were modified according to the spread of infection (Fig. 3). Vaccination in infected zones and in areas at risk commenced on a limited scale in the late autumn of In most regions and provinces involved the vaccination programme commenced in January 2002 (Fig. 4). When the new epidemic began, in July 2002, 57% of the eligible animals in Italy had already been vaccinated (Fig. 4) but vaccination coverage in the various regions varied greatly (Figs 5, 6, 7, 8, 9, 10 and 11). Sardinia (Fig. 5) and Tuscany (Fig. 6) were able to vaccinate approximately 90% (97% in Sardinia and 87% in Tuscany) before the commencement of the new epidemic. In Basilicata, on the other hand (Fig. 7) only 2% of the population was vaccinated before the new epidemic started in July 2002 and only by the end of December 2002 had 84% of the eligible population been vaccinated. In the other regions (Sicily, Latium, Calabria and Campania) (Figures 8, 9, 10 and 11), less than twothirds of the populations were vaccinated. Direct losses: number of outbreaks and diseased animals The different levels of vaccination had clear consequences on disease occurrence. In the two regions in which approximately 90% of the ruminant population was vaccinated, clinical disease either Outbreak Seroconversion Figure 2 Distribution of BTV infection in Italy during the fourth epidemic, 15 April October 2003 Vaccination appeared not to significantly reduce the spread of disease and infection in the five regions of central and southern Italy (Basilicata, Calabria, Campania, Latium and Sicily) where variable numbers of the eligible animal population were vaccinated prior to the commencement of the new epidemic. A total of 559 outbreaks were recorded before vaccination commenced in central and southern Italy. In the epidemic (after the introduction of vaccination) infection spread to two additional regions (Molise and Apulia) causing a total of 417 outbreaks. Nevertheless, vaccination appears to have limited direct losses in these regions, despite the spread of infection to neighbouring regions. Furthermore, the total number of outbreaks was similar to the previous year. The number of Veterinaria Italiana, 40 (3),

210 outbreaks recorded in the epidemic in the five regions of central and southern Italy was significantly correlated to the level of vaccination achieved by each region at the end of July 2002 (Spearman s ρ= , p<0.0001). Indirect losses: the export of cattle from Sardinia to continental Italy The vaccination of ruminant populations was conducive to a progressive reduction of virus circulation and consequently of the zones in which movement restrictions were applied. Sardinia has been taken as an example to evaluate the effect of vaccination on animal trade for the following reasons: a) before the BT epidemic, cattle were traded extensively between Sardinia and continental Italy, especially northern Italy b) after the appearance of the disease in Sardinia, the export of cattle from the island to disease-free areas in continental Italy came to a complete standstill c) the progressive relaxation of movement restrictions paved the way for the resumption of exports to disease-free areas in northern Italy. Autumn February February- 16 May 2002 BTV-2 (7) BTV-2 and BTV-9 (8) BTV-2 (17) BTV-2 and BTV-9 (19) 16 May-25 September 2002 From 25 September 2002 BTV-2 (14) BTV-2 and BTV-9 (20) BTV-2 (14) BTV-2 and BTV-9 (29) Figure 3 Changes in the distribution and numbers of provinces (in brackets) vaccinated in Italy, Veterinaria Italiana, 40 (3), 2004

211 Vaccinated (%) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2000/8 2000/ /4 2001/8 2001/ /4 2002/8 2002/ /4 Month Percent vaccinated Number of outbreaks Number of outbreaks Figure 4 Monthly curve of animal populations vaccinated and outbreaks of bluetongue in Italy, August October 2003 Vaccinated (%) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2000/8 2000/ /4 2001/8 2001/ /4 2002/8 2002/ /4 Month Percent vaccinated Number of outbreaks Figure 5 Monthly curve of animal populations vaccinated and outbreaks of bluetongue in Sardinia, August October 2003 Vaccinated (%) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2000/8 2000/ /4 2001/8 2001/ /4 Month 2002/8 Percent vaccinated Number of outbreaks 2002/ / Figure 6 Monthly curve of animal populations vaccinated and outbreaks of bluetongue in Tuscany, August October 2003 Number of outbreaks Number of outbreaks Vaccinated (%) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2000/8 2000/ /4 2001/8 2001/ /4 Month Percent vaccinated Number of outbreaks 2002/8 2002/ /4 Figure 7 Monthly curve of animal populations vaccinated and outbreaks of bluetongue in Basilicata, August October 2003 Vaccinated (%) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2000/8 2000/ /4 2001/8 2001/ /4 Month Percent vaccinated 2002/8 Number of outbreaks 2002/ / Number of outbreaks Number of outbreaks Figure 8 Monthly curve of animal populations vaccinated and outbreaks of bluetongue in Sicily, August 2000-October 2003 Vaccinated (%) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2000/8 2000/ /4 2001/8 2001/ /4 2002/8 Month Percent vaccinated Number of outbreaks 2002/ / Number of outbreaks Figure 9 Monthly curve of animal populations vaccinated and outbreaks of bluetongue in Lazio, August 2000-October 2003 Veterinaria Italiana, 40 (3),

212 Vaccinated (%) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2000/8 2000/ /4 2001/8 2001/ /4 2002/8 Month Percent vaccinated Number of outbreaks 2002/ / Figure 10 Monthly curve of animal populations vaccinated and outbreaks of bluetongue in Calabria, August October 2003 Vaccinated (%) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2000/8 2000/ /4 2001/8 2001/ /4 2002/8 Month Percent vaccinated Number of outbreaks 2002/ / Figure 11 Monthly curve of animal populations vaccinated and outbreaks of bluetongue in Campania, August October Figure 12 Spread of bluetongue in Sardinia and Tuscany before and after vaccination Number of outbreak Number of outbreaks The total number of bovine animals sent from Sardinia to continental Italy in 2001 was (Fig. 13). Of these, 92% were moved during the last two months of the year when the effects of vaccination were visible and when extensive areas of northern Italy were free from vectors and could consequently receive animals from the surveillance zones without the risk of losing their free status (2, 9). Towards the end of 2002, a new risk assessment (3, 5) led to the authorisation of movements (for direct slaughter) of vaccinated animals from infected areas where at least 80% of the susceptible population has been vaccinated; the risk assessment also led to a new approach to define those areas under movement restrictions. From January to June 2002, a total of cattle were exported from Sardinia to continental Italy, compared to eight animals in the same period of the previous year. The arrival and spread of BTV-4 to most of the island in August 2003 once again brought the exports from Sardinia to a new halt. Number of animals / / Month Figure 13 Animals exported from Sardinia to continental Italy, January 2002-June 2003 Future perspectives The presence of four virus serotypes (BTV-2, BTV-4, BTV-9 and BTV-16) in Italy complicates the choice of vaccine to be applied in any given area: monovalent, bivalent, trivalent or tetravalent. The suspected role played by illegal animal trade in the spread of bluetongue thwarts further the adoption of a national vaccination strategy based upon stable geographical realities and known trade channels. One objective of the Italian vaccination strategy is to reduce the intensity and duration of viraemia following contact between vaccinated animals and the wild-type strain of the virus. This appears to have been achieved successfully using either monovalent or bivalent vaccines even though the combined BTV-2/BTV-9 vaccine seemed to give weaker protection (8, 10). The apparent need in some areas for a tetravalent vaccine demands that carefully controlled trials be conducted first to establish the level of interference that may occur between 258 Veterinaria Italiana, 40 (3), 2004

213 serotypes that are administered concurrently. Based on the results of these trials, and upon the level of efficacy achieved in the second vaccination programme (January-May 2003), a further risk assessment will be performed. Its primary aim will be to establish whether it is feasible to trade in vaccinated animals originating from areas in which BTV is still actively circulating. This finding could contribute to further revision of current trade regulations. References 1. Calistri P., Giovannini A., Conte A., Nannini D., Santucci U., Patta C., Rolesu S. & Caporale V. (2004). Bluetongue in Italy: Part I. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), European Commission (2001). Commission Decision of 9 November 2001 on protection and surveillance zones in relation to bluetongue and on rules applicable to movements of animals in and from those zones (2001/783/EC). Off. J., L 293, 10/11/2001, European Commission (2003). Commission Decision of 27 March 2003 on protection and surveillance zones in relation to bluetongue, and on rules applicable to movements of animals in and from those zones and repealing Decision 2001/783/EC (2003/218/EC). Off. J., L 082, 29/03/2003, Giovannini A., MacDiarmid S., Calistri P., Conte A., Savini L., Nannini D. & Weber S. (2003). The use of risk assessment to decide the control strategy for bluetongue in Italian ruminant populations. J. Risk Anal., 24 (6), Giovannini A., Conte A., Calistri P., Di Francesco C.E. & Caporale V. (2004). Risk analysis on the introduction into free territories of vaccinated animals from restricted zones. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (4), Giovannini A., Paladini C., Calistri P., Conte A., Colangeli P., Santucci U., Nannini D. & Caporale V. (2004). Surveillance system of bluetongue in Italy. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Ministero della Sanità (2001). Ordinanza 11 maggio Misure urgenti di profilassi vaccinale obbligatoria contro la febbre catarrale degli ovini (Blue-tongue). Gazz. Uff. Repubbl. Ital., Serie gen., 128, 5 June, Monaco F., De Luca N., Spina P., Morelli D., Liberatore I., Citarella R., Conte A. & Savini G. (2004). Virological and serological response of cattle following field vaccination with bivalent modified-live vaccine against bluetongue virus serotypes 2 and 9. In Bluetongue, Part II (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (4), Office International des Épizooties (OIE) (2003). Bluetongue, Chapter In Terrestrial animal health code, 12th Ed. OIE, Paris ( eng/normes/mcode/a_00038.htm accessed on 12 September 2004). 10 Savini G., Monaco F., Conte A., Migliaccio P., Casaccia C., Salucci S. & Di Ventura M. (2004). Virological and serological response of sheep following field vaccination with bivalent modified-live vaccine against bluetongue virus serotypes 2 and 9. In Bluetongue, Part II (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (4), Siegal S. & Castellan Jr N.J. (1988). Nonparametric statistics for the behavioral sciences, 2nd Ed. McGraw Hill Book Company New York, Veterinaria Italiana, 40 (3),

214 Vet. Ital., 40 (3), Entomological surveillance of bluetongue in Italy: methods of capture, catch analysis and identification of Culicoides biting midges M. Goffredo (1) (1, 2) & R. Meiswinkel (1) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario Teramo, Italy (2) Research affiliate: Agricultural Research Council (ARC)-Onderstepoort Veterinary Institute (OVI), Private Bag X05, Onderstepoort 0110, South Africa Summary To elucidate the epidemiology of vector-borne diseases that can affect livestock in the Mediterranean Basin and elsewhere, it is essential to obtain a clear understanding of the life-cycle and habits of the vector insects involved. One purpose of such investigations is to provide data for an epidemiological surveillance system. As this depends heavily upon the collection of specimens in the field, it is necessary to establish the kinds of information required, and how it can be obtained. This requires, in turn, that the method (and instrument) of capture be standardised, so that all data are as complete as possible, are comparable, and are informative at many levels. Within the surveillance system for bluetongue (BT) in Italy, the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche) is leading an intensive and countrywide survey for Culicoides (Diptera: Ceratopogonidae) using standardised methods and protocols developed in collaboration with the Onderstepoort Veterinary Institute in South Africa. These methods have now also been implemented outside Italy in Malta, Croatia, Albania and Romania. This system includes the field protocols developed for the collection of Culicoides, the laboratory protocols developed around the insect analyses and the computer-based recording of all field data. Finally, the authors provide an Easy key for the rapid identification of the principal BT vector C. imicola, and for grouping species that belong to the Obsoletus and Pulicaris vector complexes, and to the Nubeculosus and Schultzei potential vector complexes. Keywords Bluetongue Culicoides Culicoides imicola Entomology Italy Surveillance Trap Vector. The entomological surveillance of bluetongue (BT) in Italy is based primarily on a net of permanently sited traps (at least one trap every km 2, and operated weekly throughout the year). This survey is augmented by the random use of mobile traps (to investigate more thoroughly outbreaks or any other kind of epidemiological situation such as the risks associated with annual transhumance movements). As hundreds of weekly collections are made throughout Italy, it is essential to conduct these surveys simply and in a standardised way, both in the field and in the laboratory. The light-trap collections are mostly undertaken by the local Veterinary Services, and sometimes also by the farmers; these activities are co-ordinated by the ten Istituti Zooprofilattici Sperimentali located across Italy. In each Institute, a veterinarian or biologist has been trained in the various field and laboratory protocols developed by the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche). The aims of the entomological surveillance system are as follows: to obtain field data for the continuous refinement of the distribution map of Culicoides imicola for Italy to detect and elucidate the prevalence of other suspected vector Culicoides in the absence of C. imicola to elucidate the seasonal dynamics of all vector Culicoides, in order to identify zones that are seasonally free of these vectors to collect samples for the detection of BT virus (BTV) in Culicoides during outbreaks of the disease. 260 Veterinaria Italiana, 40 (3), 2004

215 Clearly these aims can be achieved only with data that are both comparable and sensitive. Field protocols for the collection of Culicoides Culicoides can be captured using several methods. The adults are most easily captured using various lighttrap models (these vary in power, in the use of different colours, with or without an additional attractant), truck traps, aspirators, bait-traps and also emergence traps for the sampling of larval habitats. In addition, the immature stages can be harvested from various moist habitats. Each method has specific advantages and disadvantages but when taken in combination, they are capable of yielding a wide range of valuable information on the biology of many species simultaneously. In any BT surveillance system, the principal aim is to capture adult Culicoides in the near vicinity of vertebrate hosts, and to employ a powerful trap (to enhance surveillance sensitivity at low Culicoides population levels and, furthermore, to increase the number of midges captured for virus isolation studies). Species identifications have to be performed by a specialised entomologist, but with a large influx of collections, it becomes important to carefully store the captured Culicoides, so that they can still be used for more detailed taxonomic studies months later. Equipment Onderstepoort blacklight suction traps of the type described by Venter and Meiswinkel in 1994 (3) are used; blacklight has been demonstrated to be 8 to 10 times more attractive than white light (4) and will increase monitoring sensitivity in areas where vector abundances are low (and will also provide more midges for the isolation of BTV). The traps are of a robust design and can remain in the field for a number of years. Surrounding the blacklight tube is a netting (apertures of about 4 mm) to exclude larger insects, especially moths whose body scales can contaminate a collection and make the analysis of a sample more laborious. Below the suction fan of the light trap, there is a white fine-gauzed bag, which is linked to a white 500 ml collection beaker. White is the colour of preference for the beaker as it enables one to more easily see the tiny Culicoides with the naked eye and establish quickly whether a successful collection has been made. The collection kit (Fig. 1) comprises the following: blacklight trap (complete with net, white gauze bag and cord to suspend the trap). If electricity is not available, modified traps can be operated using a 12 V car battery (these need to be recharged every three days) two white 500 ml collection beakers water and detergent (odourless commercial liquid soap) a cheap tea strainer with a separate square of fine white gauze (the size of half a handkerchief and of a smooth quality to which insects do not easily adhere) a labelled screw-lid container with ml of 70% ethanol a pencil a maximum-minimum thermometer a global positioning system (GPS) SBT06 forms of the BT National Surveillance System phosphate-buffered saline (PBS) with antibiotics, refrigerated at 4 C, to be used instead of water and/or ethanol when collections are made for the isolation of virus. Figure 1 Culicoides collection kit Choice of trapping sites The choice of trapping sites is made as follows: either cattle, sheep, goats or horses must be present large livestock holdings are preferred, i.e. >10 animals livestock to be located in the near vicinity of the light trap all night; stabling can be of any type (but must be open) priority should be given to farms where conditions such as pools of water or mud are found, created either naturally (rain) or by irrigation or overflows Veterinaria Italiana, 40 (3),

216 it must be established whether electricity is available to operate the light trap it is essential to establish whether the farmer/owner is willing to collaborate. Catch procedure The catch procedure should be conducted as follows: the trap must be suspended at a height of between m and be as close to the animals as possible (i.e. in the stable entrance or on a nearby tree) (Fig. 2) Figure 2 A blacklight trap suspended near sheep on a farm affected by bluetongue in Sardinia Figure 3 The blacklight trap collecting beaker being half-filled with water and detergent through a gauze bag the minimum-maximum thermometer must be placed close to the trap use the GPS to take altitude, longitude and latitude, and record the data on the SBT06 form attach the white beaker to the gauze bag and fill the beaker (through the bag) with about 200 ml of water to which a few drops of detergent have been added (Fig. 3) activate the trap and the thermometer at least 1 h before sunset switch the trap off the following morning soon after daybreak place the gauze square over the strainer and filter the insects from the white collection beaker through it; to prevent insects from adhering to the wall of the beaker, it should first be gently swirled a few times to get all the insects into suspension (Fig. 4) once all the insects have been trapped in the gauze square, it is folded and placed into 70% ethanol in a labelled bottle (Fig. 5), which is then sealed tightly if the collection is being made to attempt the isolation of virus, PBS with antibiotics must be used instead of water and/or ethanol; in addition, such collections must be sent immediately to the laboratory and must be refrigerated at 4 C during transit Figure 4 An insect collection being concentrated into the gauze square using a tea strainer Figure 5 The concentrated insect collection is then placed into a labelled screw-lid container with 70% ethanol 262 Veterinaria Italiana, 40 (3), 2004

217 complete labelling of the collection, using a pencil, indicating the trapping site and the date (that of the morning after the night of collection) store the catch in a cool place, out of direct sunlight, until transport to the laboratory complete all sections of the SBT06 form. automatically upgrade the distribution maps (and/or graphs) of C. imicola daily; these can be accessed on the website ( Similarly, the abundance maps (or graphs) of the Culicoides captured at the permanent sites are updated daily. Laboratory protocols for the analysis of catches Each light-trap collection arrives in the laboratory with the attached SBT06 form; the collection is then registered with a code number (which is also recorded on the form and on the bottle label). This code number links all field and laboratory information associated to a particular collection. Analysis of the catches The following procedure should be used when analysing catches: rinse the insects from the gauze square (by using a gentle but persistent spray) and place them back into the same labelled bottle using fresh 70% ethanol first separate all Culicoides midges from other insects; this usually is done under a stereomicroscope at low magnifications (6-10 ) sort into separate Petri dishes the species C. imicola, Obsoletus Complex, Pulicaris Complex and other Culicoides spp. and count the number of specimens in each category count all insects other than Culicoides to obtain an insect:culicoides ratio and then dispose of the non- Culicoides insects for large collections containing over 500 Culicoides, a subsample is analysed as described previously (2), and recorded (Fig. 6) the sorted Culicoides are stored in fresh 70% ethanol, in glass bottles, each labelled with the code number, the region of capture and the species or species complex these collections are stored in a cool, dark place to preserve the midges for further investigations (i.e. for the mounting of slides, and for the detection of virus using PCR). The results obtained following this protocol are both qualitative (positive/negative for C. imicola) and quantitative: total insects (other insects + Culicoides), total Culicoides, number of C. imicola, number of Obsoletus Complex, and number of Pulicaris Complex. At the end of each working day, all these collection data are inserted into a database designed to Figure 6 A subsample analysis form Samples for virus detection Virus detection can be performed by PCR or by isolation. In the first instance, midges stored in ethanol can be used and the sample can be screened months after it was collected. However, if virus isolation is to be attempted, the sample must be analysed immediately and the sorting of midges (not stored in ethanol) should be done on a chill table using only PBS with antibiotics. The kind of sample to be sorted for virus detection is dependent upon the aim of the analysis, as follows: 1. To detect the presence of BTV circulation in an area, pools of engorged and parous females of Culicoides numbering 100 individuals (or less) and of mixed species can be submitted for isolation, or PCR. Such pools can be prepared rapidly, but if positive for BTV the infected species cannot be determined. 2. To identify the vector in areas where BTV has been shown to circulate in the vertebrate host, pools of parous females of a single species or species complex are submitted for virus isolation. Maximum pool size should be 100 individuals, but single individuals can be submitted (such as is done in vector competence studies). These investigations require expertise in the accurate identification and age-grading of Culicoides (as only parous and not bloodfed females are required). Identification of Culicoides Within a surveillance system, the identification of insects of epidemiological interest has to be precise but also rapid; therefore, it is important to know which morphological characters to search for. The first step is to separate the genus Culicoides from the rest of the insect world. Within the family Veterinaria Italiana, 40 (3),

218 Ceratopogonidae there are more than 100 genera, but only four genera feed on warm-blooded vertebrate hosts: Leptoconops, Austroconops, Forcipomyia (subgenus Lasiohelea) and Culicoides. The body-shape of Culicoides is highly characteristic, but is roughly similar to other ceratopogonid genera. However, a distinguishing feature is that Culicoides usually have spotted wings (all species implicated to date in the transmission of BTV possess a wing pattern); however, some species of Culicoides lack a pattern, and, also, a wing pattern is not exclusive to Culicoides. If one relies only on body-shape and on a patterned wing (at low magnification under the stereoscope), it becomes possible to identify Culicoides with 80% specificity and 80% sensitivity; the 20% false positives will be other insects with spotted wings (e.g. Chironomidae) and the 20% false negatives will be plain-winged Culicoides. Identification of the latter group is best achieved by careful examination of the wing venation (Fig. 7), using higher magnification. Regarding identification to species, not all species of Culicoides can be reliably identified based solely upon wing pattern. Accurate identification relies upon highly developed taxonomic expertise; at best the majority of species, especially those of the vector complexes, can only be identified to species complex level; fortunately this is achieved with relative ease. Within the Obsoletus and Pulicaris Complexes (which, after C. imicola, contain the most important species involved in the transmission of BTV in Europe), there are at least six and twelve species respectively in Italy alone (R. Meiswinkel, personal observation). Cross vein Anal cell Cubital cell 2nd radial cell Figure 7 A diagrammatic representation of the Culicoides wing showing the principal nomenclature of the various veins (e.g. M1, M2) and cells (e.g. r5, m2) Easy key for the rapid identification of C. imicola and species of other vector complexes Ultimately, the correct identification of Culicoides is reliant upon a highly developed taxonomic expertise. This can require that many specimens be dissected and slide-mounted, but this procedure is m2 r5 m1 M1 M2 impracticable in those field and laboratory situations where large numbers of captures have to be identified. Thus, the majority of species can be identified to the species complex level only using a stereo-microscope at low power when the identification is based almost solely upon wing pattern. Adding to the confusion is the fact that some species complexes are poorly defined; for example, species of the Grisescens Complex (subgenus Silvicola) and the Fagineus Complex (subgenus unknown) are currently lumped (incorrectly) in the Pulicaris Complex (subgenus Culicoides). Although most of the species within these possess characteristic wing patterns, the variations are so subtle that one is not able to select simple key characters that are infallibly diagnostic. Thus the Easy key presented below has limitations and consequently it is advised that identifications be cross-referenced against published species descriptions or a taxonomist be consulted (Fig. 8). 1. Wing with a distinct pattern of light and dark spots... 2 Wing without a pattern... Ignore 2. Wing: second radial cell entirely dark... 3 Second radial cell partly or wholly pale Wing with the pale spot in the centre of cell r5, irregularly shaped (this spot not to be confused with pale spot at apex which may extend to the centre of r5)... 4 Pale spot in centre r5 round or absent... Ignore 4. Dark spot in centre of cubital cell eccentrically-shaped... No dark spot in centre of the cubital cell... Nubeculosus Complex 5. Wing with a neatly rounded dark spot in the centre of the cubital cell... Pulicaris Complex No dark spot in centre of the cubital cell Wing mostly or entirely hairy including anal cell... 8 Ignore Wing hairy in apical a-2 only and excluding anal cell Pale spot in the apex of r5 sometimes absent; if present, it ranges from being small/weakly defined to large/better defined; this spot either ovoid, round or square in shape; vein M2 without a distinct preapical excision... Obsoletus Complex Pale spot in apex of r5 large/welldefined, its anterior margin distinctly pointed; vein M2 with a distinct preapical excision... C. imicola 8. Wing hairy, with a small distinct round spot straddling vein M1 medianally... C. paolae Wing distinctly less hairy, without a round spot straddling vein M1 medianally... Schultzei Complex Figure 8 Easy key for the rapid identification of Culicoides imicola and species of other vector complexes found in the Mediterranean Region 264 Veterinaria Italiana, 40 (3), 2004

219 The key deals mostly with species and species complexes known, or suspected, to be involved in the transmission of bluetongue and African horse sickness (AHS) in the Mediterranean Region. It facilitates the identification of C. imicola sensu stricto, and the correct placement of species into the Obsoletus, Pulicaris, Schultzei and Nubeculosus vector complexes. Although rarely captured in the western half of the Mediterranean, the Schultzei Complex has been included as it is considered a potential vector and because it may be confused with the superficially similar (and relatively prevalent) C. paolae Boorman (1). The Nubeculosus Complex is included as orbiviruses replicate in this species under laboratory conditions in Europe, and because it is closely related to C. sonorensis, the most important vector of BTV in North America. 3. Venter G.J. & Meiswinkel R. (1994). The virtual absence of Culicoides imicola (Diptera: Ceratopogonidae) in the colder, high-lying area of the eastern Orange Free State, South Africa, and its implications for the transmission of arboviruses. Onderstepoort J. Vet. Res., 61, Wieser-Schimpf L., Foil L.D. & Holbrook R.F. (1990). Comparison of New Jersey light traps for collection of adult Culicoides variipennis (Diptera: Ceratopogonidae). J. Am. Mosquito Control Assoc., 6, References 1. Boorman J., Mellor P.S. & Scaramozzino P. (1996). A new species of Culicoides (Diptera: Ceratopogonidae) from southern Italy. Parassitologia, 38, Van Ark H. & Meiswinkel R. (1992). Subsampling of large light-trap catches of Culicoides (Diptera: Ceratopogonidae). Onderstepoort J. Vet. Res., 59, Veterinaria Italiana, 40 (3),

220 Vet. Ital., 40 (3), Improving light-trap efficiency for Culicoides spp. with light-emitting diodes A.L. Bishop, R.J. Worrall, L.J. Spohr, H.J. McKenzie & I.M. Barchia NSW Agriculture, Locked Bag 26, Gosford, NSW 2250, Australia Summary The robustness of light traps used to monitor Culicoides spp. throughout Australia was improved with stainless steel and heavy duty plastic fittings. Printed circuit boards and light-dependent resistors were modified to be compatible with recent advances in electronics. In experiments with light-emitting diodes (LEDs), C. brevitarsis Kieffer was significantly attracted to green light. This species is the major vector of Akabane and bluetongue viruses in Australia and is the main target of a national monitoring programme using light traps. This response was significantly greater than the response to the incandescent lights currently used in the light traps. Catches of C. brevitarsis were also related to the intensity of the green LEDs. These were more effective than the currently used incandescent globes at intensities between 46% and 142% of the incandescent intensity. The response of seven other Culicoides spp. to the LEDs was also determined. Keywords Arbovirus monitoring Australia Bluetongue Culicoides brevitarsis Light-emitting diodes Light intensity Light trap Vectors Viruses. Introduction The Australian Culicoides fauna is extensive and diverse. Distributions are restricted by geography, weather and habitat availability. Several species from the genus are vectors of viruses affecting native animals (10). Culicoides brevitarsis Kieffer is the main species responsible for the transmission of bluetongue (BT) and Akabane viruses to livestock (9). Distribution is chiefly coastal and it is endemic in the eastern half of Australia from the Northern Territory (8) to the northern/mid-northern coastal plains of New South Wales (NSW) (2, 3). Light traps with incandescent globes are used to monitor the presence of Culicoides species (as part of the National Arbovirus Monitoring Program [NAMP]) (7), to compare their relative abundances and to conduct research (2). Until recently, these traps had changed little from those first described by Dyce et al. (6). In the field, the traps are subjected to rough treatment and breakages were common and disruptive to the monitoring programme. Original printed circuit boards (PCBs) proved incapable of maintaining consistent outputs as new electronic technology developed. Replacement component (transistor) burnout increased as there was no definite switch point in the old system. This caused the transistors to half-switch and to dissipate more heat energy than could be tolerated. Improvements to the robustness and circuitry of traps were therefore major requirements for maintaining an efficient trapping system. There have been instances at the margins of the distribution of C. brevitarsis in NSW where virus activity has been detected by the serological testing of sentinel cattle herds in the apparent absence of C. brevitarsis in light traps (P.D. Kirkland, personal communication). It was considered possible that the traps failed to record low numbers of infective C. brevitarsis and that trap efficiency should be reconsidered. Improved trapping of mosquitoes has been achieved by determining mosquito responses to the colour and intensity of light sources (1, 5). There have been no similar studies on Culicoides spp. Insects can generally perceive and respond to light in the nm range and their relative response can vary considerably over this range. The standard incandescent light sources used in these light traps generally have a maximum output at 700 nm with little or no output below 400 nm. Advances in light- 266 Veterinaria Italiana, 40 (3), 2004

221 emitting diodes (LEDs) in the last 5 to 10 years have produced LEDs that are energy efficient, often producing a greater total photon flux (TPF) than incandescent globes in the nm range for the same power input making them suitable for battery operation such as that used in light traps. LEDs also can provide closely defined outputs across narrow spectral ranges enabling responses to colour to be investigated more effectively. A study was therefore conducted to determine the response of a range of Australian species of Culicoides to different colours in the visible spectrum with the use of LEDs. Materials and methods Modifications to light traps Strengthening of the traps and changes to electronic circuitry were made progressively by trial and error over several years. Light-emitting diodes All experiments were conducted in the Hunter Valley (NSW) in 2002 and The light source (incandescent globe) in standard light traps was replaced in treatment traps with a range of LED treatments and compared to the incandescent light. The traps were powered by three 1.5 V alkaline D cells, which were replaced after two nights of operation. The quantum output of incandescent globes was measured at 20 C with a LI- COR Model LI-250 light meter with a LI-COR quantum sensor (approximately linear over nm). The current to the LEDs (three for blue, green, white and red, and five for yellow) was then adjusted until the quantum output was the same as the incandescent light source. The LEDs (other than yellow) were mounted in polycarbonate plastic diffusers (120 apart) to ensure an even distribution of light. The yellow LEDs were mounted facing directly outwards on the same plastic caps at 72 apart as five were necessary due to their lower quantum output per current input compared to other LEDS used. Three trials were conducted. The first trial was carried out in March and April 2002 using blue, green, white, yellow and red LEDs and the standard incandescent globes. The traps were hung from 2 m-high L shaped frames placed at 20 m intervals on one side of 30 ha paddocks containing cattle. Yellow was not included in the first two experiments and replaced red in the next two experiments. The five treatments were arranged in five randomised blocks and re-randomised between each of four experiments. Collections were made over two nights into bottles containing 70% alcohol. C. brevitarsis was identified under 10 magnification and total numbers were recorded. The second trial was conducted in February The experimental site was chosen because it frequently has the greatest diversity of species at sites monitored in coastal NSW (A.L. Bishop, unpublished data). It is also marginal for C. brevitarsis in most years. Blue, green, yellow, red and incandescent treatments were used in this trial. The treatments were arranged in five randomised blocks which were re-randomised at the start of each of four experiments. Traps were hung as before at 12 m intervals on two sides of a 10 ha paddock containing cattle. Collections were made over one night, the samples sorted and numbers of C. brevitarsis, C. austropalpalis Lee and Reye, C. bundyensis Lee and Reye, C. bunrooensis Lee and Reye, C. dycei Lee and Reye, C. marksi Lee and Reye, C. nattaiensis Lee and Reye and C. victoriae Macfie recorded. The third trial was also conducted in February Green LEDs at four intensities relative to the intensity of incandescent globes were compared with the incandescent light against C. brevitarsis. The intensities were varied by adjusting the current to the LEDs. The five treatments were arranged in five randomised blocks in a 36 ha paddock containing cattle. The treatments were re-randomised between each of four experiments. Collections were made over one night and C. brevitarsis numbers counted as before. Statistical methods The influence of light frequency or intensity on counts of Culicoides spp. was modelled using a mixed linear regression approach which allowed the separation of variance components into fixed and random effects. Insect counts were log e transformed for Trial 1, Trial 3 and for C. austropalpalis in Trial 2. A square root transformation was used for counts of all other species in Trial 2 due to their low numbers. Analysis of the transformed counts was conducted using the restricted maximum likelihood (REML) directive in Genstat 5.4.1, Release 3. Treatment effects were examined for significance using Wald tests while treatment means were compared using the least significant difference (LSD) technique at the 5% level. In Trial 2, the counts for C. brevitarsis and C. nattiensis were very low and were subsequently pooled for each block and an analysis of variance performed. Veterinaria Italiana, 40 (3),

222 Results Modifications to light traps The robustness of the light traps was improved by the addition of: heavy-duty PVC battery boxes and lightdependent resistor (LDR) covers stainless steel weather protection plates, ribs on cones and bottle connections. The PCB design was modified by the additions of: screw terminals to prevent wire breakage a 10 turn 10 K trim pot to allow easy adjustments a definite 100% switch point to eliminate half switching a timer chip incorporating a hysteresis loop so that the light level required to switch the unit off would be higher than the level required to switch the unit on. Light-emitting diodes The treatment effect in Trial 1 was highly significant with catches of C. brevitarsis highest with green LEDs and with significant differences between each singleband treatment (Table I). White was similar to the blue and green treatments but included all wavelengths with peak emissions in the blue and yellow ranges. Significant treatment effects were recorded in Trial 2 for each of the eight species recorded. Catches of C. brevitarsis were again highest with the green LEDs (Table I). C. austropalpalis, C. bunrooensis and C. marksi each exhibited highest responses to blue LEDs although green LEDs were also more effective than the incandescent light. Significantly higher responses to the blue and green LEDs relative to the incandescent could not be separated for C. bundyensis, C. dycei, C. nattaiensis and C. victoriae. The overall treatment effect of different intensities of green LEDs was significant in Trial 3. Catches increased with intensity but were not significantly different at the two highest intensities (Table I). Significantly more C. brevitarsis were caught at all intensities tested than in the incandescent traps. Discussion Changes to the structure and circuitry of the light traps have significantly reduced trap breakage and breakdown and improved the efficiency of the NAMP. Light trapping of C. brevitarsis was more efficient when incandescent globes were replaced with green LEDs. Attraction was also more effective as the intensity of the green light was increased, with catches at four intensities significantly greater than those with the incandescent light. An upper threshold of intensity suggested by the two highest intensities require confirmation. While trapping of C. brevitarsis was the major aim, trapping of several other species would also be improved with the green LEDs. Specific trapping of some of these species could be maximised with blue LEDs. Most predictions of the activity and spread of C. brevitarsis in NSW are based on population monitoring with light traps and are more dependent on the species occurrence than its density (2, 4). Table I Predicted (back-transformed) means of Culicoides brevitarsis in response to coloured light-emitting diodes arranged in spectral order in two trials and to different intensities of green LEDs in relation to the intensity of standard incandescent globes in Australia in 2002 and 2003 LED experiments Treatment C. brevitarsis* C. brevitarsis** LED intensity experiment LED: incandescent TFD ratio C. brevitarsis (%) Red LED 17.0 e c Yellow LED 53.5 d 0.7 b b Green LED a 4.7 a a Blue LED b 1.3 b a Incandescent c 0.5 b 100 (incandescent) d White LED a b * Trial 1 TFD total photon flux ** Trial 2 LED light-emitting diode Means in columns with the same superscript letter are not significantly different (P <0.05) Larger catches in endemic or established areas where the occurrence of C. brevitarsis is not in question may therefore be of little value and the extra time taken to count increased numbers be unnecessary. Greatest 268 Veterinaria Italiana, 40 (3), 2004

223 benefit for monitoring use would be for first occurrences outside endemic areas and at sites with low C. brevitarsis density. Further benefits could be derived where larger catches may be required for virus isolation from vectors, for experimental use of vectors with animals or for detecting vectors at key locations involved in the export of livestock (staging areas and ports). Colours with higher attraction could possibly be used in trapping systems designed to control the insects particularly where important livestock are kept in confined areas. Other Culicoides vectors of the Akabane and bluetongue viruses also exist in Australia s far north. Determination of responses to colour in a wider range of Culicoides species throughout Australia and overseas could therefore be an important adjunct to the understanding and control of these pest species and this could easily be carried out with LEDs in currently used light traps. Acknowledgements The support of staff from the C.B. Alexander Agricultural College and co-operating farmers was greatly appreciated. Part funding was received from Biosecurity Australia. References 1. Ali A., Nayar J.K., Knight J.W. & Stanley B.H. (1990). Attraction of Florida mosquitoes (Diptera: Culicidae) to artificial light in the field. Proceedings and papers. Ann. Conf. Calif. Mosquito Vector Control Assoc., 57, Bishop A.L., Kirkland P.D., McKenzie H.J., Spohr L.J., Barchia I.M. & Muller M.J. (1995). Distribution and seasonal movements of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) at the southern limits of its distribution in New South Wales and their correlation with arboviruses affecting livestock. J. Aust. Entomol. Soc., 34, Bishop A.L., Kirkland P.D., McKenzie H.J. & Barchia I.M. (1996). The dispersal of Culicoides brevitarsis in eastern New South Wales and associations with the occurrences of arbovirus infections in cattle. Aust. Vet. J., 73, Bishop A.L., Barchia I.M. & Spohr L.J. (2000). Models for the dispersal in Australia of the arbovirus vector, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae). Prev. Vet. Med., 47, Burkett D.A., Butler J.F. & Kline D.L. (1998). Field evaluation of coloured light-emitting diodes as attractants for woodland mosquitoes and other Diptera in north central Florida. J. Am. Mosquito Control Assoc., 14, Dyce A.L., Standfast H.A. & Kay B.H. (1971). Collection and preparation of biting midges (Fam. Ceratopogonidae) and other small Diptera for virus isolation. J. Aust. Entomol. Soc., 11, Kirkland P.D., Ellis T., Melville L.F. & Johnson S. (1995). The national arbovirus monitoring program as a model for studying the epidemiology of bluetongue in China. In Bluetongue disease in South- East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, Muller M.J., Murray M.D. & Edwards J.A. (1981). Blood-sucking midges and mosquitoes feeding on mammals at Beatrice Hill, NT. Aust. J. Zool., 29, Muller M.J., Standfast H.A., St George T.D. & Cybinski D.H. (1982). Culicoides brevitarsis (Diptera: Ceratopogonidae) as a vector of arboviruses in Australia. In Proc. Third Symposium on arbovirus research in Australia (T.D. St George & B.H. Kay, eds). CSIRO, Brisbane, Standfast H.A., Dyce A.L., St George T.D., Muller M.J., Doherty R.L., Carley J.G. & Fillipich C. (1984). Isolation of arboviruses from insects collected at Beatrice Hill, Northern Territory of Australia, Aust. J. Biol. Sci., 37, Veterinaria Italiana, 40 (3),

224 Vet. Ital., 40 (3), Distribution and abundance of Culicoides imicola, Obsoletus Complex and Pulicaris Complex (Diptera: Ceratopogonidae) in Italy M. Goffredo (1), A. Conte (1) (1, 2) & R. Meiswinkel (1) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy (2) Research affiliate: Agricultural Research Council (ARC)-Onderstepoort Veterinary Institute (OVI), Private Bag X05, Onderstepoort 0110, South Africa Summary Between 2000 and 2003, thousands of light-trap collections for Culicoides were made throughout Italy and a detailed distribution map of the primary vector of bluetongue (BT) virus (BTV), C. imicola compiled. In some areas, however, where clinical BT occurred and C. imicola could not be captured, the virus was isolated from biting midges belonging to the Obsoletus and/or the Pulicaris Complexes. Thus, the distribution and abundance of these two species complexes in Italy, as determined from about collections, are reported here also and compared to that of C. imicola (from about collections). The probable spread of the main vector of BT, C. imicola, into the northern third of Italy, and the widespread prevalence of additional vectors of the Obsoletus and Pulicaris Complexes, indicate nearly all regions of Italy to be at some risk to incursions of BTV. However, these complexes comprise at least six and twelve species, respectively, so precisely which species are able to transmit BTV remains incompletely known. Keywords Bluetongue Culicoides imicola Italy Obsoletus Complex Pulicaris Complex. Introduction Since 2000, when bluetongue (BT) first affected Italy, thousands of light-trap collections have been made for Culicoides throughout the country, and detailed distribution maps for C. imicola Kieffer, 1913 compiled (3). During 2000 and 2001 the disease occurred in all regions where this vector was detected, even in those areas where it was found to be extremely rare (2, 3). However, in 2002, no specimens of C. imicola could be captured in some areas of Italy where clinical bluetongue (and fatalities) occurred amongst sheep. In three of these outbreaks, BT virus (BTV) serotype 2 and/or BTV serotype 9 were successfully isolated from biting midges of the Obsoletus Complex (7); BTV-2 was isolated also from a species of the Pulicaris Complex on the island of Sicily (1). In this study, the authors report on the distribution and abundance of C. imicola, and of the Obsoletus and the Pulicaris Complexes across Italy. Materials and methods The collection and the identification of C. imicola and species of the Obsoletus and Pulicaris Complexes were performed in accordance with the protocols developed by the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche) (4). The latter two species complexes comprise at least six and twelve species, respectively; these constituent species could not be identified to the species level, and therefore the distribution of individual species could not be determined. A more complete discussion of the taxonomy of these two species complexes is given elsewhere (6). The map for C. imicola, the only species of the Imicola Complex to occur in the Mediterranean Basin, was compiled from almost light-trap collections made over the last four years in municipalities. The distribution maps for the Obsoletus and the Pulicaris Complexes were compiled from approximately collections made in (629 municipalities sampled). 270 Veterinaria Italiana, 40 (3), 2004

225 Results The distributions and abundances of the Obsoletus and Pulicaris Complexes are shown in Figures 1 and 2, respectively; those for C. imicola are shown in Figure 3. These Figures show the log abundances of the largest light-trap collection/municipality and are colour-coded to aid visual interpretation of the maps. These abundances are also graphically depicted (in two forms) in Figure 4, but in this instance the data are combined to produce a single figure/region. All 20 regions of Italy were sampled; these are ranked from north to south (roughly) with the islands of Sicily and Sardinia ranked last Figure 2 Maximum light-trap abundances of the Pulicaris Complex collections from 629 municipalities in all regions of Italy, The complex comprises at least twelve species Figure 1 Maximum light-trap abundances of the Obsoletus Complex: collections from 629 municipalities in all regions of Italy, The complex comprises at least six species Discussion The absence of historical data on the presence of C. imicola in Italy before 2000 does not allow us to establish whether this vector is a recent invader or not (3). Continuous and thorough monitoring across all seasons in all regions has revealed that C. imicola occurs beyond the 44th parallel, but at this northern end of its range only one or two specimens are collected at any one site/season (Fig. 3). Continued monitoring of such sites is required to establish whether C. imicola is spreading northwards and whether it is increasing in abundance. The maps (and graphs) show that species of the Obsoletus Complex occur abundantly throughout Italy, whilst species of the Pulicaris Complex tend to be more abundant in the southern regions (Fig. 4). The probable spread of the main vector, C. imicola, into the northern third of Italy, and the widespread prevalence of additional vectors of the Obsoletus and Pulicaris Complexes, would seem to indicate that nearly all regions of Italy are at some risk to incursions by BTV. However, an important caveat is that the information on the Obsoletus and the Pulicaris Complexes remains extremely limited. A number of species comprise these complexes, and so their distributional and abundance data are difficult to interpret. In addition, it is almost certain that not all species within these complexes will be able to transmit BTV but precisely which species do, remains unknown. This conflation of data is due simply to the fact that these species are extremely difficult to identify. Recent molecular analyses (combined with morphological studies) (5, 6), have revealed that at least eighteen species of these two complexes (combined) occur in Italy. It remains to be seen whether the separate mapping of these species can be achieved in the future using molecular tools, or whether a greater effort should be made towards locating diagnostic morphological characters easily scored under the dissecting microscope. Veterinaria Italiana, 40 (3),

226 A B Log midges (maximum/month /2001 1/2002 2/2002 3/2002 4/2002 5/2002 6/2002 7/2002 8/2002 9/ / / /2002 1/2003 2/2003 3/2003 4/2003 5/2003 6/2003 7/2003 8/2003 Month/year Log midges (maximum/month /2002 2/2002 4/2002 5/2002 6/2002 7/2002 8/2002 9/ / / /2002 1/2003 2/2003 3/2003 4/2003 5/2003 6/2003 7/2003 8/2003 9/ Month/year Culicoides (total) Culicoides imicola Figure 3 Maximum light-trap abundances of Culicoides imicola; collections from municipalities in all regions of Italy ( ) and seasonal abundances of total Culicoides in the two northernmost collection sites where C. imicola was found in the Provinces of Genoa (A) and Parma (B) A Region Piemonte Veneto Friuli Venezia Giulia Valle d Aosta Liguria Lombardia Trentino Alto Adige Emilia Romagna Marche Umbria Lazio Toscana Puglia Basilicata Molise Campania Abruzzo Calabria Sicilia B Log midges Piemonte Veneto Friuli Venezia Giulia Valle d Aosta Liguria Lombardia Trentino Alto Adige Emilia Romagna Marche Umbria Lazio Toscana Puglia Basilicata Molise Campania Abruzzo Calabria Sicilia Sardegna Culicoides imicola Pulicaris Complex Region Obsoletus Complex Sardegna Log midges Culicoides imicola Obsoletus Complex Pulicaris Complex Figure 4 Maximum light-trap abundances of each of the species of Culicoides imicola, Obsoletus Complex and Pulicaris Complex captured in the various regions of Italy Log numbers represented as a bar graph (A) and a linear graph (B) ranked roughly from north to south Multiple species comprise the Obsoletus and Pulicaris Complexes, each complex mapped from about collections ( ); C. imicola is mapped from about catches ( ) 272 Veterinaria Italiana, 40 (3), 2004

227 References 1. Caracappa S., Torina A., Guercio A., Vitale F., Calabrò A., Purpari G., Ferrantelli V., Vitale M. & Mellor P.S. (2003). Identification of a novel bluetongue virus vector species of Culicoides in Sicily. Vet. Rec., 153, Goffredo M., Satta G., Torina A., Federico G., Scaramozzino P., Cafiero M.A., Lelli R. & Meiswinkel R. (2001). The 2000 bluetongue virus (BTV) outbreak in Italy: distribution and abundance of the principal vector Culicoides imicola Kieffer. In Proc. Tenth International Symposium of the American Association of Veterinary Laboratory Diagnosticians (AAVLD), Salsomaggiore, Parma, 4-7 July. AAVLD, Ames, Goffredo M., Conte A.M., Cocciolito R. & Meiswinkel R. (2003). The distribution and abundance of Culicoides imicola in Italy. Vet. Ital., 39 (47), Goffredo M. & Meiswinkel R. (2004). Entomological surveillance of bluetongue in Italy: methods of capture, catch analysis and identification of Culicoides biting midges. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Gomulski L.M., Meiswinkel R., Goffredo M., Torti C., Malacrida A.R. & Gasperi G. (2002). Differenziazione genetica di specie criptiche del genere Culicoides, sottogenere Avaritia (Diptera: Ceratogonidae), presenti in Italia. In Proc. 63rd Congresso Nazionale Unione Zoologica Italiana, Rende, Cosenza, September, Meiswinkel R., Gomulski L.M., Delécolle J.-C., Goffredo M. & Gasperi G. (2004). The taxonomy of Culicoides vector complexes unfinished business. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Savini G., Goffredo M., Monaco F., de Santis P. & Meiswinkel R. (2003). Transmission of bluetongue virus in Italy. Vet. Rec., 152 (4), 119. Veterinaria Italiana, 40 (3),

228 Vet. Ital., 40 (3), Association between the bluetongue epidemic in Lazio and Tuscany (central Italy) and distribution and abundance of Culicoides imicola and C. obsoletus vectors G. Scavia, G.L. Autorino, C. De Liberato, F. Farina, G. Ferrari, M. Guidoni, A. Magliano, M. Miceli, F. Scholl, M.T. Scicluna & P. Scaramozzino Istituto Zooprofilattico Sperimentale delle Regioni Lazio e Toscana, Via Appia Nuova 1411, Rome, Italy Summary During the epidemic of bluetongue (BT) in Lazio and Tuscany between 2001 and 2003, the distribution pattern of Culicoides imicola did not always correspond either geographically or seasonally, with virus circulation. Culicoides obsoletus was observed to be abundant, ubiquitous and active throughout the year. The geographical and seasonal distribution of BT virus (BTV), C. imicola and C. obsoletus was compared. The territory of the two regions was divided into 30 cells each measuring km 2. The presence of C. obsoletus was recorded in every cell, while C. imicola was detected in 18 of the 30 cells, but was absent in 6 of the 21 cells that indicated the presence of BTV. The occurrence of seroconversions appeared to be positively correlated with maximum C. obsoletus catches. Seroconversions were recorded throughout the year, even when C. imicola was not active, whereas C. obsoletus was detected during the entire period. The occurrence of BTV circulation in areas and periods where C. imicola was absent, and the abundant and constant presence of adult C. obsoletus in all the cells, suggest the active role of the latter species in BTV circulation in central Italy. Keywords Bluetongue Culicoides imicola Culicoides obsoletus Italy Lazio Tuscany. Introduction Since 2000, the circulation of bluetongue (BT) virus (BTV) has been recorded in central and southern Italy, causing one of the largest epidemics of BT in recent decades in the Mediterranean Basin. At present, the serotypes involved are mainly 2, 4 and 9. In the summer-autumn of 2001, the Lazio and Tuscany regions of central Italy experienced 220 clinical outbreaks of BT caused by BTV serotype 2, which was also involved in the 15 outbreaks observed in southern Lazio in Surveillance, based on the regular assessment of the serological status of sentinel animals (58 in each of the 20 km 20 km quadrants) and on extensive Culicoides collections, commenced in the two regions in the summer of 2001, in the framework of a BT National Surveillance Programme. During the same year, because of the occurrence of BT outbreaks in the study area, serological surveillance of sentinels was suspended between September and December and surveillance concentrated on the detection of clinical cases. In 2002, the Health Ministry authorised the vaccination, using a live attenuated vaccine, of all susceptible domestic species in areas considered at risk for BT. In both 2002 and 2003, over 80% of the susceptible animal population in large areas of Lazio and Tuscany was vaccinated in two consecutive campaigns. Culicoides imicola Kieffer, 1913 is the only known and confirmed BTV vector in the Mediterranean region, and its presence in Sardinia, Sicily and mainland Italy was confirmed during the entomological surveillance programme (4). However, C. imicola was absent in many sites in Lazio and Tuscany where BTV circulated and, if present, its populations were small 274 Veterinaria Italiana, 40 (3), 2004

229 (3). On the other hand, C. obsoletus (Meigen), 1818 was almost ubiquitous and active throughout the year (C. De Liberato, unpublished findings). Culicoides obsoletus is a species known to be capable of sustaining BTV replication (6) and also strongly suspected of being the BT vector in Eastern Europe (5, 9). In fact, C. obsoletus is known to be a group of species and, when C. obsoletus is referred to in this paper, the Obsoletus Complex is intended. The aim of this study was to investigate the role of C. imicola and C. obsoletus as BTV vectors by evaluating the association between the geographical and seasonal distribution of both species, together with BTV circulation in the two regions. Materials and methods The study was conducted from July 2001 to August According to the criteria of the entomological surveillance programme in the study area, which includes the territories of Lazio and Tuscany, the region was divided into thirty km 2 quadrants. Some of the outer quadrants had different shapes and areas (Fig. 1). Figure 1 Regions of Lazio and Tuscany divided into thirty geographical units The assessment of BTV circulation was made through BT clinical outbreaks and seroconversions, the latter defined as a positive serological diagnosis in a recent previously seronegative tested sentinel, according to the BT National Surveillance Programme. For serological diagnosis a positive and confirmed scheme was used with ELISA (7) as the first screening step, followed by serum neutralisation; the latter was performed by the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche), to confirm seropositivity and serotype. Each seroconversion detected was included in the present data set, regardless of BTV serotype and whether or not it was a field or vaccine virus. Entomological surveillance was based on Culicoides collections, with trapping and sorting activities as described by standardised methodologies (1). A total of catches were performed in 415 different trapping sites over the entire study area. Culicoides imicola was counted in the whole set, whereas C. obsoletus in only a subset of catches made at 160 trapping sites. For statistical analysis, all data were grouped according to the 30 geographical units (Fig. 1). The relationship between presence and abundance of C. imicola and/or C. obsoletus and virus circulation was evaluated. The amount of BTV circulation in each cell was estimated by a seroconversion rate (SR) per month, calculated as the mean number of seroconversions per month, over a period (follow-up) ranging from the first seroconversion observed in every cell to 100 days after the last. Counts of C. obsoletus and C. imicola were log transformed because of the high abundance variability. Maximum catches of C. obsoletus (CobMax) and C. imicola (CimMax) were considered more appropriate than mean catches to express population sizes (2). All catches for C. obsoletus were classed into three equally sized categories. The proportion of C. obsoletus catches in the maximum class (P-obs) was calculated for each cell. Due to the generally scarce catches of C. imicola, only presenceabsence was considered for each cell. Nine cells that had no evidence of BTV circulation were not included in the statistical analysis. Pearson s correlation coefficient was calculated between SR and P-obs, CobMax and CimMax. The mean SR value was compared between C. imicolapresent and C. imicola-absent cells. Multiple linear regression analysis was performed to assess the association between SR and C. imicola presence and P-obs, as well as with CobMax and CimMax. Analysis of variance (ANOVA) was performed to compare P-obs, CobMax and CimMax mean values between cells with and without clinical outbreaks. In this analysis, only outbreaks that occurred in 2001 (before the commencement of the vaccination campaign), were considered. The seasonality of seroconversions and C. obsoletus and C. imicola activity in the entire study area are presented in Figure 2. Veterinaria Italiana, 40 (3),

230 Mean monthy catch Jan. 02 Mar. 02 May 02 July 02 Sept. 02 Nov. 02 Jan. 03 Mar. 03 May 03 July 03 Date C. imicola C. obsoletus Seroconversions Seroconversions Figure 2 Mean monthly catches of Culicoides imicola and C. obsoletus and number of bluetongue virus seroconversions per month, January 2002-August 2003 The commencement of the 2002 and 2003 vaccination campaigns are indicated by arrows Results Table I summarises the results of serological, clinical and entomological surveillance in the 30 cells. Culicoides obsoletus was present in all cells, whereas C. imicola was found in only 18 of the 30 cells. In 6 of the 21 cells (28.6%) in which BTV circulation was detected, C. imicola was not found. On the other hand, this species was detected in 6 of the 9 cells with clinical outbreaks. The SR was significantly positively correlated (r=0.46, p<0.05) only with CobMax. Results from regression analysis confirm the statistically significant (β=0.46, p<0.05) association between SR and CobMax, even when taking into account CimMax. Linear regression showed a positive, although not statistically significant, association between SR and P-obs, even when taking into account the presence Table I Virus circulation (seroconversions and outbreaks) and entomological data (number of catches, C. imicola and C. obsoletus maximum catch) for each cell Cell Seroconversions ( ) Follow-up (months) SR (a) Outbreaks (b) Catches (c) No. of Culicoides CimMax (d) CobMax (e) P-obs (f) (49) % 2 (g) 31 (15) % (50) % 4 (g) 112 (46) % 5 (g) 11 (4) % 6 (g) 13 (8) % (15) % 8 (g) 92 (46) % 9 (g) 197 (88) % (75) % (18) % (72) % 13 (g) 93 (65) % (156) % (165) % (49) % (14) % (43) % (48) % (102) % (28) % 22 (g) 57 (23) % (1) 58 (49) % (9) % (7) 22 (18) % (75) % (142) % (7) 92 (36) % (4) % 30 (g) 3 (3) % a) Seroconversion rate per month e) C. obsoletus maximum catch b) In brackets: 2002 outbreaks not included in statistical analysis f) Proportion of C. obsoletus catches in the highest abundance class c) In brackets: number of catches of C. obsoletus g) Cells not included in statistical analysis d) C. imicola maximum catch 276 Veterinaria Italiana, 40 (3), 2004

231 of C. imicola. A slightly lower mean SR was found in cells with C. imicola presence, if compared with cells without this species, although this was not significant. Mean monthly catches of C. imicola and C. obsoletus and seroconversion numbers for each month from January 2002 to August 2003 are shown in Figure 2. Seroconversions in sentinel bovines were recorded throughout the year, even during the winter months when C. imicola was not active, whereas activity of C. obsoletus was detected during the entire period. Discussion Occurrence of BTV circulation was recorded in areas from which C. imicola was absent and during periods of the year when they were not found to be active elsewhere. On the other hand, the abundant and constant presence of adult, active C. obsoletus, indicates it played a role in the epidemic considered here. This evidence appears to be confirmed by the positive association between C. obsoletus abundance and BTV circulation, as indicated by the correlationship between SR and CobMax and, even if not statistically significant, P-obs. Culicoides imicola, the only proven BT vector in the Palaearctic Region, did not affect the SR in these two regions. A possible limit of the present study could be the geographical units being too large, hence too limited in number. Despite the large quantities of data provided by the present study (especially by entomological surveillance), which meant very robust indicators were obtained for every cell, the low number of units reduces the possibility of performing a more powerful statistical analysis. Further analysis, based on the same surveillance system, could be performed using single sites. In October 2002, BTV serotype 2 was isolated from a pool of C. obsoletus, caught on a farm in southern Lazio clinically affected by BT (C. De Liberato, unpublished findings). Isolation of BTV from C. obsoletus, together with data from serological, clinical and entomological surveillance, could prove to be an important breakthrough in BT epidemiology. Areas until now considered risk-free, could now be considered at risk. The presence of C. obsoletus extends from southern Italy to Great Britain and can be recorded throughout the year in the southern areas of distribution (5, 8). Furthermore, the presence of this species throughout the year in the study area (Fig. 2) suggests the possibility of BTV overwintering in infected midges. Finally, the contemporaneous presence of large numbers of viraemic animals (i.e. during vaccination campaigns), in an atypical season for BTV circulation, together with the concomitant presence of active C. obsoletus, could explain the occurrence of many seroconversions at the beginning of 2003 (Fig. 2). This finding raises the question about possible vaccine virus transmission among susceptible animals through active vectors. Finally, the bovine population was also included in the vaccination programme, thus introducing a new element in the interaction between vaccine virus hosts and vectors on a large scale. References 1. Calistri P., Goffredo M., Caporale V. & Meiswinkel R. (2003). The distribution of Culicoides imicola in Italy: application and evaluation of current Mediterranean models based on climate. J. Vet. Med., 50, Capela R., Purse B.V., Pena I., Wittmann E.J., Margarita Y., Capela M., Romao L., Mellor P.S. & Baylis M. (2003). Spatial distribution of Culicoides species in Portugal in relation to the transmission of African horse sickness and bluetongue viruses. Med. Vet. Entomol., 17 (2), De Liberato C., Purse B.V., Goffredo M., Scholl F. & Scaramozzino P. (2003). Geographical and seasonal distribution of the bluetongue virus vector Culicoides imicola, in central Italy. Med. Vet. Entomol., 17 (4), Goffredo M., Conte A.M., Cocciolito R. & Meiswinkel R. (2003). The distribution and abundance of Culicoides imicola in Italy. Vet. Ital., 39 (47), Jennings D.M. & Mellor P.S. (1988). The vector potential of British Culicoides species for bluetongue virus. Vet. Microbiol., 17, Mellor P.S. & Pitzolis G. (1979). Observations on breeding sites and light-trap collections of Culicoides during an outbreak of bluetongue in Cyprus. Bull. Entomol. Res., 69, Office International des Épizooties (OIE) (2000). Chapter In Manual of standards, diagnostic tests and vaccines. 4th Ed. OIE, Paris, Rawlings P. & Mellor P.S. (1994). African horse sickness and the overwintering of Culicoides spp. in the Iberian peninsula. Rev. Sci. Tech. Off. Int. Épiz., 13 (3), Wittmann E.J. & Baylis M. (2000). Climate change: effects on Culicoides-transmitted viruses and implications for the UK. Vet. J., 160, Veterinaria Italiana, 40 (3),

232 Vet. Ital., 40 (3), Entomological surveillance for bluetongue on Malta: first report of Culicoides imicola Kieffer M. Goffredo (1), M. Buttigieg (2), R. Meiswinkel (1, 3), J.-C. Delécolle (4) & S. Chircop (2) (1) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy (2) Ministry for Rural Affairs and the Environment Food Veterinary Regulation Division, Alberttown, Marsa, Malta (3) Research affiliate: Agricultural Research Council (ARC)-Onderstepoort Veterinary Institute (OVI), Private Bag X05, Onderstepoort 0110, South Africa (4) Musée Zoologique de l Université Louis Pasteur et de la Ville de Strasbourg (MZS), 29 bd de la Victoire, Strasbourg, France Summary A survey for Culicoides Latreille, 1809, was made on Malta in 2002 to establish whether Culicoides imicola Kieffer, 1913, the principal vector of bluetongue virus (BTV) in the Mediterranean Basin, or any other suspected vector species, was present. The collections and analyses were performed in accordance with the protocols of the National Reference Centre for Exotic Diseases (CESME Centro Studi Malattie Esotiche) in Teramo, Italy. Eighty-four catches were made between May and October at six permanent sites, namely: Mellieha, Rabat, San Gwann, Zejtun (Malta), Gharb and Sannat (Gozo island). The traps were placed near cattle (four farms), cattle and sheep (one farm: Rabat) and sheep and goats (one farm: Mellieha). Culicoides midges were found in 91.66% (77/84) of the catches and the highest number of midges per catch was Culicoides imicola was confirmed on Malta for the first time in October 2002 and was found at four sites (San Gwann, Sannat, Gharb and Mellieha) but at very low abundance levels (<0.1% of the total Culicoides collected). Culicoides paolae Boorman, 1996 was the most widespread and abundant species (more than 80% of total Culicoides). Midges of the Obsoletus Complex were rare, with less than 10 individuals captured. Other species of Culicoides identified in the collections were: C. submaritimus Dzhafarov, 1962, C. cataneii Clastrier, 1957, C. circumscriptus Kieffer, 1918, C. jumineri Callot and Kremer, 1969, C. kingi Austen, 1912, C. maritimus Kieffer, 1924 and C. newsteadi Austen, Keywords Bluetongue Culicoides Culicoides imicola Culicoides paolae Entomological surveillance Malta. Introduction The Maltese islands are located in the middle of the Mediterranean Basin, where bluetongue (BT) disease occurred recently in several countries. Significant populations of C. imicola have been discovered in Italy (4) and Corsica (3). Culicoides imicola is the principal vector of BT virus (BTV) in this area but occurs fragmentarily across the Mediterranean Basin. Little data are available on the Maltese Culicoides fauna, and C. imicola has never been reported to occur there (1). A survey for Culicoides was organised on the islands in 2002 to establish whether Culicoides imicola Kieffer, 1913 (or any other suspected vector species) was present. The authors report on and discuss the results of collections made between May and October Materials and methods Six farms were chosen as permanent sites in the districts of Mellieha, Rabat, San Gwann, Zejtun (Malta island), Gharb and Sannat (Gozo island) (Fig. 1). The traps were placed near cattle (four farms), cattle and sheep (one farm: Rabat) and sheep and goats (one farm: Mellieha). A total of 84 catches was made between May and October (Table I). The collections and analyses were performed in 278 Veterinaria Italiana, 40 (3), 2004

233 Figure 1 Location of the collection sites on the Maltese islands, 2002 Table I Sites positive for Culicoides imicola on the Maltese islands, May-October 2002 Collection sites Positive collections of C. imicola/total collections (%) C. imicola/ total Culicoides (%) Gharb 1/12 (8.33) 1/256 (0.39) Mellieha 3/20 (15) 5/10484 (0.048) Rabat 0/16 (0) 0/454 (0) San Gwann 1/8 (12.5) 1/161 (0.62) Sannat 1/16 (6.25) 4/287 (1.39) Zejtun 0/12 (0) 0/414 (0) Total 6/84 (7.14) 11/12056 (0.09) accordance with the protocols of the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche) in Teramo, Italy (5). Results Total Culicoides and total C. imicola for each site are reported in Figure 2. Culicoides midges were found in 91.7% (77/84) of the catches and the highest number of midges per catch was C. imicola was identified on Malta for the first time in October 2002 and was found at four sites (San Gwann, Sannat, Gharb and Mellieha) but at very low abundance levels (<0.1% of total Culicoides collected and always less than five specimens per catch). Culicoides paolae Boorman, 1996 was by far the most widespread and abundant species (more than 80% of total Culicoides) (Fig. 3). Midges of the Obsoletus Complex were rare (<10 individuals were captured). In a single catch (Mellieha in October), a species of the Schultzei Complex was found to be as abundant as C. paolae and was identified as C. kingi Austen, The other species of Culicoides found on Malta were: C. submaritimus Dzhafarov, 1962, C. cataneii Clastrier, 1957, C. circumscriptus Kieffer, 1918, C. jumineri Callot and Kremer, 1969, C. maritimus Kieffer, 1924 and C. newsteadi Austen, Discussion and conclusions The presence of C. imicola indicates that if BTV was introduced, it could circulate on the islands of Malta and Gozo. The absence of comparable historical data on the Culicoides fauna does not enable an assessment to be made as to whether or not the vector has colonised the Maltese Islands recently. Other vectors of orbiviruses (including BTV) also occur on the Maltese Islands and include species of the Obsoletus, Pulicaris and Schultzei Complexes. However, in most instances, their abundances were extremely low. Culicoides paolae has been suspected in the past to feed on horses (and thus classed as a potential vector of arboviruses) (2), but its antennal and palpal morphology would indicate it to feed preferentially on birds (6). This species may prove to be a synonym of the Central American C. jamaicensis Edwards, 1922 (6). It is important to now establish the host preference and larval habitat of C. paolae, since this species is very widespread and abundant on the islands of Malta. Veterinaria Italiana, 40 (3),

234 Mellieha Log midges /06/ /07/ /07/ /07/ /08/2002 Date 19/08/ /09/ /09/ /10/ /10/2002 Gharb Log midges /07/ /07/ /07/ /08/ /08/ /08/ /08/ /09/ /09/ /10/ /10/ /10/2002 Date Rabat San Gwann Log midges Log midges /06/ /06/ /07/ /07/ /08/ /08/ /09/ /10/ /05/ /05/ /06/ /07/ /10/ /10/ /10/ /06/2006 Date Date Sannat Log midges 16/07/ /07/ /08/ /08/ /09/ /09/ /10/ /10/2002 Zejtun Date O Total Culicoides O Culicoides imicola Figure 2 Abundance of Culicoides biting midges on the Maltese islands, 2002 Log midges References 16/05/ /05/ /05/ /06/ /06/ /06/ /06/ /07/ /07/ /08/ /08/ /08/2002 Date Figure 3 Culicoides paolae Boorman, 1996: the most abundant and widespread Culicoides species on the Maltese islands 1. Boorman J., Jennings M., Mellor P.S. & Wilkinson P. (1985). Further data on the distribution of biting midges in southern Europe and the Mediterranean area, with special reference to Culicoides imicola. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, Inc., New York, Boorman J., Mellor P.S. & Scaramozzino P. (1996). A new species of Culicoides (Diptera: Ceratopogonidae) from southern Italy. Parassitologia, 38, Veterinaria Italiana, 40 (3), 2004

235 3. Delécolle J.-C. & de La Rocque S. (2002). Contribution à l étude des Culicoides de Corse. Liste des espèces recensées en 2000/2001 et redescription du principal vecteur de la fièvre catarrhale ovine : C. imicola Kieffer, 1913 (Diptera, Ceratopogonidae). Bull. Soc. Entomol. Fr., 107 (4), Goffredo M., Conte A.M., Cocciolito R. & Meiswinkel R. (2003). The distribution and abundance of Culicoides imicola in Italy. Vet. Ital., 39 (47), Goffredo M. & Meiswinkel R. (2004). Entomological surveillance of bluetongue in Italy: methods of capture, catch analysis and identification of Culicoides biting midges. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Meiswinkel R., Labuschagne K. & Goffredo M. (2004). Christopher Columbus and Culicoides: was C. jamaicensis Edwards, 1922 introduced into the Mediterranean and 500 years later renamed C. paolae Boorman, 1996? In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Veterinaria Italiana, 40 (3),

236 Vet. Ital., 40 (3), Laboratory survival and blood feeding response of wild-caught Culicoides obsoletus Complex (Diptera: Ceratopogonidae) through natural and artificial membranes M. Goffredo, G. Romeo, F. Monaco, A. Di Gennaro & G. Savini Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy Summary In late summer 2002, live wild-caught midges of the Obsoletus Complex were collected using blacklight traps placed at a horse stable in Teramo (Abruzzo, Italy). For the survival study under laboratory conditions, Obsoletus Complex midges were kept at 17 C-25 C and provided only with a sucrose solution. Of these, 150 (10%) survived for at least 40 days and 3 midges were still alive after 92 days. In addition, 10 midges survived 10 days at 4 C. For the feeding trials, 40 bloodmeals (9 440 midges) were administered, 27 of which were successful (67.5%); the feeding rate ranged from 0.3% to 16.7%, with a total of 592 engorged midges. Similar feeding rates (U Mann- Whitney test=129.5 p>0.05) were obtained when natural (day-old chicken skin) and artificial (stretched parafilm) membranes were used. To infect the insects, a field strain of bluetongue (BT) virus (BTV) serotype 2 isolated from the spleen of a sheep during the 2000 Italian outbreak was added to the blood-meal. Two different viral solutions, with titres of 10 6 TCID 50 /ml and 10 7 TCID 50 /ml, were prepared. Uninfected blood was significantly more appetising (U Mann- Whitney test=88.5 p<0.05) than the infected meal and the midges preferred (U Mann-Whitney test=48 p<0.05) to feed on blood containing BTV-2 at a lower titre. A total of 251 midges were fed on BTV-2 infected blood and were then incubated at 23 C-25 C and fed with a sucrose solution for 10 days. During the incubation period, the dead insects were collected daily and analysed for evidence of virus infection. Of the 251 engorged midges, 54 (21.5%) died in the feeding chambers or during sorting on the chill table, 136 died within the first 10 days and 61 survived longer. BTV was isolated only from those which died just after feeding (52.6%; 10/19) or 24 h later (47.8%; 11/23). Considering the small number of midges tested after 10 days of incubation, the prevalence of infection detected in this study (95% probability) would have been higher than 4.74%. These preliminary results appear very promising as this is the first time that midges of the Obsoletus Complex have been successfully fed under laboratory conditions. Keywords Bluetongue Obsoletus Complex Culicoides obsoletus Blood feeding Laboratory survival. Introduction The Obsoletus Complex includes a group of Culicoides species belonging to the subgenus Avaritia (Diptera: Ceratopogonidae). There is strong evidence that the species of the Obsoletus Complex are vectors of bluetongue (BT) virus (BTV). They have been found associated with BT outbreaks in areas where the main vector, C. imicola, is rare or absent (1). They have also been capable of sustaining BTV replication when inoculated intrathoracically (2); perhaps more importantly, BTV serotypes 4, 2 and 9 have been isolated from parous individuals caught during clinical outbreaks (4, 5). Being abundant and widely distributed, they might play an important role in the spread of BTV in Europe. A crucial step to assess the vector competency of a Culicoides species is to isolate the virus 10 days after artificial feeding (incubation period). According to the literature, all attempts to feed Culicoides obsoletus under laboratory 282 Veterinaria Italiana, 40 (3), 2004

237 conditions have been unsuccessful (2, 3). The aim of this study was to develop methods to improve both insect survival rates and feeding under laboratory conditions. Materials and methods In the late summer of 2002, live wild-caught midges of the Obsoletus Complex were collected using blacklight traps placed at a horse stable in Teramo (Abruzzo, Italy). Two years of daily trapping at this site demonstrated that the Obsoletus Complex represented 90%-95% of the total Culicoides (Fig. 1) and that the males present in the catches belonged to at least two species of the group, C. obsoletus (Meigen), 1818 and C. scoticus Downes and Kettle, For the survival study under laboratory conditions, Obsoletus Complex midges were kept at 17 C-25 C and were provided only with a sucrose solution. The feeding trials were performed in accordance to the method described by Venter et al. (6). Forty afternoon or evening blood-meals were given, involving midges ( midges for each meal). Log Culicoides (maximum/month) Nov Jan March 2002 May 2002 July 2002 Month Sept Nov Figure 1 Seasonal dynamics of the Obsoletus Complex population in the collection site, 2002 The midges were fed at least four days after collection and did not receive a sucrose solution in the 24 h that preceded meals. Sheep defibrinated blood was used for feeding. During feeding, the blood-meal was kept at 37 C and agitated using a magnetic stirrer (Fig. 2). For the infection, a BTV serotype 2 field strain isolated from the spleen of an infected sheep during the 2000 BT outbreak in Italy, was added to the blood-meal. Two different viral solutions with titres of 10 6 TCID 50 /ml and 10 7 TCID 50 /ml were prepared. Jan Figure 2 Midges of the Obsoletus Complex feeding through an artificial membrane Day-old chicken skin and stretched parafilm were used as natural and artificial feeding membranes, respectively. After the infected meal, the engorged midges were incubated at 23 C-25 C for 10 days and fed with a sucrose solution (Fig. 3). Dead insects were collected daily and analysed for evidence of virus infection. Figure 3 Infected midges of the Obsoletus Complex fed with a sucrose solution during the incubation period Results Of the midges fed with a sucrose solution alone, 150 (10%) survived for at least 40 days and 3 midges were still alive after 92 days (Fig. 4). This is the longest survival period recorded for Culicoides. In addition, 10 midges survived 10 days at 4 C. Engorged midges of the Obsoletus Complex were obtained in 27/40 (67.5%) blood-meals. Veterinaria Italiana, 40 (3),

238 No. of midges alive Days Figure 4 Number of midges of the Obsoletus Complex that survived at least 40 days under laboratory conditions Similar feeding rates (U Mann-Whitney test =129.5 p>0.05) were obtained when natural and artificial membranes were employed (Table I). In the 27 successful blood-feeding trials, the feeding rate ranged from 0.3% to 16.7%, with a total of 592 engorged midges of the Obsoletus Complex of a total of Table I Blood feeding response of midges of the Obsoletus Complex, fed through natural and artificial membranes Membrane No. successful/ No. total meals (%) Engorged/total midges fed (%) Natural (day-old chicken skin) 7/10 (70) 130/2 000 (6.5) Artificial 20/30 (66.67) (stretched parafilm) 462/7 440 (6.21) Total 27/40 (67.5) 592/9 440 (6.27) Uninfected blood was significantly more appetising (U Mann-Whitney test = 88.5 p<0.05) than the infected blood and midges of the Obsoletus Complex preferred (U Mann-Whitney test = 48 p<0.05) to feed on blood containing BTV-2 at a lower titre (Table II). Table II Blood feeding response of midges of the Obsoletus Complex fed with uninfected and infected meals Sheep blood No. successful/ total meals (%) Engorged/total midges fed (%) No virus 11/13 (84.62) 341/3 030 (11.25) With BTV TCID50/ml (76.47) 204/4 750 (4.29) With BTV TCID50/ml 3/10 (30) 47/1 660 (2.83) Total 27/40 (67.5) 592/9 440 (6.27) Of the 251 midges engorged with infected blood, 54 (21.5%) died in the feeding chambers or during sorting on the chill table. Of the remaining 197 midges, only 61 survived for 10 days (Fig. 5). All the midges that survived were analysed for evidence of virus and BTV was isolated only from those which died within 24 h after feeding (Table III). No. of midges alive Days Figure 5 Number of midges of the Obsoletus Complex that survived the incubation period after oral infection with BTV-2 Table III Bluetongue virus serotype 2 isolation from artificially infected midges of the Obsoletus Complex during a 10-day incubation period Incubation day Positive/analysed midges (%) 0 10/19 (52.6) 1 11/23 (47.8) 2 0/5 (0) 3 0/27 (0) 4 0/39 (0) 5 0/0 (0) 6 0/1 (0) 7 0/7 (0) 8 0/2 (0) 9 0/13 (0) 10 0/61 (0) Total 21/197 (10.66) Discussion and conclusions These results clearly show that females of the Obsoletus Complex, at least those belonging to the species C. obsoletus sensu stricto and C. scoticus, were able to survive for up to 92 days at 17 C-25 C without a blood-meal (Fig. 4). This study also demonstrated that midges of the above species 284 Veterinaria Italiana, 40 (3), 2004

239 recovered with ease after being kept for 10 days at 4 C. This lengthy life-span and resistance to low temperature could play an important role in BTV persistence if their vector competence for BTV is confirmed. After infection, the survival rate could probably be improved if antibiotics, never used in this study, were added to the sugar solution. The blood-feeding response of the wild-caught midges through natural and artificial membranes appeared very promising and this is the first time that midges of Obsoletus Complex have been successfully fed under laboratory conditions. It also appears evident that the presence of BTV and the titre thereof in some way influenced the feeding rate (Table II). The fact that BTV was isolated from engorged midges 24 h after the infected meal would suggest that the virus was still alive in the gut; no virus was isolated from midges which survived 10 days after feeding on an infected meal (Table III). According to the number of midges tested 10 days after feeding, it is predicted that a prevalence of infection of up to 4.74% could be detected (95% probability). Consequently, to better assess the vector competence (if any) of the Obsoletus Complex, the study should now be performed with a larger sample of midges incubated for a period of 10 days. 3. Jones R.H., Schmidtmann E.T. & Foster N.M. (1983). Vector-competence studies for bluetongue and epizootic hemorrhagic disease viruses with Culicoides venustus (Ceratopogonidae). Mosquito News, 43 (2), Mellor P.S. & Pitzolis G. (1979). Observations on breeding sites and light-trap collections of Culicoides during an outbreak of bluetongue in Cyprus. Bull. Entomol. Res., 69, Savini G., Goffredo M., Monaco F., de Santis P. & Meiswinkel R. (2003). Transmission of bluetongue virus in Italy. Vet. Rec., 152 (4), Venter G.J., Hill E., Pajor I.T.P. & Nevill E.M. (1991). The use of a membrane feeding technique to determine the infection rate of Culicoides imicola (Diptera: Ceratopogonidae) for two bluetongue virus serotypes in South Africa Onderstepoort J. Vet. Res., 58, 5-9. References 1. Goffredo M., Conte A.M., Cocciolito R. & Meiswinkel R. (2003). The distribution and abundance of Culicoides imicola in Italy. Vet. Ital., 39 (47), Jennings D.M. & Mellor P.S. (1988). The vector potential of British Culicoides species for bluetongue virus Vet. Microbiol., 17, Veterinaria Italiana, 40 (3),

240 Vet. Ital., 40 (3), The isolation of bluetongue virus from field populations of the Obsoletus Complex in central Italy G. Savini (1), M. Goffredo (1), F. Monaco (1), A. Di Gennaro (1), P. de Santis (1), R. Meiswinkel (1, 2) & V. Caporale (1) (1) Istituto Zooprofilattico Sperimentale, dell Abruzzo e del Molise G. Caporale, via Campo Boario, Teramo, Italy (2) Agricultural Research Council (ARC)-Onderstepoort Veterinary Institute (OVI), Private Bag X05, Onderstepoort 0110, South Africa Summary Between July and September 2002, bluetongue (BT) virus (BTV) serotypes 2 and 9 caused mortalities amongst sheep in the communities of San Gregorio Magno (Salerno, Campania), Laviano (Salerno, Campania) and Carpino (Foggia, Puglia), central Italy. On three of the affected farms, approximately specimens of Culicoides were captured, representing fifteen species. Not a single specimen of the classical Afro-Asiatic BT vector, C. imicola Kieffer, was found; species of the Obsoletus Complex dominated the light-trap collections (90%) and included C. obsoletus (Meigen), C. scoticus Downes and Kettle and C. dewulfi Goetghebuer. Fifty-eight pools of the Obsoletus Complex (excluding C. dewulfi), each numbering 100 individuals per pool, and containing only parous and gravid females, were assayed for virus. BTV serotype 2 (BTV-2) was isolated from three pools (San Gregorio and Carpino) and BTV-9 from one (Laviano). These results indicate clearly that a species other than C. imicola is involved in the current re-emergence of BT in the Mediterranean Basin, but whether this is only C. obsoletus sensu stricto, or only C. scoticus, or both together, has yet to be established. Keywords Bluetongue Bluetongue virus serotype 2 Bluetongue virus serotype 9 Culicoides Italy Obsoletus Complex Virus isolation. Introduction The Mediterranean Basin is currently being affected by the most severe and long-lasting outbreak of bluetongue (BT) ever experienced. Since 1999, at least six countries that had never before reported BT have now reported outbreaks of the disease (12, 13, 14). In the outbreaks that affected Bulgaria in 1999 and 2002, entomological surveys confirmed Culicoides imicola to be absent. Instead, species of the Obsoletus Complex were found to dominate and represented up to 90% of all Culicoides captured (11). Although no field isolations of BT virus (BTV) have been made from the Obsoletus Complex in Bulgaria, it seems reasonable to assume that it was the vector of the virus as it has been isolated once before from this complex in Cyprus (10). Since August 2000 in Italy, the distribution of the many thousands of foci of BT has been found to overlap that of C. imicola (6). However, with each successive season, cases of BT have appeared in areas where C. imicola was either rare, or absent (7). The possibility that another species of Culicoides was involved in the transmission of BT in parts of Italy, prompted us to investigate entomologically outbreaks of BT in areas where C. imicola was not known to occur. Accordingly, three outbreak areas were visited during the 2002 season; the Culicoides captured were identified, agegraded and then assayed for the presence of BTV. The results are reported here. Materials and methods Bluetongue outbreaks Three outbreaks of BT were investigated: two in Campania and one in Puglia (Fig. 1). To enhance the possibility of infected Culicoides being captured, the 286 Veterinaria Italiana, 40 (3), 2004

241 selected flocks had to show severe clinical symptoms of BT. Disease symptoms and morbidity and mortality rates were recorded for each flock. Diagnosis was confirmed by virus isolation from spleen samples and from the blood of dead and infected animals. In addition, blood samples were collected twice, once while the animal presented clinical signs, and again a few weeks later to detect seroconversion. collections were made for Culicoides on the night of the 25 July. On 10 September, 62 days after the first clinical appearance of disease, a post-mortem was performed on two adult sheep and the spleens and lymph-nodes collected for virus isolation. The third flock (flock 3: 180 adult sheep) was located in the region of Puglia in the Province of Foggia near the village of Carpino on the Gargano Peninsula. Clinical signs typically associated with BT disease appeared in the middle of August and by 5 September, 26 animals had died. EDTA blood was collected from six animals that showed heightened temperatures and facial oedema. The spleen and lymph-nodes were obtained from one of the dead animals and serum samples were taken from 16 animals; all samples were treated as described above. Six days later, Culicoides were collected in association with this flock. Laboratory studies Culicoides imicola Absent Present Figure 1 Distribution of Culicoides imicola in Italy, showing areas (blue circles) in which bluetongue virus was isolated from midges of the Obsoletus Complex The red line delimits the area in which C. imicola has been found (8) The first flock selected (flock 1) was in the municipality of San Gregorio Magno, in the Province of Salerno, Campania. It comprised 124 sheep (118 adults and 6 lambs) and 20 goats (18 adults and 2 kids). Sheep began showing clinical signs on 1 July 2002 and ten animals died over the next ten days. Ethylene-diaminetetra-acetic acid (EDTA) blood and serum samples were collected from two sick animals and tested against BTV. On the night of 23 July, Culicoides were collected in the sheep paddock using Onderstepoort-type blacklight traps. Seventy days after the first clinical signs of disease were detected in the flock, serum samples were collected from eleven fully recovered animals. The second flock (flock 2: 260 adults and 5 lambs) was located also in the Province of Salerno but in the municipality of Laviano. Here the first signs of clinical BT were reported on 9 July and by 22 July, ten animals had died. EDTA blood and serum samples were collected from each of four sick animals and tested against BTV. Light-trap Virus isolations from blood and tissue samples were performed by intravenous inoculation of embryonating chicken eggs (ECE) followed by repeated passages in Vero cells (4, 17). A reverse transcriptase-polymerase chain reaction (RT-PCR) using group-specific and type-specific primers, was used for the detection of BTV nucleic acid in tissues and in blood samples according to described methods (15). Serum samples were tested for the presence of BTV antibodies using a competitive ELISA (c-elisa) (VMRD, USA) and virus neutralisation (VN) assays (5). Positive and negative controls for the VN assays were kindly provided by the OIE reference laboratory of the Onderstepoort Veterinary Institute (OVI) in South Africa. Studies on insects Onderstepoort-type blacklight traps were used to capture Culicoides according to the method developed by the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche) in Teramo, Italy, in collaboration with the OVI, South Africa (9). At each site, two to four traps were operated for one night. Culicoides were collected into 300 ml of phosphate-buffered saline (PBS) with antibiotics, to which 5 ml of soap detergent had been added. Collections were maintained on ice during transport to the laboratory of the Istituto Zooprofilattico Sperimentale in Teramo, and stored there at 4 C. All Culicoides were age-graded in chilled PBS according to the method of Dyce (3). Parous and gravid categories, which comprise the older individuals that had already had one or more bloodmeals prior to capture, were separated out as these Veterinaria Italiana, 40 (3),

242 are the only ones likely to yield BTV. Care was taken to not include freshly bloodfed individuals. Pools of 100 (mostly parous) individuals of species belonging to the Obsoletus Complex were sorted for virus isolation. These pools comprised a mixture of C. obsoletus and C. scoticus; specimens of C. dewulfi were excluded. For the isolation of virus, insects were processed as described previously (16). Pools were homogenised in 100 µl of cold Dulbecco s minimum Eagle s medium (DMEM) using a battery powered microtissue grinder. Homogenised samples were centrifuged at g for 10 min. and the supernatants collected into 1.5 ml microfuge tubes. One hundred µl of DMEM containing antibiotics (penicillin 100 IU/ml, streptomycin 100 µg/ml, gentamycin 5 µg/ml and nystatin 50 IU/ml) and 20% foetal calf serum (FCS) were added. Ten µl of each insect pool was diluted 1:10, and then three additional ten-fold dilutions were prepared. Twentyfive µl of undiluted sample and 25 µl of each dilution were put into each of four wells of a 96 flatbottomed well plate to which was added 100 µl of Vero cell suspension in DMEM containing approximately cells/ml. The final concentration of FCS in the total volume (125 µl) was 5%. The inoculated microplates were incubated at 37 C in 5% CO 2 and observed microscopically for cytopathic effect (CPE) every day for 6 to 7 days post-inoculation. On day 7, the tissue culture fluid was collected from wells showing CPE and inoculated onto a two-day-old monolayer of Vero cells. The second passage of the virus was collected at a time when 90%-100% of the cells showed CPE and then used for virus identification. Immunofluorescence (IF) using BTV monoclonal antibodies (VMRD, USA) was used as the confirming assay. Virus characterised as BTV was subsequently typed by virus microneutralisation assays using type-specific antisera. In the absence of CPE, cultures were scraped, the cells centrifuged, and the supernatant re-passaged for two blind passages. At the end of the third passage, cells were checked by IF for the presence of BTV. The remaining volume of each individual insect homogenate was also tested using the intravenous inoculation of embryonating chicken eggs followed by two passages in Vero cells as previously described. The BTV infection rate (and its confidence limits) in the midge populations collected were calculated using a Bayesian approach. The virus isolation data were analysed using the Beta (s+1, n-s+1) distribution where s, the number of successes, is the total number of positives and n, the number of trials, is the total number of tested individuals. As BTV was recovered from pools of 83 to 100 individuals, the minimum and maximum infection prevalence rates were determined. Results Clinical, serological and virological examinations All flocks suffered clinical disease and deaths. Clinical signs persisting over two months in flock 2 included fever, depression, nasal discharge, facial oedema, hyperaemia and ulceration of the oral mucosa, coronitis, muscle weakness, diarrhoea and cachaessia. The mortality rate averaged 12% in the first two flocks (11.8% in flock 1 and 12.3% in flock 2), and approached 17% in the third flock. Subsequent virological and serological studies confirmed the presence of BTV in all three flocks (Tables I and II). Table I Bluetongue virus isolations from blood and organ samples from three infected sheep flocks in central Italy, July-September 2002 Flock No. Sample RT-PCR (positive/ No. tested) Method ECE (positive/ No. tested) Bluetongue virus serotype 1 Blood 2/2 0/2 2 Spleen 2/2 0/2 Blood 4/5 0/5 3 Spleen 1/2 1/2 BTV-2 Lymphnodes 0/1 0/1 Blood 6/6 0/6 RT-PCR reverse transcriptase-polymerase chain reaction ECE embryonating chicken eggs Culicoides identification Approximately midges, representing fifteen species, were captured on the affected farms. Not a single specimen of the classical Afro-Asiatic BT vector, C. imicola was found. Instead, other species of the subgenus Avaritia dominated the light-trap collections, and included C. obsoletus (Meigen), C. scoticus Downes and Kettle and C. dewulfi Goetghebuer, these comprising 90% of all Culicoides captured. In addition, more than 95% of the older parous and gravid females (these being the only ones likely to harbour virus) belonged to the Obsoletus Complex only. Male specimens indicated that two species of this complex were present: C. obsoletus sensu stricto and C. scoticus. Owing to the fact that the female of C. scoticus is inseparable from that of C. obsoletus, these two species were combined during 288 Veterinaria Italiana, 40 (3), 2004

243 Table II Serological results from three bluetongue-infected sheep flocks in central Italy, July-September 2002 Flock No. Date of sampling c-elisa (positive/no. tested) Method Virus neutralisation (positive/no. tested) BTV-2 BTV-4 BTV-9 BTV-16 1 July /4 2/4 0/4 2/4 0/4 September /1 0/1 0/1 0/1 0/1 2 July /2 0/2 0/2 2/2 0/2 September /11 5/11 0/11 4/11 0/11 3 September /16 3/3 0/3 0/3 0/3 the selection of pools for BTV isolation. C. dewulfi comprised 2% of the collections; these were excluded from the pools assayed for virus. Bluetongue virus isolation from insects Fifty-six pools (of mostly 100 parous individuals each) were sorted for virus isolation. Twenty-eight pools derived from flock 1, 26 from flock 2 and 2 (each containing 83 individuals) from flock 3. Four isolations of BTV (three of BTV-2 and one of BTV-9) were made. All isolates were recovered by ECE (7.1%) but only two by tissue culture (TC) inoculation (3.6%) (Table III). BTV-2 was found in Culicoides collected around flocks 1 and 3, whereas BTV-9 was recovered from those collected around flock 2 (Table III). The estimated BTV infection prevalence rates in midges are shown in Table IV. Discussion To determine the vector status of any arthropod for an arbovirus, it has to be demonstrated that it can become infected with the virus after feeding on a viraemic host (or on an artificial substitute) and that it is able to then transmit the virus biologically to a healthy host by bite after an extrinsic viral incubation period of one week or more. Furthermore, the virus should be recovered from wild-caught specimens whose abdomens are free of visible fresh blood; also, there should be field evidence confirming the association between the arthropod and the diseaseinfected vertebrate host (1). However, before it can be classified as a proven vector of BTV, conclusive proof of the ability of the arthropod to transmit BTV to a vertebrate host is required also via the establishment of an infection by bite. Whilst the latter requirement is imperative towards establishing the vector competency of species of the Obsoletus Complex, it remains indisputable that in this study most of the above vector status criteria were fulfilled, i.e. BTV was recovered from parous individuals belonging to species of the Obsoletus Complex collected in the vicinity of infected flocks (that had received no new animals during the preceding year) and in which BTV infection had been demonstrated clinically, serologically and virologically. Furthermore, almost half the specimens of the Obsoletus Complex captured in the outbreaks were parous, demonstrating that a significant number of midges were surviving sufficiently long enough to ingest, replicate and then transmit BTV some 7 to 14 days later. As 97% of all the parous midges captured belonged to one or two species of the Obsoletus Complex, it is almost certain that they were the only species transmitting the virus locally. Table III Bluetongue virus isolations from 56 pools of Culicoides of the Obsoletus Complex in central Italy, July-September 2002 Flock No. Method Bluetongue virus serotype (positive/no. pool tested) ECE TC BTV-2 BTV-4 BTV-9 BTV /28 0/ /26 1/ /2 1/2 1 Total 4/56 2/ ECE embryonating chicken eggs TC tissue culture Veterinaria Italiana, 40 (3),

244 Table IV Estimated bluetongue infection prevalence rates in Culicoides of the Obsoletus Complex captured on affected farms in central Italy, July-September 2002 Flock No. BTV serotype No. of pools* positive/examined No. of midges examined No. of possible positive individuals (min./max.) 95% confidence limits of prevalence estimate (%) 1 2 2/ / , /2** / Total 9 1/ and 9 4/ * each pool comprises 100 individuals ** two pools of 83 individuals Observations in Bulgaria and Croatia indicate C. obsoletus to be the principal or only vector of BTV-9 in this part of Europe (14). Our study suggests that species of the Obsoletus Complex are able to transmit more than one serotype of BTV in Europe, and, furthermore, that it may involve more than one species of the complex. As to the epidemiology of BT in Europe, the isolation of the virus from the Obsoletus Complex should be viewed with some concern. Culicoides imicola, the proven classical vector of BTV, is restricted to the southern parts (mostly) of only some countries that adjoin the northern border of the Mediterranean Sea. The converse is true for species of the Obsoletus Complex containing C. obsoletus, one of the commonest Culicoides to occur across central and northern Europe and where it is known to attack both livestock and man (2). Essentially its distribution only seldom penetrates into that of C. imicola but this area of overlap may enable species of the Obsoletus Complex to spread BTV further north. Stated simply, C. imicola remains the principal vector involved in the rapid and explosive incursions of BT that periodically affect the Mediterranean Basin but in carrying the virus northwards hands it over to a member of the Obsoletus Complex: the baton effect (11). Italy falls within this zone of vector overlap and thus is one of the Mediterranean countries most at risk to incursions by BT as both vectors are now known to occur widely there and in occasional sympatry (Figs 1 and 2). This implies that in areas of intense infection, both vectors may have been involved and there may have been many instances in which the Figure 2 Distribution of Culicoides obsoletus in Italy, showing areas (red circles) in which bluetongue virus has been recovered from pools of midges comprising both C. obsoletus and C. scoticus baton effect could have occurred (6, 8). It has not been established whether this spread occurs only from areas of overlap or whether a species of the Obsoletus Complex, in the absence of C. imicola, is able to carry the virus northwards and to then sustain outbreaks across a season or more. In view of the location of the affected farms (Figs 1 and 2) it would appear that a species of the Obsoletus 290 Veterinaria Italiana, 40 (3), 2004

245 Complex was indeed able to carry the virus and to sustain outbreaks in the absence of C. imicola. References 1. Anon. (1967). Arbovirus and human disease. Technical Report Series 369. World Health Organization, Geneva, 22 pp. 2. Campbell J.A. & Pelham-Clinton E.C. (1960). A taxonomic review of the British species of Culicoides Latreille (Diptera, Ceratopogonidae). Proc. R. Soc. Edin., Sec. B (Biol.), 67, Dyce A.L. (1969). Recognition of nulliparous and parous Culicoides (Diptera: Ceratopogonidae) without dissection. J. Aust. Entomol. Soc., 8, Foster N.M., Jones R.H. & Luedke A.J. (1968). Transmission of attenuated and virulent bluetongue virus with Culicoides variipennis infected orally via sheep. Am. J. Vet. Res., 29, Gard G.P. & Kirkland P.D. (1993). Bluetongue virology and serology. In Australian standard diagnostic techniques for animal diseases (L.A. Corner & T.J. Bagust, eds). CSIRO Information Services, Melbourne, Goffredo M., Satta G., Torina A., Federico G., Scaramozzino P., Cafiero M.A., Lelli R. & Meiswinkel R. (2001). The 2000 bluetongue virus (BTV) outbreak in Italy: distribution and abundance of the principal vector Culicoides imicola Kieffer. In Proc. Tenth International Symposium of the American Association of Veterinary Laboratory Diagnosticians (AAVLD), Salsomaggiore, Parma, 4-7 July. AAVLD, Ames, Goffredo M., Conte A.M., Cocciolito R. & Meiswinkel R. (2003). The distribution and abundance of Culicoides imicola in Italy. Vet. Ital., 39 (47), Goffredo M., Conte A. & Meiswinkel R. (2004). Distribution and abundance of Culicoides imicola, Obsoletus Complex and Pulicaris Complex (Diptera: Ceratopogonidae) in Italy. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Goffredo M. & Meiswinkel R. (2004). Entomological surveillance of bluetongue in Italy: methods of capture, catch analysis and identification of Culicoides biting midges. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium. Taormina, October Vet. Ital., 40 (3), Mellor P.S. & Pitzolis G. (1979). Observations on breeding sites and light-trap collections of Culicoides during an outbreak of bluetongue in Cyprus. Bull. Entomol. Res., 69, Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean Basin Vet. J., 164, Office International des Épizooties (OIE) (2000). Dis. Info., 13 (oie.int/eng/info/hebdo/a_isum.htm accessed on 5 September 2004). 13. Office International des Épizooties (OIE) (2001). Dis. Info., 14 (oie.int/eng/info/hebdo/a_isum.htm accessed on 5 September 2004). 14. Office International des Épizooties (OIE) (2002). Dis. Info., 15 (oie.int/eng/info/hebdo/a_isum.htm accessed on 5 September 2004). 15. Shad G., Wilson W.C. & Everman J.F. (1997). Bluetongue virus detection: a safer reverse transcriptase-polymerase chain reaction for prediction of viraemia in sheep. J. Vet. Diag. Invest., 9, Venter G.J., Paweska J.T., van Dijk A.A., Mellor P.S. & Tabachnick W.J. (1998). Vector competence of Culicoides bolitinos and C. imicola for South African bluetongue virus serotypes 1, 3 and 4. Med. Vet. Entomol., 12, Wechsler S.J. & McHolland L.E. (1988). Susceptibilities of 14 cell lines to bluetongue virus infection. J. Clin. Microbiol., 26, Veterinaria Italiana, 40 (3),

246 Vet. Ital., 40 (3), Seasonal abundance of Culicoides imicola and C. obsoletus in the Balearic islands M.A. Miranda, C. Rincón & D. Borràs Department of Biology, University of the Balearic Islands, Cra. Valldemosa km 7.5, CP:07122, Palma de Majorca, Spain Summary An outbreak of bluetongue (BT) was declared on the Balearic islands of Spain in September and October In 2001 and 2002, an intensive survey was conducted on cattle farms in Majorca and Minorca for the principal vectors in the Mediterranean Basin, Culicoides imicola and C. obsoletus. Adult Culicoides were collected once a week between June 2001 and December 2002 using CDC light traps. The results from 348 light-trap collections revealed that, in addition to other species of the genus Culicoides, both C. imicola and C. obsoletus appear to be well established on both Majorca and Minorca. Furthermore, both species showed a different seasonal abundance pattern: peak adult populations of C. obsoletus occurred in July, while those of C. imicola peaked in October. These findings indicate that the principal vector in the outbreak of BT in the Balearic islands in 2000 was probably C. imicola. Keywords Balearic islands Bluetongue Culicoides Majorca Minorca Spain. Introduction The disease of ruminants known as bluetongue (BT) is caused by a virus of the genus Orbivirus (family Reoviridae) and in the Mediterranean Basin is mainly transmitted by the species of biting midge Culicoides imicola Kieffer and C. obsoletus (Meigen) (Diptera; Ceratopogonidae) (1, 6, 7). Recently, another species, C. pulicaris (Linnaeus), has also been incriminated as a vector following the isolation of BT virus (BTV) from field specimens (3). There are 24 serotypes of the virus distributed worldwide. Of these, at least six (1, 2, 4, 6, 9 and 16), occur in the Mediterranean Basin. In Europe, BT mainly affects sheep, but cattle and goats could act as reservoirs of the virus, without showing clinical signs of the disease. There have been previous incursions of the disease into the Mediterranean Basin, for example into Cyprus and Israel in 1943 and into the Iberian Peninsula between 1956 and 1960, causing high mortalities in sheep. However, in the last seven years ( ), BT has spread massively around the Mediterranean Basin (17), affecting countries including Algeria, Bulgaria, France Greece, Italy, Morocco, Spain, Tunisia and Turkey. In the case of Spain, the recent outbreak of BT was detected in September 2000 in the Balearic islands (9), firstly on Majorca, and then on Minorca. The virus involved was serotype 2, the same as that identified in Corsica, Sardinia and Sicily and which probably originated from Tunisia where it appeared in Subsequently, serotype 9 was also identified on mainland Italy, but was never to reach the Balearic islands. Since BT is one of the List A diseases of the Office International des Épizooties (OIE), the declaration of the epizootic had important consequences on livestock farming in the Balearic islands, especially in Minorca, since movement of animals amongst the islands was prohibited and also because many animals were slaughtered in an effort to stop the disease spreading further. During the first outbreak, the number of foci totalled 391 in Majorca and 114 in Minorca (M.J. Pourtau and M.J. Rubio, personal communication) thus demonstrating the efficiency with which the vectors transmitted the virus from 292 Veterinaria Italiana, 40 (3), 2004

247 one animal to the next over a short period of time. Accordingly, an intensive survey of the major vectors was conducted on cattle farms. The results demonstrated, for the first time, that C. imicola and C. obsoletus occur on Majorca and Minorca (8). The former has been reported recently in other areas in which previous surveys had not revealed its presence, for example on some islands of Greece and in Italy (7). This may suggest that C. imicola has been spreading into new areas of the Mediterranean Basin in the past eight years. The seasonal abundance of vectors has been broadly correlated with the transmission of the virus. In this sense, transmission is enhanced under optimal climatological conditions for virus replication, vector development and activity. Large numbers of C. imicola and C. obsoletus have been linked to the recent BT outbreaks recorded around the Mediterranean Basin, as well as with previous BTV and African horse sickness virus (AHSV) incursions in Portugal and southern Spain. In this study, data is reported for the first time on the seasonal abundance of C. imicola and C. obsoletus in the Balearic islands. At the time of writing, October 2003, a new outbreak of BT was detected in Minorca. Material and methods Midges were captured on four cattle farms in Majorca and on three in Minorca (Table I); the farms were selected in collaboration with the Conselleria d Agricultura i Pesca (Government of the Balearic islands). All farms were situated in the area affected by BT during the outbreak of Table I Location of farms and number of animals involved in the study of Culicoides vectors on the islands of Majorca and Minoca, Spain, Island Site name Location No. of animals (bovines) Majorca Ca n Centes 39º27 N 3º14 E 90 Ca n Roig Nou 39º29 N 3º05 E 111 Sa PlanaVella 39º27 N 3º16 E 155 Ses Veles 39º39 N 3º26 E 261 Minorca Algaiarens 40º03 N 3º57 E 125 Son Gornes 39º58 N 4º00 E 107 Cases Noves 40º 02 N 4º05 E 77 A CDC (Centers for Disease Control) UV 4W light trap provided with a suction fan (blacklight model 912, John Hock Company) was placed outside stables at a height of between 1.7 and 2 m from the ground. Traps were powered by 220 V using a transformer and all were located close to where the livestock remained during the night (less than 7 m from the animals). Collections were made one night per week from June 2001 to December The traps were operated from one hour before sunset to one hour after sunrise. Insects collected were transported to the laboratory in the same trap mesh and preserved in 70% ethanol. Culicoides species were separated from other insects. C. imicola was classified by using the wing pattern, and confirmed by microscopic examination using the methods described by Wirth and Marston (16) and Delécolle (4). Due to the difficulty encountered in the differentiation of C. obsoletus females from other similar species (such as C. scoticus) these were classified, using the wing pattern, as the C. obsoletus group. Results and discussion A total of 348 light-trap collections were made, 242 from Majorca and 106 from Minorca. C. imicola and species of the C. obsoletus group were captured on all the farms sampled and on both islands. The total number of C. imicola found on Majorca was (25.5 adults per night per trap), whereas on Minorca the total was 595 (5.6 adults per night per trap). For the C. obsoletus group, (21.45 adults per night per trap) were collected on Majorca and 590 (5.6 adults per trap per day) on Minorca. Of the total Culicoides captured on Majorca, C. imicola comprised 37%, C. obsoletus 32.3%, and other Culicoides spp. 30.7%. On Minorca, the respective data were 9.3%, 9.2% and 81.5%. Results showed that C. imicola is the predominant species on Majorca, but its numbers did not differ significantly from those of C. obsoletus. However, on Minorca, they clearly are not the predominant species; instead C. newsteadi Austen, represented a high percentage of the captures (data not shown). We also found important differences between the abundances of C. imicola and C. obsoletus on Majorca and Minorca. Different farming practices, or differing environmental conditions, should be investigated to find the explanation. If low populations of both vectors are confirmed on Minorca, this would clearly show that low numbers of vector midges are sufficient to cause epizootics of BT. The low predominance of C. imicola and C. obsoletus found in this survey differed from other studies conducted in the Iberian Peninsula during the AHS Veterinaria Italiana, 40 (3),

248 epizootics that occurred in the 1990s, when C. imicola clearly predominated (10, 11). Recently, extensive surveys performed in BTV-affected areas, such as mainland Italy, Sicily, Sardinia, Corsica and some of the Greek islands, have demonstrated a high prevalence of this vector, despite the fact that it had been proved absent during previous surveys. It is widely accepted that the area of spread of C. imicola has been increasing in recent years (7). Furthermore, due to its spread into the Iberian Peninsula, the geographic range of this species has recently been extended on account of the presence of a single female in Catalonia (15); at this site, C. obsoletus was the predominant species. On the other hand, C. imicola has not been detected in other areas of the Mediterranean; for example, in Bulgaria and in parts of mainland Greece which have experienced BT epizootics, and where the predominant species were C. obsoletus and C. pulicaris (7). Regarding the seasonal abundance of both species, periods with high numbers of adults were observed in both years, indicating that C. imicola and C. obsoletus are well established in the Balearic islands. Since no previous surveys had been conducted in this area, it is impossible to establish whether C. imicola has been introduced recently. The seasonal abundance of vectors has important implications in the efficiency of virus transmission and the occurrence of outbreaks. It is recognised that the peak in vector population determines the optimum period for transmission (5). The seasonal abundances of C. imicola and of the C. obsoletus group on Majorca and Minorca are shown in Figures 1 and 2, respectively. Our results from 2001 and 2002 indicate that the seasonal abundance of C. imicola was highest during September and October, whereas C. obsoletus showed a period of peak abundance between April and July. High abundances of C. imicola during the summer and autumn months were also reported in different areas within the Iberian Peninsula, for example southern Spain (11, 12, 13) and Portugal (2). The C. obsoletus group was also recorded in high numbers during spring in southern Spain (12, 13) and in Catalonia (15). We also found that the seasonal occurrence of C. obsoletus extended longer than that of C. imicola. Adults of C. obsoletus were captured almost throughout the year, confirming the observations of Rawlings and Mellor (14). This could be explained by the fact that C. obsoletus is a cooladapted Palaearctic species, whereas the seasonal and geographic occurrence of C. imicola, an Afrotropical species, is limited by lower temperatures when compared to C. obsoletus (14). Mean adults/night/trap Jun-01 Aug-01 Oct-01 Dec-01 Feb-02 Apr-02 Jun-02 Aug-02 Oct-02 Dec-02 Date Culicoides imicola Culicoides obsoletus Other Culicoides Figure 1 Seasonal fluctuation of Culicoides imicola, C. obsoletus and Culicoides spp. adult populations on the island of Majorca, Spain Mean adults/night/trap Jun-01 Aug-01 Oct-01 Dec-01 No data available Feb-02 Apr-02 Jun-02 Aug-02 Oct-02 Dec-02 Date Culicoides imicola Culicoides obsoletus Other Culicoides Figure 2 Seasonal fluctuation of Culicoides imicola, C. obsoletus and Culicoides spp. adult populations on the island of Minorca, Spain Results obtained in this study indicate that the BT outbreak that occurred in the Balearic in 2000 coincided with maximum abundances of C. imicola indicating it to have been the principal vector at that time. C. imicola has been also linked to the recent BT outbreaks in mainland Italy, Sardinia, Sicily and Corsica. However, in areas such as mainland Greece and Bulgaria where C. imicola has not been detected, C. obsoletus appears to be the main vector, although some laboratory studies have indicated this species to be a less efficient vector (5). Furthermore, in previous studies conducted in the Iberian Peninsula 294 Veterinaria Italiana, 40 (3), 2004

249 and in relation to the epizootics of AHS, C. imicola was also the predominant species in the months of September and October (11). Considering data for 2002 alone, C. imicola was first detected on Majorca in April and persisted until December; for the C. obsoletus group, the first specimens were captured in January with activity ceasing in November. The situation is unknown for Minorca as insufficient light-trap collections were made. These preliminary results indicate that proven vectors of BTV, such as C. obsoletus, overwinter in the Balearic islands and probably also C. imicola, depending on the severity of the winter. The possibility of C. obsoletus overwintering in the Iberian Peninsula was mooted earlier by Rawlings and Mellor (14). Acknowledgements We would like to thank the staff of Sanitat Animal in the Conselleria d Agricultura i Pesca (Goverment of the Balearic islands) and the Consell Insular de Menorca for providing us with the insect collections, especially to Antonio Riera, Carlos Aranga, Alexandra, Maria José Rubio, Federico Martín, Lucía Tascón, Cristina Ramos and Maria José Portau. This study was supported by the Instituto Nacional de Investigación Agropecuaria (INIA) and the Conselleria d Agricultura i Pesca. References 1. Boorman J., Jennings M., Mellor P.S. & Wilkinson P. (1985). Further data on the distribution of biting midges in southern Europe and the Mediterranean area, with special reference to C. imicola. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, Inc., New York, Capela R., Sousa C., Pena I. & Caeiro V. (1993). Preliminary note on the distribution and ecology of C. imicola in Portugal. Med. Vet. Entomol., 7, Caracappa S., Torina A., Guercio A., Vitale F., Calabro A., Purpari G., Ferrantelli V., Vitale M. & Mellor P.S. (2003). Identification of a novel bluetongue virus vector species of Culicoides in Sicily. Vet. Rec., 153, Delécolle J.-C. (1985). Nouvelle contribution à l étude systématique et iconographique des espèces du genre Culicoides (Diptera: Ceratopogonidae) du nord-est de la France. PhD dissertation. Université Louis Pasteur de Strasbourg, UFR des sciences de la vie et de la terre, Strasbourg, 238 pp. 5. Mellor P.S. (1992). Culicoides as potential orbivirus vectors in Europe. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Press, Mellor P.S., Jennings D.M., Wilkinson P.J. & Boorman J.P. (1985). Culicoides imicola: a bluetongue virus vector in Spain and Portugal. Vet. Rec., 116, Mellor P.S., Boorman J. & Baylis M. (2000). Culicoides biting midges: their role as arbovirus vectors. Ann. Rev. Entomol., 45, Miranda M.A., Borràs D., Rincón C. & Alemany A. (2003). Presence of Culicoides imicola and Culicoides obsoletus in the Balearic islands. Med. Vet. Entomol., 17, Office International des Epizooties (2001). Bluetongue in Spain. Dis. Info., 13, Ortega M.D. & Holbrook F.R. (1994). Presence of Culicoides imicola (Diptera: Ceratopogonidae) in Jaen, Spain. J. Am. Mosquito Control Assoc., 10, Ortega M.D., Lloyd J.E. & Holbrook F.R. (1997). Seasonal and geographical distribution of Culicoides imicola Kieffer (Diptera, Ceratopogonidae) in southwestern Spain. J. Am. Mosquito Control Assoc., 13, Ortega M.D., Mellor P.S., Rawlings P. & Pro M.J. (1998). The seasonal and geographical distribution of Culicoides imicola, C. pulicaris group and C. obsoletus group biting midges in central and southern Spain. Arch. Virol., 14, Ortega M.D., Holbrook F.R. & Lloyd J.E. (1999). Seasonal distribution and relationship to temperature and precipitation of the most abundant species of Culicoides in five provinces of Andalusia, Spain. J. Am. Mosquito Control Assoc., 15, Rawlings P. & Mellor P.S. (1994). African horse sickness and the overwintering of Culicoides spp. in the Iberian peninsula. Rev. Sci. Tech. Off. Int. Épiz., 13, Sarto i Monteys V. & Saiz-Ardanaz M. (2003). Culicoides midges in Catalonia (Spain), with special reference to likely bluetongue virus vectors. Med. Vet. Entomol., 17, Wirth W. & Marston N. (1968). A method for mounting small insects on microscope slides in Canada balsam. Scientific Notes, 61, Wittmann E.J., Mellor P.S. & Baylis M. (2001). Using climate data to map the potential distribution of Culicoides imicola (Diptera: Ceratopogonidae) in Europe. Rev. Sci. Tech. Off. Int. Épiz., 20, Veterinaria Italiana, 40 (3),

250 Vet. Ital., 40 (3), Multiple vectors and their differing ecologies: observations on two bluetongue and African horse sickness vector Culicoides species in South Africa R. Meiswinkel (1), K. Labuschagne (2), M. Baylis (3) & P.S. Mellor (3) (1) Instituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, via Campo Boario, Teramo, Italy (2) Onderstepoort Veterinary Institute, Private Bag X05, Onderstepoort 0110, South Africa (3) Institute for Animal Health, Pirbright Laboratory, Pirbright, Surrey GU24 0NF, United Kingdom Summary Blacklight traps were used to collect Culicoides biting midges weekly between September 1996 and August 1998 at 40 sites distributed equidistantly across South Africa. The seasonal and geographic prevalences of 86 species of Culicoides were elucidated simultaneously, and included C. imicola Kieffer and C. bolitinos Meiswinkel the principal vectors of bluetongue (BT) and African horse sickness (AHS) in the region. These two species were amongst the most prevalent Culicoides to be found and, together, comprised >50% of the more than three million biting midges captured. The data are presented as coloured matrices, and are transformed also into inverse distance weighting (IDW) interpolative maps. The data reveal that the prevalence of each vector is somewhat fractured and it is posited that this is (in part) due to significant differences in their respective breeding habitats. The results illustrate also that the presence of multiple vectors (in any region of the world) will complicate both the epidemiology of the orbiviral diseases they transmit and the formulation of rational livestock movement and disease control strategies. This is especially true for southern Europe where the recent devastating cycle of BT has been shown to involve at least three vectors. Finally, the influence that man has on the development of large foci of vector Culicoides around livestock may be less important than previously suggested but must be investigated further. Keywords Culicoides vectors Culicoides imicola Culicoides bolitinos Seasonal distribution Geographic distribution Interpolative maps South Africa Vector. Introduction Pantropically some 30 species of Culicoides biting midges are involved in the transmission of orbiviral diseases injurious to livestock; these diseases include bluetongue (BT) and African horse sickness (AHS). Because of their negative economic impact there is a great need to model the distributions of the insect vectors involved, not only for disease control, but also for predictive purposes. To obtain data that are reliable, modellers depend heavily upon an accurate taxonomy as each vector species, in occupying a specific niche in nature, possesses a unique biology that determines its distribution in both space and time. However, accurate large-scale species-linked data sets are surprisingly few for Culicoides. Equally cogent is that these diseases are, in all regions, being transmitted by more than one vector species. Not only does this complicate disease epidemiology, but also invalidates (partially) single vector risk models. As to the involvement of multiple vectors, South Africa is a case in point: 60 years were to pass before it was realised that C. imicola is not the only vector of BT and AHS (5) but that C. bolitinos is also involved (11). To assess objectively whether the relative abundances of these two species, and their seasonal and geographic prevalences, differed significantly, a network of 40 blacklight traps was established and operated weekly across South Africa from September 1996 to August The data reveal the two species to occur widely but profound differences 296 Veterinaria Italiana, 40 (3), 2004

251 in their respective ecologies impact significantly on their local prevalences and abundances. These data are used also to strengthen the argument that an unresolved taxonomy at the base of the decisionmaking pyramid will hobble the development of rational disease management strategies at the apex. Materials and methods Blacklight traps of the type described by Venter and Meiswinkel (14) were used to capture Culicoides. The choice of the forty sites sampled across South Africa (Fig. 1) was determined by: 1) the need for equidistant monitoring 2) the necessary presence of livestock (of any species that included cattle, horses, sheep, goats and/or pigs) 3) the collection site being near (1-50 km) to an automatic weather station (AWS) maintained by the South African Weather Bureau 4) the need to sample a diverse range of habitats i.e. one (or two) sites in each of the 10 climatic zones of South Africa, and two (or more) sites in each of the seven broad ecological or vegetative zones. 1 Welkom 21 George 2 Venterstad 22 Struisbaai 3 De Aar 23 Porterville 4 Graaff Reinet 24 Springbok 5 Port Elizabeth 25 Alexander Bay 6 Elliot 26 Calvinia 7 Port Edward 27 Beaufort West 8 Mtunzini 28 Middelburg 9 Vryheid 29 Lydenburg 10 Greytown 30 Komatipoort 11 Giant s Castle 31 Phalaborwa 12 Ficksburg 32 Tshipese 13 Vrede 33 Pietersburg 14 Lichtenburg 34 Thabazimbi 15 Kuruman 35 Ellisras 16 Upington 36 Onderstepoort 17 Prieska 37 Vereeniging 18 Kimberley 38 St Lucia 19 Fort Beaufort 39 Mabula 20 Joubertina 40 Laingsburg Figure 1 Culicoides sampling sites in South Africa, September 1996-August 1998 Light-trap collections were made weekly by the livestock owner whose co-operation was solicited first and, once obtained, he/she would be trained on site and provided immediately with the entire Culicoides collection kit. To promote sustained collaboration, owners were paid upon receipt of their light-trap collections (at six-weekly intervals) at the Onderstepoort Veterinary Institute (OVI). Culicoides species identifications were performed using a wingpicture atlas developed at the OVI over the last 20 years. Listed vertically in Figures 2 and 3 are 38 of the 40 sites sampled; the week number is listed horizontally (commencing 11 September 1996 [week 37] and ending in the first week of September 1997 [week 36]). In these figures the weekly abundances of the two vector species are depicted for the first year of the survey only, and are converted to log(n+1); these abundances are colour-coded to aid rapid visual discernment of their geographical and seasonal prevalences: white (no collection made), green (0 Culicoides), yellow (1-9 Culicoides), orange (10-99), red ( ), purple ( ) and dark blue ( ). The data were used to produce also the interpolation maps (Figs 4 and 5) with Mapinfo 7,5 s inverse distance weighting (IDW) interpolator employing a distance-weighted average of data points to calculate grid cell values. Raw data were the greatest total catch of each species (males and females) made at any given site in a single night between October 1996 and September The IDW was undertaken with an exponent of 2 and a search radius of 200 km; the exponent was chosen arbitrarily, its low value increasing the influence of distant data points. The search radius was chosen to be about half of the distance between the two most distant neighbouring sites (400 km), thereby ensuring that all sites are subject to the influence of at least one other. Results Of a projected collections, (81%) were eventually made, and included a total of >3 million Culicoides. These represented 86 species i.e. 70% of the >120 species of Culicoides known to occur in South Africa. The ten most abundant species (which together comprised 90% of the midges collected) were: C. imicola Kieffer (47.5%), C. subschultzei Cornet and Brunhes (9.8%), C. magnus Colaço (7.7%), C. zuluensis de Meillon (6.9%), C. bolitinos Meiswinkel (4.1%), C. pycnostictus Ingram and Macfie (3.9%), C. leucostictus Kieffer (3.6%), C. nivosus de Meillon (3.0%), C. schultzei Enderlein (2.4%) and C. enderleini Cornet and Brunhes (1.3%). These data show that the two principal vectors of BT and AHS, C. imicola and C. bolitinos, comprised >50% of all the Culicoides captured; they were also amongst the most prevalent Veterinaria Italiana, 40 (3),

252 Giant s Castle Mtunzini Port Edward Hellsgate George Vryheid Elliot Ficksburg Vrede Vereeniging Port Elizabeth Middelburg Lydenburg Komatipoort Onderstepoort Warmbad Pietersburg Phalaborwa Thabazimbi Ellisras Kuruman Struisbaai Porterville Joubertina Fort Beaufort Lichtenburg Welkom Tshipese Kimberley Prieska Laingsburg Venterstad Calvinia De Aar Upington Springbok Beaufort West Alexander Bay Figure 2 Culicoides imicola: log(n+1) numbers captured weekly at 38 livestock holdings spread across South Africa between September 1996 (week 37) and September 1997 (week 36) Sites are ranked from highest to lowest rainfall Giant s Castle Mtunzini Port Edward Hellsgate George Vryheid Elliot Ficksburg Vrede Vereeniging Port Elizabeth Middelburg Lydenburg Komatipoort Onderstepoort Warmbad Pietersburg Phalaborwa Thabazimbi Ellisras Kuruman Struisbaai Porterville Joubertina Fort Beaufort Lichtenburg Welkom Tshipese Kimberley Prieska Laingsburg Venterstad Calvinia De Aar Upington Springbok Beaufort West Alexander Bay Figure 3 Culicoides bolitinos: log(n+1) numbers captured weekly at 38 livestock holdings across South Africa between September 1996 (week 37) and September 1997 (week 36) Sites are ranked from highest to lowest rainfall species, being found at 39 and 38 of the 40 sites sampled, respectively. In Figures 2 and 3 the collection sites are ranked according to average rainfall ranging from >1 200 mm/annum (sites 1-3: Giant s Castle to Port Edward) to <250 mm/annum (site 38; Alexander Bay). The pattern reveals C. imicola (Fig. 2) to occur most abundantly in those areas where the annual rainfall averages between mm. This embraces largely the sweetveld grazing areas favoured for the raising of livestock, and previously famed for their teeming hordes of game animals (wildebeest, zebra, etc.). This pattern of prevalence suggests that the larvae of C. imicola, being soilinhabiting, favour a breeding substrate that is not only moisture-retentive but one which is also rich in micro-nutrients; it is known that at annual average rainfall levels above 700 mm, such soils will become leached of nutrients, and so may help explain why C. imicola is virtually absent from the high rainfall Interpolated maximum catch Figure 4 Inverse distance weighting interpolated maximum catches of Culicoides imicola, using a search radius of 200 km Interpolated maximum catch Figure 5 Inverse distance weighting interpolated maximum catches of Culicoides bolitinos, using a search radius of 200 km equatorial regions of Africa. Judging from the more randomised scatter of the high abundance collections of C. bolitinos (Fig. 3) it would appear that higher annual average precipitation levels influence less the prevalence of this cattle dung-inhabiting species. Culicoides imicola was most abundant in the warmer northern and eastern areas of South Africa (sites Middelburg, Komatipoort and Tshipese); such areas have long been considered enzootic for orbiviral diseases, and so may act as source points for their southward spread. However, C. imicola was found abundantly also in the southern half of the country (Kimberley, Joubertina, Porterville); the latter two sites occur in the Cape Province where devastating outbreaks of AHS have occurred both in the past (1, 6) with smaller incursions more recently (8). Although C. bolitinos was as widespread as C. imicola, it was an order of magnitude less abundant. Also, and unlike C. imicola, it tended to predominate in cooler, higher-lying areas (site 12: Ficksburg), near where it has recently been linked unequivocally to significant outbreaks of AHS (11), and where BT can 298 Veterinaria Italiana, 40 (3), 2004

253 recur almost annually. Of importance is that the latter areas mostly abut those in which C. imicola predominates, and so this partial allopatry in the geographic distributions of these two vector species effectively increases the total area at risk to orbivirus transmission in South Africa. At a number of sites it was found that both species peaked in abundance and prevalence around week 12 (the end of March) which is mid-autumn in the southern hemisphere, and which marks (roughly) the annual commencement of the BT and the AHS seasons. One prerequisite for a bloodsucking invertebrate to be classed as a vector is that its period of heightened abundance (and hence biting activity) should correlate with the appearance of the relevant disease in the vertebrate host that it may attack. Judging from the data both C. imicola and C. bolitinos fulfil this requirement. During this period, the parity rates of these two species were also at their highest seasonal levels (data not shown) i.e. the proportion of older females was found to peak also in the late summer/autumn months, and across the entire range of populations sampled. These older females are the ones to most likely harbour virus, and so transmit it to their ruminant hosts. The dual presence of two vectors with partially overlapping distribution patterns would suggest that during outbreaks the virus will (or could) be transmitted more widely. It is a common belief that areas experiencing the coldest winters (including snow) are also those with the smallest populations of C. imicola. This is generally true (Giant s Castle and Elliot sites) but, unexpectedly, C. imicola was found to be even rarer (or entirely absent) at three sites (Port Elizabeth, Struisbaai and Alexander Bay) which possess a milder, frost-free, climate. These latter three sites occur along the southern and western coastline, and so, intuitively, one might be tempted to attribute the local absence of C. imicola at these sites to excessive windiness. However, the dominance of the second vector, C. bolitinos, at Port Elizabeth (site 5) would seem to mitigate against wind being an important inhibitor of Culicoides activity. Thus, it must be considered whether the sandiness of the soil (which will promote the rapid drainage of moisture from its top layer) is not the primary factor disrupting larval development in C. imicola locally (9, 10); if so this single edaphic negative variable would cancel the effect of the more immediately obvious positive variables (such as a frost-free climate, rainfall of 600 mm/annum, and abundance of livestock) and which would normally support the development of large populations of C. imicola at any given site. As hinted above, soil-type does not seem to act as a barrier to C. bolitinos due to its predilection for cattle dung as a larval habitat. It is this predilection that enables C. bolitinos to penetrate into sandy areas, and also into steeply-sloped, mountainous terrain where C. imicola is unable to persist (in this instance presumably because of the aridification of the soils top layer due to water runoff). The interpolation maps (Figs 4 and 5) show (broadly) that C. imicola is more abundant than C. bolitinos almost everywhere (except around Lesotho, down to Port Edward and across to Port Elizabeth). They show also that C. imicola is present in two major clusters the Cape Province and the north/north-east/east of South Africa; there is also a central cluster around Kimberley (site 18). As noted earlier this correlates well with the pattern of historical outbreaks (especially of AHS) in South Africa. Unfortunately, it means also that the AHS surveillance and protection zone, which is situated in the Western Cape Province (area south of site 23), can expect to be periodically affected by outbreaks of the disease, and is a supposition confirmed by the recent incursions of AHS into this zone in 1999 and 2004 (8). The interpolation maps show clearly that the hyper-arid north-western corner of South Africa (sites 24-26) are virtually vector free, and so would serve the country better as both export and quarantine zones. The geographic distribution of these two vectors was found to be stable across the two seasons sampled (data not shown for the second season) and so might be harnessed as a predictive tool in risk analysis. However, there are important caveats that may profoundly influence modelling; some are natural cyclical phenomena, others are artificial (or maninduced). For example, C. imicola is the only species of afrotropical Culicoides known to develop extraordinarily large populations during the episodic rains that affect South Africa every 10 to 15 years (10) correlating perfectly with the old adage that horse sickness appears as a plague following heavy rains (13). Not only must such climatic oscillations be factored into predictive risk maps, but, taken a step further, they imply also that the constant irrigation of grazing pastures (coupled with the maintenance of fenced livestock) not only interrupt natural wet-dry cycles but, also, may be causally linked to the establishment of extremely large (and artificial) foci of C. imicola locally. In the case of C. bolitinos, cattle husbandry may be inducing similar modifications to its natural distribution across South Africa. This study revealed that vector-free areas do exist in South Africa. One of these sites (Struisbaai) was surveyed because historical records (1) showed it to have escaped the ravages of the largest outbreak of Veterinaria Italiana, 40 (3),

254 AHS that occurred in South Africa 150 years ago. Although Struisbaai was monitored weekly, not a single specimen of C. imicola was captured amongst 15 other species of Culicoides during 140 weeks of uninterrupted sampling. Can the absence of C. imicola be ascribed solely to the local soils being both sandy and calcareous? Soil type has been cited to not only explain the absence of C. imicola at specific points along the southern margin of its range (9, 10) but also at its northern end (4) in the region of Puglia, Italy, and where the prevailing climatic conditions would seem to favour its presence (3). These eyecatching local gaps in the distribution of C. imicola indicate strongly that this vector is not able to penetrate into all landscape zones (neither in the long term nor under changing climatic conditions). Although ameliorating climate change will have little foreseeable impact on local edaphic conditions, it would still facilitate the northward movement of C. imicola even if such a progression was to occur only patchily. Whatever the eventuality, if the C. imicola-free zones discovered in this survey were to be characterised precisely, and their distribution mapped across the Old World range of C. imicola, then vector-free enclaves could be identified. Whilst these would be of inestimable value for the quarantining and export of livestock, the absence of other competent vector Culicoides would still have to be assured. From Figure 3 it can be seen that the second vector, C. bolitinos, does occur in low numbers at Struisbaai and so introduces an element of risk. It is possible that C. bolitinos can be controlled (or even eliminated) locally if all cattle were to be removed. For C. imicola, however, its control through environmental management would appear to be well-nigh impossible to achieve. This is because it breeds in moist (and clay-type soils) which become widely available following periods of rainfall, and also under sustained irrigation, which is practised extensively on many thousands of farms throughout South Africa. Conclusions With objective monitoring an immense amount of valuable data on the distribution and seasonality of Culicoides can be collected quite rapidly from across an extensive geographic transect. Furthermore, if the taxonomy of the local Culicoides fauna is resolved, then data on 50 species or more will be obtained simultaneously. Not only are such data indispensable for predictive modelling, but they arm us also with a more precise understanding of the ecological niche occupied by each species. These can help us to track changes (if any) in the dispersal pattern of any given species over time or to monitor changing prevalence amplitudes under various climatic circumstances. Had a similar surveillance system been present in Europe, there would be a much stronger evidence base for assessing whether C. imicola has recently spread across the Mediterranean region. Another hypothesis to emerge recently is that the distribution of C. imicola is determined not only by climate but also (and fundamentally) by soil type as it has been found to be consistently absent from sandy (and calcareous) areas which appear otherwise suitable, i.e. which possess abundant livestock and an equable climate. For C. bolitinos, the scenario is different: its local prevalence does not seem linked to soil but more intimately to that of cattle, buffalo and wildebeest (their dung being essential to the development of its immature stages). Thus, while some climatic variables may favour equally the adult activity preferences of both C. imicola and C. bolitinos, their geographic distributions can still differ dramatically due to their vastly different breeding habitats. These differences should prompt modellers to incorporate a separate range of breeding habitat variables because, stated simply, if the larvae are not able to mature then no adult midges can emerge to fly. The IDW maps, though broadly accurate, have shortcomings. For example, C. imicola from being abundant at one site can disappear at another over a fairly short distance as illustrated by the respective data (Fig. 2) collected at Struisbaai (site 22) and at Porterville (site 23). Although this heterogeneity was captured on the IDW map (Fig. 4), sampling should be performed at a finer scale (of about 50 km) to improve the resolution at a more localised level. However, this would increase the number of trapping stations (and subsequent laboratory analyses) six-fold. Clearly then, these maps are crude and show the greater value of using satellite imagery with scales down to 8 km or 1 km, and using those types of satellite data that correlate with the insect catches (using climate and soil proxies). Such a satellite map has been produced for C. imicola in South Africa (2) but was found also to have a limited predictive capacity probably because not all factors influencing the local distribution of C. imicola had been considered. Ultimately, it is necessary that the distributional data on insect vectors be highly detailed as livestock owners, especially those involved in the regular movement of animals, require information that is applicable (and reliable) at the farm level. In an AHS- and BT-endemic region such as South Africa this implies that each farmer should consider assessing the Culicoides vector situation on his/her holding, rather than relying on the broader fuzziness of current risk maps. 300 Veterinaria Italiana, 40 (3), 2004

255 Previously (9, 10), it was mooted that man, by husbanding livestock in confined spaces ( sedentary bloodbanks ), and on grazing pastures that are irrigated constantly, helps establish and maintain localised foci of both C. imicola and C. bolitinos. If so, it forces one to consider whether the currently mapped distribution of these two vectors are therefore somewhat artificial, and if this is the case, how then should man s influence be calibrated for its factoring into predictive risk models? It is not possible to provide the answer here. All that may be said is that the present data would indicate that the role of man should perhaps not be exaggerated because although both vector species were found to occur almost throughout South Africa, there was enormous geographic variability in their respective abundance levels and seasonal prevalence patterns despite the universal presence of livestock at all trapping sites. Thus we can fairly safely interpret man s impact to be but minor when compared to that exerted by climate and by soil. Ultimately it is the presence of the insect vector that places a given area at risk to disease incursion and maintenance. As illustrated above, the ecologies of C. imicola and C. bolitinos differ markedly, and so increases the total area at risk. Furthermore, these patterns of vector prevalence will never be entirely annectant but will overlap; this applies particularly to southern Europe where the distribution of C. imicola, to a varying extent, overlaps that of the more northerly additional vector species of the Obsoletus and Pulicaris species complexes (7). This zone of overlap will serve to facilitate the further movement (and maintenance) of the virus and so complicates the development of rational livestock movement and disease control strategies. Multiple vectors exist in nearly all regions of the world affected by Culicoides-borne orbiviral diseases. Added to this is the concern that other, and more prevalent non-vector species, may possess a nascent vectorial competence that will emerge only in the future under changing climatic conditions. These possibilities warn us that vector surveillance studies should be planned with a view to mapping the distributions of all species of Culicoides simultaneously, and especially in those instances where the disease transmission potentials of suspected vectors still remain unknown. This is especially true for the Mediterranean Basin, where the taxonomy of the 100 or more species of Culicoides is incompletely resolved (12), and where an enormous body of work (both in the laboratory and in the field) is still needed to understand which species are most competent at transmitting the various diseases that threaten the livestock industry there. As long as the taxonomy, and the precise ecologies, of vector Culicoides in Europe remain unclear, we can expect to make only limited progress in terms of predictive risk modelling and in the development of rational control strategies. References 1. Bayley T.B. (1856). Notes on the horse-sickness at the Cape of Good Hope, in Saul Solomon & Co., Steam Printing Office, Cape Town. 2. Baylis M., Meiswinkel R. & Venter G.J. (1999). A preliminary attempt to use climate data and satellite imagery to model the abundance and distribution of Culicoides imicola (Diptera: Ceratopogonidae) in southern Africa. J. Sth Afr. Vet. Ass., 70, Baylis M., Mellor P.S., Wittmann E.J. & Rogers D.J. (2001). Prediction of areas around the Mediterranean at risk for bluetongue by modelling the distribution of its vector using satellite imaging. Vet. Rec., 149, Conte A., Ippoliti C., Calistri P., Pelini S., Savini L., Salini R., Goffredo M. & Meiswinkel R. (2004). Towards the identification of potential infectious sites for bluetongue in Italy: a spatial analysis approach based on the distribution of Culicoides imicola. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Du Toit R.M. (1944). The transmission of bluetongue and horse-sickness by Culicoides. Onderstepoort J. Vet. Sc. Anim. Ind., 19, Edington A. (1893). Report of the Colonial bacteriological Institute for the year 1892, Part II. W.A. Richards & Sons, Government Printers, Cape Town. 7. Goffredo M., Conte A. & Meiswinkel R. (2004). Distribution and abundance of Culicoides imicola, Obsoletus Complex and Pulicaris Complex (Diptera: Ceratopogonidae) in Italy. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Koekemoer J.J.O., Paweska J.T., Pretorius P.J. & van Dijk A.A. (2003). VP2 gene phylogenetic characterization of field isolates of African horsesickness virus serotype 7 circulating in South Africa during the time of the 1999 African horsesickness outbreak in the Western Cape. Virus Res., 93, Meiswinkel R. (1997). Discovery of a Culicoides imicola-free zone in South Africa: preliminary notes and potential significance. Onderstepoort J. Vet. Res., 64, Meiswinkel R. (1998). The 1996 outbreak of African horse sickness in South Africa the entomological perspective. Arch. Virol. [Suppl.], 14, Veterinaria Italiana, 40 (3),

256 11. Meiswinkel R. & Paweska J.T. (2003). Evidence for a new field Culicoides vector of African horse sickness in South Africa. Prev. Vet. Med., 60, Meiswinkel R., Gomulski L.M., Delécolle J.-C., Goffredo M. & Gasperi G. (2004). The taxonomy of Culicoides vector complexes unfinished business. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Theiler A. (1921). African horse sickness (Pestis equorum). Department of Agriculture Science Bulletin No. 19. The Government Printing and Stationery Office, Pretoria. 14. Venter G.J. & Meiswinkel R. (1994). The virtual absence of Culicoides imicola (Diptera: Ceratopogonidae) in a light-trap survey of the colder, high-lying area of the eastern Orange Free State, South Africa, and implications for the transmission of arboviruses. Onderstepoort J. Vet. Res., 61, Veterinaria Italiana, 40 (3), 2004

257 Vet. Ital., 40 (3), Epidemiology and vectors Modelling the distribution of outbreaks and Culicoides vectors in Sicily: towards predictive risk maps for Italy B.V. Purse (1), S. Caracappa (2), A.M.F. Marino (2), A.J. Tatem (3), D.J. Rogers (3), P.S. Mellor (1), M. Baylis (1) & A. Torina (2) (1) Institute for Animal Health, Ash Road, Pirbright, Surrey GU24 0NF, United Kingdom (2) Istituto Zooprofilattico Sperimentale della Sicilia, A. Mirri, Palermo, Italy (3) TALA Research Group, Department of Zoology, University of Oxford, South Parks Road, Oxford OX1 3PS, United Kingdom Summary Vector (911 light-trap catches from 269 sites) and serological surveillance data were obtained during recent bluetongue (BT) outbreaks in Sicily. The distributions of Culicoides vectors are compared with that of bluetongue virus (BTV) to determine the relative role of different vectors in BTV transmission in Sicily. The best climatic predictors of distribution for each vector species were selected from 40 remotely-sensed variables and altitude at a 1 km spatial resolution using discriminant analysis. These models were used to predict species presence in unsampled pixels across Italy. Although Culicoides imicola, the main European vector, was found in only 12% of sites, there was close correspondence between its spatial distribution and that of the 2000 and 2001 outbreaks. All three candidate vectors C. pulicaris, C. newsteadi and C. obsoletus group were widespread across 2002 outbreak sites but C. newsteadi was significantly less prevalent in outbreak versus nonoutbreak sites in Messina and BTV has been isolated from wild-caught adults of both C. pulicaris and C. obsoletus in Italy. The yearly distribution and intensity of outbreaks is attributable to the distribution and abundance of the vectors operating in each year. Outbreaks were few and coastal in 2000 and 2001 due to the low abundance and prevalence of the vector, C. imicola. They were numerous and widespread in 2002, following hand-over of the virus to more prevalent and abundant novel vector species, C. pulicaris and C. obsoletus. Climatic determinants of distribution were species-specific, with those of C. obsoletus group and C. newsteadi predicted by temperature variables, and those of C. pulicaris and C. imicola determined mainly by normalised difference vegetation index (NDVI), a variable correlated with soil moisture, vegetation biomass and productivity. The predicted continuous presence of C. pulicaris along the Appenine mountains, from north to south Italy, suggests BTV transmission may be possible in a large proportion of this region and that seasonal transhumance between C. imicola-free areas should not generally be considered safe. Future distribution models for C. imicola in Sicily should include non-climatic environmental variables that may influence breeding site suitability such as soil type. Keywords Bluetongue Climate Culicoides imicola Culicoides obsoletus Culicoides pulicaris Discriminant analysis Distribution Risk map Sicily Vector. Introduction Since 1998, an unprecedented epidemic of bluetongue (BT) has affected many countries in the Mediterranean Basin (east and west) including regions where the main vector, Culicoides imicola, is absent (6). This indicates that novel vector species are involved in BT virus (BTV) transmission, and recently, BTV has been isolated from wild-caught adults of C. pulicaris in Sicily (15) and from those of the C. obsoletus group in other regions of Italy (28). Although present in all nine BT-affected regions of Italy (11), C. imicola is abundant only in Sardinia and Veterinaria Italiana, 40 (3),

258 eastern Calabria (12, 19). In Sicily, one of the first regions of Italy to be affected by BT (13 October 2000), C. imicola is particularly scarce. Given these factors and the isolations described above, Sicily is particularly suitable for the investigation of the relative role of different Culicoides vector species. In collaboration with the National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche) in Teramo, a vector and serological surveillance programme across the entire landmass of Sicily was established in These data are used here to quantitatively compare vector distributions with that of the virus to investigate the relative role of these vectors in BTV transmission in Sicily. In addition, it is essential to model the relationship between environmental factors and the distribution of different vectors (3, 4, 5) so that risk-maps can be formulated for unsampled regions on the basis of their environment. Although all Culicoides species share the same basic habitat requirements, i.e. presence of host for blood-meals and breeding sites for egg and larval development (22), they differ in their life-history characteristics, and in turn, the extent to which their distribution and abundance is affected by environmental factors. The distributions of different vectors in Sicily are modelled in relation to satellite-derived climate variables and the climatic determinants of distribution are compared between vectors. These relationships are extrapolated to predict the distribution of these vectors across Italy. Materials and methods From 2000 to 2002, a total of 911 light-trap collections were made using Onderstepoort-type blacklight traps in 269 trap-sites across Sicily between May and December (1 to 88 catches per site). Individuals of C. imicola were counted and the presence of C. obsoletus group (hereafter referred to as C. obsoletus) and C. pulicaris, C. newsteadi and C. circumscriptus species scored in each sample. The analyses of distribution are restricted to the summerautumn period since, in Europe, this period corresponds with outbreaks of Culicoides-borne disease and peak Culicoides abundance (25). Maximum catches in this period are also consistently related to the annual abundance of Culicoides across sites in Morocco (2). For C. imicola, the analysis was further restricted to 767 catches from 248 sites sampled between July and December since this species reaches peak numbers later than the other species. Serological and sentinel surveillance, i.e. sampling of unvaccinated sentinel bovines, began across all Sicilian provinces in November 2000 and is on-going (14). The geographical co-ordinates of trap sites and outbreaks were determined using a Garmin GPS 12 receiver (those of 16 outbreaks were obtained from websites (1, 23). Altitudes were derived from the 1 km 1 km spatial resolution global topography (GTOPO30) digital elevation model (17). Four variables of environmental significance were considered in the development of climate models of species distribution, namely: normalised difference vegetation index (NDVI), middle infra-red reflectance (MIR), land surface temperature (LST) and air temperature (TAIR). NDVI specifically measures chlorophyll abundance and light absorption, but is correlated with soil moisture, rainfall and vegetation biomass, coverage and productivity (13). MIR is correlated with the water content, surface temperature and structure of vegetation canopies (9). LST is a general index of the apparent environmental surface temperature (whether soil or vegetation) and TAIR is an estimate of the air temperature a few metres above the land surface (18). The method by which these four variables are obtained from 1 km 1 km spatial resolution Pathfinder AVHRR imagery (17) and fourierprocessed to summarise their seasonal variations (producing 40 remotely sensed variables and altitude in all) are given elsewhere (26, 29). The discriminant analysis procedure used to select (by forward stepwise selection) the 10 environmental variables that best divide trap sites into presence-absence classes for each species, and then to predict species presence in unsampled pixels is described by Rogers (27). Four methods of internal validation were used for assessment (see Results ). Results and discussion Relationship between the distribution of C. imicola and the distribution of outbreaks Between 2000 and 2002, 74 outbreaks of bluetongue were reported in Sicily (Fig. 1). In the first two years, there were few outbreaks (11 in 2000, 6 in 2001) and these were restricted to coastal regions (<9 km from the sea). In 2000, they occurred in the provinces closest to Sardinia. In 2001, outbreaks began in Messina and Siracusa, over 175 km from the 2000 outbreak sites. This suggests there was a second incursion of virus in 2001 rather than overwintering of the virus in the 2000 outbreak sites by continual host-to-midge cycling. In 2002, outbreaks were much more numerous and widespread, occurring in five provinces and further inland (>40 km from the sea). They commenced in sites near (<12 km) the outbreak sites of the previous year, suggesting that 304 Veterinaria Italiana, 40 (3), 2004

259 the virus probably overwintered between 2001 and Maximum catch of C. imicola BT outbreak areas Presence of BT outbreaks Figure 1 Distribution of bluetongue outbreaks each year in Sicily Key to provinces: TP = Trapani; PA = Palermo; ME = Messina; EN = Enna, CT = Catania; AG = Agrigento; CL = Catanisetta; RG = Ragusa; SR = Siracusa Culicoides imicola occurred in only 12% of trap sites (mostly in coastal areas <30 km from the sea) and generally in low abundances (Fig. 2, Table I). However, it was the only vector found in all 2000 and 2001 outbreak sites and was particularly abundant in Palermo and Siracusa. In stepwise logistic regression models comparing sites in which seroconversion to BTV occurred in 2000 to those in which seroconversion did not occur, the presence of C. imicola was the only species presence variable added to the model (model χ 2 =10.6, p=0.001, 1 d.f.). Thus, C. imicola was probably the vector responsible for most BTV transmission in the first two years of the outbreak in Sicily. However, it was trapped at only a few 2002 outbreak sites in Paterno, Catania. Thus, in 2002, other vectors must have been responsible for BTV transmission. Table I Maximum catches and prevalence of different vector species across different subsets of trap sites in Sicily, 2002 Species All sites (sampled May to December) Prevalence No. (%) of sites 2002 outbreak sites n = Messina outbreak sites n =18 C. imicola 33 (12) 8 (19) 0 (0) C. obsoletus 161 (60) 24 (57) 13 (72) C. pulicaris 103 (38) 22 (52) 14 (78) C. newsteadi 176 (65) 25 (60) 13 (72) C. circumscriptus 138 (51) 13 (31) 7 (39) Figure 2 Distribution of Culicoides imicola across sites in Sicily from catches made between June and December, from 2000 to 2002 Relationship between the distribution of potential novel vector groups and the distribution of outbreaks Culicoides pulicaris, C. obsoletus and C. newsteadi were widespread across trap sites in Sicily including inland and high altitudes sites (Fig. 3, Table I) and were all present in over 50% of the 2002 outbreak sites. However, the four vector species were rarely present alone in a trap site, complicating the detection of statistical relationship between the distribution of one particular candidate vector and that of outbreaks. For example, C. obsoletus was the sole vector species in only 6% of sites positive for this species whilst it was found with two or more other potential vectors in 41% of such sites. Thus, no species presence variables were added to a stepwise logistic regression model comparing 2002 outbreak and non-outbreak sites. Culicoides newsteadi, being less prevalent in outbreak versus non-outbreak sites in Messina (χ 2 = 4.8, 1 d.f., p =0.028), was the least likely candidate vector species. Relationship between vector distribution of potential novel vector groups and environmental variables The climatic determinants of distribution differ between Culicoides species in Sicily, as indicated by the differing rank and order of climatic variables added to the presence-absence models (Table II). The proportion of correct predictions from the presence-absence models were high (range: 60% to 96% across species) and kappa values of around 0.6 indicated substantial agreement between observed and predicted species presence (Table III). Veterinaria Italiana, 40 (3),

260 Culicoides circumscriptus Culicoides obsoletus Culicoides pulicaris Culicoides newsteadi Absent Present Figure 3 Distribution of Culicoides circumscriptus, C. obsoletus, C. pulicaris and C. newsteadi across sites in Sicily from catches made between May and December, from 2000 to 2002 Table II The first four of ten variables, ranked in order of importance, which best allocated the Sicilian trap sites to the observed Culicoides species presence-absence classes Rank C. obsoletus K C. newsteadi K C. pulicaris K C. imicola K 1 LST mean 0.24 MIR mean 0.26 NDVI mean 0.42 LST tri-ann. phase LST variance 0.32 LST ann. amp NDVI tri-ann. amp MIR tri-ann. phase MIR bi-ann. amp TAIR min NDVI variance 0.50 NDVI ann. phase TAIR min LST min LST variance 0.48 NDVI mean 0.49 K kappa value for the model at the step when the variable was added LST land surface temperature NDVI normalised difference vegetation index MIR middle infrared reflectance TAIR air temperature x m above ground DEM altitude derived from digital elevation model amp. amplitude ann. annual bi-ann. bi-annual tri-ann. tri-annual Table III Kappa coefficients, sensitivity, specificity and consumer accuracy for each species presence-absence model Validation method C. obsoletus C. newsteadi C. pulicaris C. imicola Sensitivity (a) Specificity (b) Consumer accuracy (c) Kappa-coefficient a) percentage of positive observations predicted correctly, equivalent to the producer s accuracy b) percentage of negative observations predicted correctly c) percentage of predictions observed to be correct Kappa coefficient (estimate of the agreement between two variables, accounting for the degree of overlap expected by chance) Key to levels of agreement indicated by kappa values from Landis and Koch (20): K= slight; K = fair; K = moderate; K = substantial; K = near perfect 306 Veterinaria Italiana, 40 (3), 2004

261 The distributions of C. obsoletus and C. newsteadi were primarily related to remotely-sensed temperature variables in Sicily since mostly LST or RAIR variables were added to their distribution models (and made up 5 and 7 of 10 variables added respectively). For C. obsoletus, both median LST and minimum TAIR was slightly higher in absence than presence sites (LST medians: 43.1 C versus 42.9 C, W = 20121, p = 0.01; TAIR medians: 16.0 C versus 14.3 C; W = 19744, p = 0.001) whilst the variance in LST was slightly lower (medians 9.2 C versus 9.6 C; W = 13097, p = 0.01). This preference for warmer, less variable temperature regimes may reflect the fact that C. obsoletus is a northern Palaearctic species (i.e. primarily found in Europe, North Asia and North Africa) on the southern margin of its range in Sicily. It is probably adapted to cold, requiring relatively low temperatures for optimal development and survival. For C. newsteadi, the mean MIR (medians 42.7 versus 43.4, W = 9957, p<0.001), the minimum TAIR sites (medians 13.3 C versus 15.5 C; W = , p<0.001) and the minimum LST (medians 40.3 versus 40.7, W = 10105, p<0.001) were all higher in presence versus absence sites. This association with high values of MIR and minimum temperatures may reflect high thermal requirements for development in this southern Palaearctic species. NDVI variables were the most important determinants of the distribution of C. pulicaris (six NDVI variables were added to the model). The mean NDVI was much higher (medians 0.34 versus 0.43; W = , p<0.001) and the variance in NDVI was lower (medians versus 0.006, W = 11374, p<0.001) in presence versus absence sites. This indicates that C. pulicaris prefers microclimates with high, stable levels of moisture for optimal survival and development. This is consistent with its breeding site requirements which are listed as wet soil and bogs (8). Although an LST variable was the most important predictor of C. imicola distribution, eight other variables added to the model were NDVI or MIR variables, suggesting again that moisture levels as well as temperature are important for survival of this species. The close positive relationship between NDVI and C. imicola distribution is consistent with previous models (3, 5, 7, 29) and arises because microclimate conditions that favour vegetation growth (indicated by high NDVI) also permit breeding of C. imicola. This species breeds in watersaturated soils that are high in organic matter (10). Predicted distributions of C. imicola and potential novel vectors The large differences in the predicted ranges of the different vectors (Fig. 4) have important implications for disease surveillance and control. C. pulicaris was a) Culicoides pulicaris b) Culicoides obsoletus c) Culicoides imicola Figure 4 Predicted distribution of Culicoides species in Sicily and Italy For a and b: from a model derived from the observed presence/absence data for each species at 268 sites in Sicily, sampled between May and October ( ) For c: from a model derived from the observed presence/absence data for C. imicola at 142 sites in Sicily, sampled between July and October ( ) Key: Green = model prediction of species presence Red = model prediction of species absence Grey = no prediction since Mahalanobis distance between a pixel and its assigned probability or class was two or more times greater than the maximum distance observed between any one of the sites in Sicily and the classes to which they belonged Veterinaria Italiana, 40 (3),

262 predicted to be present almost continuously across the Appenine mountains, from north to south Italy (see Fig. 5 for locations of mountains and provinces in Italy). Conte et al. (16) asserted that the identification of mountainous C. imicola-free areas in central Italy could facilitate safer transhumance. Livestock are moved annually at the beginning of summer, from lowland winter pastures to highland areas until late autumn. However, such movement of infected livestock between coastal areas (areas of BTV transmission by C. imicola) to mountainous areas at the time of peak C. pulicaris abundance could provide a mechanism for hand-over of BTV between the main and the novel vector (21). BTV could then spread along the Appenines via passive wind dispersal of midges and cycle through C. pulicaris and host populations. Culicoides obsoletus is predicted to be widespread across Italy, including areas where C. pulicaris is predicted to be absent or rare, i.e. along the Adriatic coast and in Tuscany. If this species is also widely implicated in BTV transmission in Italy and our model predictions of its range are accurate, this further increases the area of Italy at risk of BTV. When compared to the distribution observed by Conte et al. (16), our model correctly predicts the presence of C. imicola in coastal regions of Lazio and Tuscany and along the Ionian cost (Calabria, Basilicata and Puglia). However in Sardinia, the predicted distribution is much patchier than that observed. Whilst this could be due to the disparity in resolution between the two maps (1 km 1 km pixels versus 10 km 10 km pixels) (16), our predicted distribution of C. imicola is generally much less extensive than those derived from previous satellite-driven models (7, 29). Culicoides imicola occurs in a few sites in Sicily and so only a narrow portion of the potential niche of this species (in terms of environmental conditions) was represented in the model training set and, in turn, only a portion of its potential distribution could then be predicted. Models where the training set included a large numbers of sites from a core area of the species distribution perform better when extrapolated to unsampled areas (4). For Sicily, predictions both from satellite-driven climate models and models based on interpolated weather station data predict C. imicola to be much more widely distributed across Sicily than it is observed to be. Thus it is more likely that, in Sicily, the relative absence of C. imicola is caused not by climatic factors but by other environmental factors that may influence its breeding sites. For example, a Figure 5 Map of Italy showing provinces and mountains (closed triangles) 308 Veterinaria Italiana, 40 (3), 2004

263 negative relationship was found between C. imicola abundance and soil sandiness in South Africa (5). Similarly, Calistri et al. (12) suggest that the porous, freely draining volcanic soils with poor moisture content that predominate in Sicily are unsuitable as C. imicola breeding sites. Conclusions and future directions During the 2000 and 2001 outbreaks, BTV appeared to be transmitted primarily by C. imicola and, in 2002 outbreaks, primarily by C. pulicaris with a contribution by C. obsoletus. The operation of different vectors, with different distributions and abundances, produced pronounced differences in the patterns of outbreaks between years. The climatic determinants of distribution are speciesspecific in Sicily and these determinants seem to correlate to some extent with breeding site requirements and species ranges. Thus, predictive risk maps for BT derived entirely from distribution data for the main European vector, C. imicola, will omit extensive regions at risk of transmission via novel vector species. A fuller examination of the relative role of different vectors and their climatic requirements requires models of species abundance rather than presence based on data collected regularly throughout outbreak periods. In the future, the distribution and abundance of these species will be re-modelled using vector surveillance data from a current European Union project collected according to a standardised protocol across a large area of the distribution of all species. Future spatial models of C. imicola distribution in Sicily should include additional environmental variables such as soil type. Acknowledgements The authors would like to thank Gaspare Lo Bue, Rosa Filippi and Nicola Galati for technical assistance. The paper is a composite of two original papers published in Medical and Veterinary Entomology (24, 30). References 1. Alexandria Digital Library (ADL) (2003). Alexandria Digital Library Project, University of California. ADL, Santa Barbara. 2. Baylis M., El Hasnaoui H., Bouayoune H., Touti J. & Mellor P.S. (1997). The spatial and seasonal distribution of African horse sickness and its potential Culicoides vectors in Morocco. Med. Vet. Entomol., 11, Baylis M., Bouayoune H., Touti J. & El Hasnaoui H. (1998). Use of climatic data and satellite imagery to model the abundance of Culicoides imicola, the vector of African horse sickness virus, in Morocco. Med. Vet. Entomol., 12, Baylis M. & Rawlings P. (1998). Modelling the distribution and abundance of Culicoides imicola in Morocco and Iberia using climatic data and satellite imagery. Arch. Virol. [Suppl.], 14, Baylis M., Meiswinkel R. & Venter G.J. (1999). A preliminary attempt to use climate data and satellite imagery to model the abundance and distribution of Culicoides imicola (Diptera: Ceratopogonidae) in southern Africa. J. Sh Afr. Vet. Assoc., 70, Baylis M. & Mellor P.S. (2001). Bluetongue around the Mediterranean in Vet. Rec., 149, Baylis M., Mellor P.S., Wittmann E.J. & Rogers D.J. (2001). Prediction of areas around the Mediterranean at risk of bluetongue by modelling the distribution of its vector using satellite imaging. Vet. Rec., 149, Blanton F.S. & Wirth W.W. (1979). The sandflies (Culicoides) of Florida (Diptera: Ceratopogonidae). Arthropods Fla. Neighb. Land Areas, 10, Boyd D.S. & Curran P.J. (1998). Using remote sensing to reduce uncertainties in the global carbon budget: the potential radiation acquired in the middle infrared wavelenghts. Remote Sensing Rev., 16, Braverman Y., Galun R. & Ziv M. (1974). Breeding sites of some Culicoides species (Diptera: Ceratopogonidae) in Israel. Mosquito News, 34, Calistri P. & Caporale V. (2003). Bluetongue in Italy: a brief description of the epidemiological situation and the control measures applied. Bull. Off. Int. Épiz., No. 2, Calistri P., Goffredo M., Caporale V. & Meiswinkel R. (2003). The distribution of Culicoides imicola in Italy: application and evaluation of current Mediterranean models based on climate. J. Vet. Med., 50, Campbell J.B. (1996). Introduction to remote sensing. Taylor & Francis, London, 622 pp. 14. Caracappa S., Di Bella C., Guercio A., Prato F. & Torina A. (2001). Emergenza bluetongue in Sicilia: contollo del territorio ed attavita di sorveglianza. Bluetongue: Stato dell arte e possibili strategie di contenimento, Universita degli Studi di Perugia. Società Italiana di Patologica Allevamento degli Ovini e del Caprini, Perugia. 15. Caracappa S., Torina A., Guercio A., Vitale F., Calabro A., Purpari G., Ferrantelli V., Vitale M. & Mellor P.S. (2003). Identification of a novel bluetongue virus vector species of Culicoides in Sicily. Vet. Rec., 153 (3), Conte A., Giovannini A., Savini L., Goffredo M., Calistri P. & Meiswinkel R. (2003). The effect of climate on the presence of Culicoides imicola in Italy. J. Vet. Med., 50, Veterinaria Italiana, 40 (3),

264 17. Goddard Space Flight Centre (GSFC) Earth Sciences (GES) (2003). Distributed Active Archive Center, GSFC Earth Sciences, National Aeronautics and Space Administration (daac.gsfc.nasa.gov/ accessed on 21 October 2004). 18. Goetz S.J., Prince S.D. & Small J. (2000). Advances in satellite remote sensing of environmental variables for epidemiological applications. In Remote sensing and geographical information systems in epidemiology (S.I. Hay, S.E. Randolph & D.J. Rogers, eds). Academic Press, San Diego, 47, Goffredo M., Satta G., Torina A., Federico G., Scaramozzino P., Cafiero M.A., Lelli R. & Meiswinkel R. (2001). The 2000 bluetongue virus (BTV) outbreak in Italy: distribution and abundance of the principal vector Culicoides imicola Kieffer. In Proc. Tenth International Symposium of the American Association of Veterinary Laboratory Diagnosticians (AAVLD), Salsomaggiore, Parma, 4-7 July. AAVLD, Ames, Landis J.R. & Koch G.G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33, Mellor P.S. & Boorman J. (1995). The transmission and geographical spread of African horse sickness and bluetongue viruses. Ann. Trop. Med. Parasitol., 89 (1), Mellor P.S., Boorman J. & Baylis M. (2000). Culicoides biting midges: their role as arbovirus vectors. Ann. Rev. Entomol., 45, National Imagery and Mapping Agency (NIMA) (2003). GEOnet names server. NIMA, United States Board on Geographic Names (earthinfo.nga.mil/gns/html/ accessed on 21 October 2004). 24. Purse B.V., Tatem A.J., Caracappa S., Rogers D.J., Mellor P.S., Baylis M. & Torina A. (2004). Modelling the distributions of Culicoides bluetongue virus vectors in Sicily in relation to satellite-derived climate variables. Med. Vet. Entomol., 18, Rawlings P., Pro M.J., Pena I., Ortega M.D. & Capela R. (1997). Spatial and seasonal distribution of Culicoides imicola in Iberia in relation to the transmission of African horse sickness virus. Med. Vet. Entomol., 11, Rogers D.J., Hay S.I., Packer M.J. & Wint G.R.W. (1997). Mapping landcover over large areas using multispectral data derived from NOAA-AVHRR: a case study of Nigeria. Int. J. Remote Sensing, 18, Rogers D.J. (2000). Satellites, space, time and the African trypanosomiases. In Remote sensing and geographical information systems in epidemiology (S.I. Hay, S.E. Randolph & D.J. Rogers, eds). Academic Press, San Diego, 47, Savini G., Goffredo M., Monaco F., de Santis P. & Meiswinkel R. (2003). Transmission of bluetongue virus in Italy. Vet. Rec., 152 (4), Tatem A.J., Baylis M., Mellor P.S., Purse B.V, Capela R., Pena I. & Rogers D.J. (2003). Prediction of bluetongue vector distribution in Europe and North Africa using satellite imagery. Vet. Microbiol., 97 (1-2), Torina A., Caracappa S., Mellor P.S., Baylis M. & Purse B.V. (2004). Spatial distribution of bluetongue and its vectors in Sicily. Med. Vet. Entomol., 18, Veterinaria Italiana, 40 (3), 2004

265 Vet. Ital., 40 (3), Epidemiology and vectors Towards the identification of potential infectious sites for bluetongue in Italy: a spatial analysis approach based on the distribution of Culicoides imicola A. Conte, C. Ippoliti, P. Calistri, S. Pelini, L. Savini, R. Salini, M. Goffredo & R. Meiswinkel Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy Summary A geographic information system (GIS) based on grids was developed by the National Reference Center for Veterinary Epidemiology at the Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale (IZS) in Teramo to identify potential infectious sites for bluetongue (BT) disease in Italy. Geographical and climatic variables were used to build a spatial process model (SPM); the different layers were combined by sequential addition. The final grids (with a cell size of decimal degrees) were generated for each season of the year, and the suitability of each cell for the presence of C. imicola given a value ranking from 0 to 10. While this model more accurately predicts the presence of C. imicola in the Basilicata and Sicily regions, it still over-predicted its presence in the Puglia region. This could be due to the occurrence of calcareous soils which dominate the Puglia landscape. The present SPM is an additive model that assigns an equal weight to each variable. However, the results suggest the existence of hitherto unconsidered variables that significantly influence the prevalence of C. imicola. To reflect their importance, these variables should be assigned a higher weighting in future models. However, the decision in regard to precisely what this weighting should be depends on a very thorough knowledge of the ecology of C. imicola. Keywords Bluetongue Culicoides imicola Geographic information system Italy Spatial model. Introduction A significant amount of research on the distribution of bluetongue (BT) and its principal insect vector C. imicola in the Mediterranean Basin was generated following incursions of the disease into southern Europe between 1998 and Special attention has been focused on modelling the spread of C. imicola and the effect of climatic and geographic factors on its presence (1, 4). The National Reference Center for Veterinary Epidemiology at the Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale (IZS) in Teramo, developed a grid-based geographic information system (GIS) to identify potential infectious sites for BT in Italy through the analysis of areas found most suitable for the persistence of C. imicola. A spatial process model (SPM) (5, 7) was created in an effort to identify areas into which BT virus (BTV) would be most likely to penetrate, but the model is based solely on the principal vector of BT in the Mediterranean, namely C. imicola. The recent incrimination of at least two other species of Culicoides in the transmission of BTV in Italy (3, 6), the ecologies of which differ from that of C. imicola, complicates the epidemiology of the disease, and will require separate spatial analyses in future. Materials and methods The SPM was developed in four stages, as follows: 1) datasets input 2) creation of datasets for the generation of new information 3) reclassification of each dataset to a common scale 4) combining datasets for the identification of suitable locations. The analysis considered the Veterinaria Italiana, 40 (3),

266 four seasons: spring (March, April, May), summer (June, July, August), autumn (September, October, November) and winter (December, January, February). Step 1: The following grids and datasets on Italy were considered to explain the spatial interactions: a) the digitalised elevation model (DEM) built using the isohypse of 1: topographic map of Italy produced by the Military Geographical Institute (IGM) b) land use (Corine Land Cover version 12/2 000 European Environment Agency ( c) the aridity index (European Environment Agency d) the lithologic environment (derived from a 1: map on the humidity retention capacity of soil) e) the monthly normalised difference vegetation index (NDVI) for (Royal Netherlands Meteorological Institute (KNMI) Observations and Modelling Department R&D Observations Division) f) the animal population density g) the daily minimum temperatures from the 81 weather stations of the Italian Air Force Meteorological Service for Step 2: A slope grid indicating the maximum rate of change between each cell and its neighbours was derived from the DEM; four grids of mean minimum temperature (one for each season) were created through Ordinary Kriging Interpolation in ArcGis 8.2 using the data from the 81 weather stations; four mean NDVI grids (one for each season) were calculated also. Through the spatial extension in ArcGis 8.2 each grid was used to extract values for each site where midge catches were made between 2000 and From the values obtained, frequency distributions were created and the percentage of positive and negative catches calculated (an example is provided in Table I). Step 3: Each of the suitability factors was reclassified on a common scale from 0 to 10, creating new integer grids for each variable. The risk reclassification was made using the percentage of positive catches/total catches falling in each class (Table I). Step 4: For each season, the eight grids were combined, assigning the same weight to each layer. The final grids had a cell size of decimal degrees. The following formula, generating a ranking for each cell for C. imicola suitability, was adopted: 1/8*elevation + 1/8*aridity + 1/8*landuse + 1/8* density + 1/8*ndvi + 1/8*slope + 1/8* temperature + 1/8*lithologic environment. Table I Example of the aridity index reclassification in Microsoft Excel: percentage of sites positive for Culicoides imicola according to various aridity classes 1 A B C D E F Aridity Positive Negative Total Positive Risk class class sites sites sites sites (%) 2 <= % 4* % % % % % % 0 9 > % 0 * Ei * 10 Fi = round (,0) Max ( Ei ) Results Examples of input variables included in the model are shown in Figures 1, 2, 3, 4 and m 0 m Figure 1 Elevation above sea level Figure 2 Aridity index 312 Veterinaria Italiana, 40 (3), 2004

267 0.515 The characteristics of territories with the highest percentage of sites positive for C. imicola are listed in Table II. The final grids with geographic and climatic characteristics suitable for C. imicola are illustrated in Figures 6, 7, 8 and 9, where the risk classes (ranked from 7 to 10) are depicted in shades of red. Figure 10 gives the distribution of C. imicola presence from 2000 to 2003 and the distribution of BT seroconversions in cattle in 2002 and Figure 3 Normalised difference vegetation index in winter 10 C 6 C Figure 4 Mean minimum temperature in winter Table II Highest ranking values of the eight variables associated with the presence of Culicoides imicola Variable Elevation Risk values 0-50 m above sea level Slope 0-3 degrees Aridity index Landuse Permanently irrigated land Animal density 250 animals/km 2 Type of soil Intrusive rocks Temperature (winter) 6 C-7 C Temperature (spring) 10 C-11 C Temperature (summer) 19 C-20 C Temperature (autumn) 14 C-16 C NDVI (winter) NDVI (spring) NDVI (summer) NDVI (autumn) NDVI normalised difference vegetation index Intrusive rocks Effusive rocks Turbidic and alternated deposits Alluvial deposits Carbonatic rocks Sand and gravel deposits, sandstone Water bodies Clay deposits and claystone Pyroclastic rocks Metamorphic rocks < Cell with negative C. imicola catches Cell with positive C. imicola catches Figure 6 Risk map for Culicoides imicola in winter Figure 5 Lithologic environment Veterinaria Italiana, 40 (3),

268 Discussion < Cell with negative C. imicola catches Cell with positive C. imicola catches Figure 7 Risk map for Culicoides imicola in spring < Cell with negative C. imicola catches Cell with positive C. imicola catches < Cell with negative C. imicola catches Cell with positive C. imicola catches Previous models have indicated that temperature, elevation, humidity, rainfall and NDVI are the most important variables that predict the presence of C. imicola across the Mediterranean Basin (1, 2, 4). The success of these predictions has, however, been partial which indicates that the full range of climatic and geographic variables that determine the presence of this vector have yet to be identified. In an effort to refine these earlier predictions, new variables were taken into consideration, these including the aridity index, lithologic environment and land use. Other adjustments made were to consider the seasons separately since climatic variables and NDVI have a clearly seasonal pattern. In the earlier models, and specifically in those for Italy, the presence of the vector, in a number of areas, was either overpredicted or under-predicted. For example, the model of Conte et al. (4), based only on presence/absence data (i.e. not refined to using actual abundances which can fluctuate greatly), overpredicted C. imicola along the entire southern coastline of the island of Sicily, and under-predicted it for the mainland region of Basilicata. For both of these areas, the model presented here, and more specifically that for autumn (the period of greatest disease prevalence), shows a marked improvement over our earlier results (4). However, a persistent exception was that the model still incorrectly predicted the vector to be widely present in the Puglia region. However, this is not the case as demonstrated by hundreds of Culicoides light-trap collections made over more than one season in the field (Fig. 10). Figure 8 Risk map for Culicoides imicola in summer < Cell with negative C. imicola catches Cell with positive C. imicola catches Figure 9 Risk map for Culicoides imicola in autumn Cell with seroconversion Cell with negative C. imicola catches Cell with positive C. imicola catches Figure 10 Distribution of Culicoides imicola presence from 2000 to 2003 and distribution of bluetongue seroconversions in cattle in 2002 to Veterinaria Italiana, 40 (3), 2004

269 In the present model, the disagreement seems to be due to the similarity between Sardinia and the Puglia region for almost all the variables except for lithologic environment. As Figure 11 clearly shows, calcareous soils dominate the landscape in the Puglia region. This principal edaphic difference would suggest that this is a factor that significantly influences the ability of C. imicola to persist locally. However, this remains a best-fit hypothesis that requires confirmation through additional testing as the heel of Italy may also experience higher average winds, which could also decrease flight and feeding activity in C. imicola. At present, the SPM is an additive model that assigns equal weight to each variable. This would need to be improved as it would appear that a specific variable (such as calcareous soil) has a causal relationship with the total absence of C. imicola locally. From the above, it is obvious that gaps exist in the current knowledge and understanding of climatic and geographic factors (and their interactions) that affect the survival and spread of C. imicola. Much remains to be done before C. imicola can be predicted with greater accuracy around the Mediterranean Basin. Calcareous soil Other type of soil Figure 11 Calcareous areas of Italy Palaearctic Culicoides were incriminated in the transmission of the disease. By definition, each Culicoides species in nature occupies a different niche, and consequently future modelling of these species as novel vectors of BT will necessitate the inclusion of variables different from those used for C. imicola in the present study. This in turn demands a fairly detailed understanding of the life-cycles of the new vectors, which, for species in the Obsoletus and Pulicaris Complexes, may be even less well understood than for C. imicola. References 1. Baylis M., Mellor P.S., Wittmann E.J. & Rogers D.J. (2001). Prediction of areas around the Mediterranean at risk for bluetongue by modelling the distribution of its vector using satellite imaging. Vet. Rec., 149, Calistri P., Goffredo M., Caporale V. & Meiswinkel R. (2003). The distribution of Culicoides imicola in Italy: application and evaluation of current Mediterranean models based on climate. J. Vet. Med. B, 50 (3), Caracappa S., Torina A., Guercio A., Vitale F., Calabro A., Purpari G., Ferrantelli V., Vitale M. & Mellor P.S. (2003). Identification of a novel bluetongue virus vector species of Culicoides in Sicily. Vet. Rec., 153 (3), Conte A., Giovannini A., Savini L., Goffredo M., Calistri P. & Meiswinkel R. (2003). The effect of climate on the presence of Culicoides imicola in Italy. J. Vet. Med. B, 50 (3), Environmental Systems Research Institute, Inc. (ESRI) (2003). ArcGis TM spatial analyst. ESRI Redlands, California (esri.com/software/arcgis/index.html/ accessed on 15 August 2004). 6. Savini G., Goffredo M., Monaco, F., de Santis P. & Meiswinkel R. (2003). Transmission of bluetongue virus in Italy. Vet. Rec., 152 (4), Wayne C. (2003). Suitability analysis with raster data. ArcUser (April-June), ( news/arcuser/0403/files/landfill_1.pdf accessed on 15 August 2004). Another serious complication revealed was that during the current incursions of BT into Italy seroconversions in sentinel cattle and outbreaks amongst sheep also occurred in areas in which C. imicola was not to be found or, if found, its abundances were very low. In some of these outbreaks, at least two additional species of Veterinaria Italiana, 40 (3),

270 Vet. Ital., 40 (3), Factors affecting the spread of Culicoides brevitarsis at the southern limit of distribution in eastern Australia A.L. Bishop, L.J. Spohr & I.M. Barchia NSW Agriculture, Locked Bag 26, Gosford, NSW 2250, Australia Summary Culicoides brevitarsis Kieffer is the main vector of bluetongue and Akabane viruses in Australia. Its threat to animal health and livestock exports requires that areas free of the vector and viruses be defined clearly. In New South Wales, survival of the vector over winter is limited to the northern coastal plains. C. brevitarsis therefore has to reinfest areas outside the endemic area each year. Models have been developed to predict the extent and nature of its movements. It can move at different rates and this is partly due to significant delays of movement due to the barrier formed by the altitude of the Great Dividing Range. C. brevitarsis subsequently retains a coastal distribution in most years. At the end of the season, the times when activity would effectively cease can be estimated from temperature data. These data allow evidenced-based conclusions on zonal and seasonal freedom to be made in combination with light trap monitoring. Keywords Altitude Australia Bluetongue Culicoides brevitarsis Dispersal Models Vector freedom. Introduction There are several species of biting midge from the genus Culicoides (Diptera: Ceratopogonidae) that transmit viruses to animals in Australia (14). Culicoides brevitarsis Kieffer is the main species responsible for the transmission of bluetongue and Akabane viruses to livestock (10). In New South Wales (NSW), C. brevitarsis and the viruses move from the endemic mid-northern/northern coastal plain in springsummer (2, 3). Establishment and survival outside the endemic area then depends primarily on temperature (4), moisture and habitat/host availability. C. brevitarsis normally retains a coastal distribution and virus transmission occurs within its dispersive limits. The patterns of occurrence of C. brevitarsis and virus transmissions in NSW have been consistent with their failure to survive winter away from the endemic area. The vector then makes passive and gradual seasonal movements (that probably occur regularly) down the southern coastal plains and up coastal valleys towards the west (2, 3). This is consistent with epidemiological evidence which suggests that the long-distance spread of many viruses is related to the wind-borne dispersal of midge vectors (6, 13). Threats to animal health and livestock exports in Australia require that vector- and virus-free areas be clearly defined. In NSW, these can be determined from knowledge of vector survival and dispersal, supported by seasonal monitoring with light traps. The key elements of this determination require predictions of times when vector activity ceases before winter and estimates of the potential for the vector to survive over winter (1) as well as predictions of probable reinfestation in areas where it cannot survive winter. Our aims were as follows: 1) to model the observed dispersal of C. brevitarsis to the south and west of its endemic area in NSW by considering that the Hunter Valley was the most likely route of dispersal to the west 2) to model the dispersal of C. brevitarsis up three major coastal valleys leading from the endemic mid-north coast to the western slopes and plains of NSW, taking into account the effect of a barrier formed by the altitude and escarpment of the Great Dividing Range. 316 Veterinaria Italiana, 40 (3), 2004

271 Materials and methods The two aims were investigated in separate studies. The first has been described previously (5). This study provided predictive models for C. brevitarsis movement down the coastal plain and up the Hunter Valley towards the western slopes and plains where there are high densities of susceptible cattle and sheep. C. brevitarsis was sampled at selected sites throughout NSW from 1990 to Data were compiled based on the time that C. brevitarsis first occurred at a site. The objective of the analysis was to model the dispersal data as a linear function of site distance from the endemic area and of weather variables (temperature, rainfall, wind frequency from different directions and wind speed). The area of dispersal was divided into the three regions based on observed biological and geographical constraints. Four key conclusions were made, as follows: the dispersal of C. brevitarsis can be explained by distance from the endemic area in NSW. the movement of C. brevitarsis is dependent on temperature and wind speed from northerly and easterly directions. the models predict times of first occurrence within regions down the southern coastal plain or up the Hunter Valley towards (but rarely reaching) the western slopes and tablelands C. brevitarsis moves at different rates in different areas. These were mainly expressed as significantly slower movement up the Hunter Valley than down the coastal plain and suggested that the speed of dispersal could be influenced by geographical features, such as urban areas, increasing altitude and the escarpment of the Great Dividing Range acting as physical barriers. There are several valleys that originate in the eastern escarpment of the Great Dividing Range and are adjacent to the endemic coastal area (Fig. 1). The valleys are approximately oriented east to west and all movements of C. brevitarsis were assumed to be in a westerly direction with the prevailing winds. A single light trap was placed at each of fifteen sites in and beyond three of these valleys for eight seasons from 1995 to 2003 (Fig. 1). Seven additional sites were sampled in the Hunter Valley from 1993 to 2003 to test if the slower movement previously recorded was related to altitude (5). The altitudes and distances from the coast were recorded for each site. C. brevitarsis activity and numbers were recorded from catches in standardised light traps (7). Each trap had a 3.2-V globe and a small downwardlydirected fan driven by three D-cell batteries. A photoelectric cell automatically triggered operation at sunset. The traps were suspended about 2 m above the ground in areas with cattle. Collections were made into plastic bottles containing 70% alcohol. Culicoides spp. were separated from other insects under a binocular microscope, C. brevitarsis was identified by its wing pattern and numbers were recorded. Figure 1 Locations of light traps for Culicoides brevitarsis in four coastal valleys of New South Wales, Australia from 1995 to 2003 The sites were sampled over a 29-week period when C. brevitarsis was active (October to May in each year). Catches were made over two nights, three times per month with the week of the full moon excluded. Two data sets were compiled. The first was based on the time that C. brevitarsis first occurred at a site, regardless of any future event. The second aim of this study was based on the times when C. brevitarsis first occurred and when it also occurred for a second time at a site in the same season. A second occurrence was defined as happening when C. brevitarsis was found at a site at least three zero sampling weeks after the first record, a period long enough to assume that C. brevitarsis had not previously been established. A second occurrence was treated as a new event and the result of repeated movements. First and second occurrence models were developed from their respective data sets. The three valleys and the Hunter Valley data sets were analysed separately. The methodology used the same principles as those employed by Bishop et al. (5) except that it was based Veterinaria Italiana, 40 (3),

272 on weeks instead of days. S-Plus (9) was used to fit a generalised failure time model to the data (15, 16). The relationship between time, distance and altitude was written in the z variable, as follows: z = {t [b0 + b1 D+b2 A]}/ σ where the observed value t = time and follows the normal distribution, A = altitude (m) and D = distance (km) from the coast for the three coastal valleys or D = loge distance (km) from the first site in the Hunter Valley. Results The hazard function expressing the first time that C. brevitarsis was found at sites is given as follows: z = {t [ 3.3 (±1.0) (±0.02) D (±0.0022) A]}/2.2. The hazard function expressing the time taken for C. brevitarsis to be found a second time is given as follows: z = {t [ 3.3 (±1.1) (±0.026) D (±0.0025) A]}/2.2. First and second occurrences were significantly related to the distance from the coast and the altitude of sites (P <0.05). These functions can be used to estimate the time when C. brevitarsis would reach any site in the respective valleys for a first or second time. The time taken for C. brevitarsis to travel a set distance for the first time was extended by 0.48 (±0.22) weeks (approximately 3.4 days) for every 100 m increase in altitude. The time taken for C. brevitarsis to travel a set distance for the second time was extended by 1.14 (±0.24) weeks (approximately 8.0 days) for every 100m increase in altitude. The hazard function representing the time that C. brevitarsis was observed at sites in the Hunter Valley is given as follows: z = {t [ 0.7 (±1.6) (±0.53) D (±0.07) A]}/2.0. Occurrence was also related to distance travelled and the altitude of the sites (P <0.05). Discussion Knowing areas of Australia where C. brevitarsis is permanently or seasonally absent is necessary to define the epidemiology of bluetongue and Akabane viruses and is critical for the establishment of protocols for the livestock export industry. Any incursions into western regions of NSW are a serious threat to the maintenance of a certifiable virus-free area. This paper reports on defining conditions under which C. brevitarsis moves to areas that are normally free of both vector and virus. The usual dispersal of C. brevitarsis down the coastal plain of NSW is a function of distance, temperature and wind conditions but can also be explained solely by the distance to be travelled by the vector (5). Several coastal valleys (Fig. 1) provide direct routes to the west but movements up these valleys are delayed by both distance and altitude of the Great Dividing Range. Delay and declining temperatures restrict movements to the top of the Range. Slower movement up the Hunter Valley (5) was also due to its distance from the endemic area and the effects of increasing altitude. Therefore, distribution normally remains primarily along the coast. Exceptions are infrequent and possibly only occur under chance weather conditions (11, 12). Conclusions There is strong experimental evidence to support the argument for determining areas of NSW that are free of vector and virus. It has been shown that temperatures normally limit the over-wintering survival of C. brevitarsis to the northern coastal plain. Areas where it cannot survive have been calculated from historical temperature data and can also be calculated for any given year (1). C. brevitarsis must therefore reinfest areas outside the endemic area in most years. The dispersal models used enable the prediction of the times of possible occurrences to the south and west of the endemic area. The barrier formed by the Great Dividing Range makes a significant contribution to C. brevitarsis retaining a coastal distribution. There are possibly some exceptions, but any abnormal incursions can be detected from a network of light traps controlled by the National Arbovirus Monitoring Program (8). Should C. brevitarsis reinfest an area, temperature data can be used to estimate the time when its activity would effectively cease at the end of the season (1). In combination, the predictions of first, second and last occurrence of C. brevitarsis, and survival probability away from the endemic area, lead to confident evidence-based conclusions on zonal and seasonal freedom. Acknowledgements The authors thank H. McKenzie for his technical assistance. 318 Veterinaria Italiana, 40 (3), 2004

273 References 1. Bishop A.L., Barchia I.M. & Harris A.M. (1995). Last occurrence and survival during winter of the arbovirus vector Culicoides brevitarsis at the southern limits of its distribution. Aust. Vet. J., 72, Bishop A.L., Kirkland P.D., McKenzie H.J., Spohr L.J., Barchia I.M. & Muller M.J. (1995). Distribution and seasonal movements of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) at the southern limits and its distribution in New South Wales and their correlation with arboviruses affecting livestock. J. Aust. Entomol. Soc., 34, Bishop A.L., Kirkland P.D., McKenzie H.J. & Barchia I.M. (1996). The dispersal of Culicoides brevitarsis in eastern New South Wales and associations with the occurrences of arbovirus infections in cattle. Aust. Vet. J., 73, Bishop A.L., McKenzie H.J., Barchia I.M. & Harris A.M. (1996). Effect of temperature regimes on the development, survival and emergence of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in bovine dung. Aust. J. Entomol., 35, Bishop, A.L., Barchia I.M. & Spohr L.J. (2000). Models for the dispersal in Australia of the arbovirus vector, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae). Prev. Vet. Med., 47, Braverman Y. (1992). The possible introduction to Israel of Culicoides (Diptera: Ceratopogonidae) borne animal diseases by wind. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Dyce A.L., Standfast H.A. & Kay B.H. (1971). Collection and preparation of biting midges (Fam. Ceratopogonidae) and other small Diptera for virus isolation. J. Aust. Entomol. Soc., 11, Kirkland P.D., Ellis T., Melville L.F. & Johnson S. (1995). The national arbovirus monitoring program as a model for studying the epidemiology of bluetongue in China. In Bluetongue disease in South- East Asia and the Pacific (T.D. St George & Peng Kegao, eds). Proc. First South-East Asia and Pacific Regional Bluetongue Symposium, Kunming, August Australian Centre for International Agricultural Research (ACIAR), Canberra, Proceedings No. 66, MathSoft (1998). S-Plus 5 Guide to Statistics. Insightful Corporation MathSoft, Inc. Seattle, pp. 10. Muller M.J., Standfast H.A., St. George T.D. & Cybinski D.H. (1982). Culicoides brevitarsis (Diptera: Ceratopogonidae) as a vector of arboviruses in Australia. In Proc. 3rd Symposium on arbovirus research in Australia (T.D. St George & B.H. Kay, eds). CSIRO, Brisbane, Murray M.D. (1987). Akabane epizootics in New South Wales: evidence for long-distance dispersal of the biting midge Culicoides brevitarsis. Aust. Vet. J., 64, Murray M.D. & Kirkland P.D. (1995). Bluetongue and Douglas virus activity in New South Wales in 1989: further evidence for long-distance dispersal of the biting midge Culicoides brevitarsis. Aust. Vet. J., 72, Sellers R.F., Pedgley D.E. & Tucker M.R. (1977). Possible spread of African horse sickness on the wind. J. Hyg., 79, Standfast H.A., Dyce A.L., St. George T.D., Muller M.J., Doherty R.L., Carley J.G. & Filippich C. (1984). Isolation of arboviruses from insects collected at Beatrice Hill, Northern Territory of Australia, Aust. J. Biol. Sci., 37, Turnbull B.W. (1974). Nonparametric estimation of a survivorship function with doubly censored data. J. Am. Stat. Assoc., 69, Turnbull B.W. (1976). The empirical distribution function with arbitrarily grouped, censored and truncated data. J. Royal Stat. Soc., Ser. B., 38, Veterinaria Italiana, 40 (3),

274 Vet. Ital., 40 (3), Protection of cattle from Culicoides spp. in Australia by shelter and chemical treatments W.M. Doherty (1), A.L. Bishop (2), L.F. Melville (3), S.J. Johnson (1), G.A. Bellis (4) & N.T. Hunt (3) (1) Queensland Department of Primary Industries, Abbott Street, Townsville, QLD 4810, Australia (2) New South Wales Agriculture, Locked Bag 26, Gosford, NSW 2250, Australia (3) Department of Business, Industry and Resource Development, GPO Box 3000, Darwin, NT 0801, Australia (4) Northern Australia Quarantine Strategy, GPO Box 3000, Darwin, NT 0801, Australia Summary Trials were conducted in three regions of Australia to investigate the potential for improvised shelters and chemical treatments to reduce feeding by Culicoides on cattle and thereby minimise the risk of bluetongue transmission during transport of cattle to ports. Various designs and combinations of roofs and walls were placed around penned cattle. Chemical treatments were applied to other penned cattle. Culicoides were collected from the cattle by vacuum samplers or by light traps in the pens. Roofs alone did not consistently reduce the numbers of Culicoides brevitarsis or C. fulvus and increased the numbers of C. actoni collected. Walls alone reduced the numbers of C. wadai but not C. brevitarsis. Roofs and walls in combination reduced the numbers of C. brevitarsis and C. wadai. The chemical treatments Flyaway (a blend of repellents) and fenvalerate reduced the numbers of C. brevitarsis and C. wadai up to 52 h post treatment. Keywords Arbovirus Australia Bluetongue Cattle Culicoides brevitarsis Culicoides wadai Culicoides actoni Culicoides fulvus Repellent Shelter Risk reduction. Introduction Australia exports live cattle to a number of countries. Arbovirus-sensitive markets require the cattle to be free from bluetongue (BT) viruses (BTV). Suitable cattle can be sourced from inland and southern areas. The export of such cattle from inland northern Australia could be enhanced if the animals could be sent from ports in northern Australia. However, to reach these ports, the cattle must be transported through areas in which Culicoides vectors may be active. The risk of BT transmission could be minimised if the cattle could be protected from Culicoides during transport. Several species of Culicoides in northern Australia are capable of transmitting BTV to cattle (9). C. brevitarsis Kieffer (Diptera: Ceratopogonidae) is the most widespread of the vector species in Australia and occurs at least seasonally through northern Western Australia, northern Northern Territory, Queensland and northern and central coastal New South Wales. C. wadai Kitaoka and C. actoni Smith are restricted to the northernmost Northern Territory and coastal Queensland although C. wadai also occurs sporadically in northern coastal New South Wales. C. fulvus Sen and Das Gupta is restricted to northernmost Northern Territory but has previously been recorded in coastal north Queensland. However, a current review is likely to redefine the Queensland population as C. dumdumi Sen and Das Gupta (A.L. Dyce, personal communication). A previous study has suggested that C. brevitarsis is exophagic (8). It is most prevalent in open pasture and the numbers of C. brevitarsis that attack cattle can decrease in wooded areas (1). However, this is not true of all Culicoides species. A variety of other Culicoides species will attack livestock in shelters (4, 6). The habits of the other BT vector species in Australia are unknown and the effects of shelter on any Australian Culicoides have yet to be tested experimentally. Improvised covers on livestock transport compartments may afford cattle some protection from exophagic Culicoides. The cover provided by the slatted construction of cattle road transport vehicles and rail wagons may already 320 Veterinaria Italiana, 40 (4), 2004

275 provide some protection. It may also be possible to enhance this effect by adding extra covers to form roofs and partial walls. A number of chemical treatments have proved to be effective against the vector species in Australia. Bishop et al. (3) collected reduced numbers of C. brevitarsis in light traps covered in mesh treated with Flyaway, fenvalerate, deltamethrin or pyrethroid-t. Melville et al. (7) found fenvalerate, deltamethrin and permethrin reduced the numbers of C. actoni, C. brevitarsis and C. fulvus in cattle 8 h- 60 h post treatment. Doherty et al. (5) found cypermethrin and deltamethrin reduced C. brevitarsis numbers in cattle 8 h-53 h post treatment. However, no tests have been conducted against C. wadai. These trials aimed to investigate the potential of improvised shelters and chemical treatments to protect cattle against the vectors of BTV so that cattle may be transported through areas where vectors may be active. Materials and methods Trials were conducted in 2001 and 2002 in three areas, Beatrice Hill in the Northern Territory (12.39 S, E), Tocal in New South Wales (32.38 S, E) and Mena Creek in Queensland (17.62 S, E). In the Northern Territory, a shelter trial against C. actoni, C. brevitarsis and C. fulvus was conducted (Trial A). In New South Wales, shelter trials against C. brevitarsis were conducted (Trials B and C). In Queensland, shelter (Trial D) and chemical (Trial E) trials against C. brevitarsis and C. wadai were conducted. Trial A Four pens, each m and 1.9 m high, were constructed. Two pens had tarpaulin roofs and two were uncovered. Four uniform steers were held in each pen. Collections were made from the cattle with a vacuum sampler. Covered and uncovered pens were sampled simultaneously. Five collections were made at approximately hourly intervals starting at 17:00 on each of eight evenings in March and a further eight evenings in May/June. Unfed and blood-fed C. actoni, C. brevitarsis and C. fulvus were identified and counted. Data were analysed with a generalised linear model using S-plus. Models were fitted using a Poisson error distribution with a log link. Trial B Six pens each 6 m 6 m and each containing two uniform steers were used. Tarpaulins were placed 3.2 m above three of the yards and three were left uncovered. The covers were relocated between pens at random for each of four serial replicates. C. brevitarsis were collected by vacuuming the backs, sides, neck/heads and rumps of the cattle in each pen for 5 min, 2 h before sunset and 1 h and 2 h after sunset. C. brevitarsis were identified and counted. Data were analysed by restricted maximum likelihood (REML) analysis following a square root transformation to normalise the data. Wald tests were used to test the significance of the fixed effects. Trial C Five treatments were replicated three times on each of four nights. The treatments were as follows: 1) High walls (>2 m high) and high roof. Three similar-sized pens (approximately 3 3 m) were used. The first was a closed wooden stable. The second a steel shed enclosed on three sides with the open side covered to 2.5 m high with a tarpaulin. The third was a cattle yard covered on each side and above with tarpaulins. 2) Low walls (<2 m high) and low roof. These used tarpaulins to cover each side of three pens (each m 1.5 m high). A tarpaulin roof was added to each pen 3) Low walls only (as for treatment 2 without the roof) 4) Low roof only (as for treatment 2 without the walls) 5) Uncovered (larger pens were used in this treatment) (30 m 30 m). Two uniform cattle were placed in each pen. A light trap was placed in each pen at a height of 1.5 m to 2.0 m so that minimal light was visible beyond the pen and only Culicoides within the pen would be trapped. The light traps were operated overnight and C. brevitarsis collected were identified and counted. Statistical analysis used a mixed model that allowed for random effects as the errors were not correlated between times. Trial D Two steers were placed into each of three 1.1 m 3.0 m pens. Three treatments were included, namely: slatted walls, slatted walls and a tarpaulin roof and an unmodified pen as a control. The walls copied those found on cattle transport compartments. Plywood was fixed to panels of steel tubing in horizontal strips to produce a slatted wall with 35% of its area uncovered. The walls were erected around two pens creating outer pens 4.2 m 2.1 m and 1.9 m high. One was covered with a tarpaulin. The experiment was replicated four times Veterinaria Italiana, 40 (4),

276 on four consecutive nights. Each treatment was allocated to a different pen each night in a randomised pattern. Collections were made by vacuuming the backs, sides, neck/heads and rumps of the cattle in each pen for 5 min. Collections occurred at intervals of at least 30 min to give 9-13 collections for each night. The numbers of blood-fed and unfed female C. brevitarsis and C. wadai were identified and counted. Mean nightly fed and total numbers were analysed using analysis of variance and least significant difference (LSD) on the natural log transformed means (mean + 1) using Genstat 5 (P<0.05). Trial E The cattle and three unmodified pens from Trial D were used. Three treatments were included: fenvalerate, Flyaway and untreated. Fenvalerate was applied as a spray of 200 ml per steer of 1% active ingredient aqueous solution to the back, rump, flanks and head. Flyaway, (12 g/l permethrin, 50 g/l diethyltoluamide, 25 g/l n-octyl bicycloheptene dicarboximide, 25 g/l piperonyl butoxide, 20 g/l dibutyl phthalate and 10 g/l lavender oil) was applied as an undiluted spray to a similar area at 20 ml-25 ml per steer. Treatments were applied at 16:30. C. brevitarsis and C. wadai were collected as in Trial D. Collections were made at 25 min intervals from 17:30-20:25 for each of the three nights following treatment giving eight collections each night covering the periods 1 h-4 h, 25 h-28 h and 49 h- 52 h post treatment. Four serial replicates were performed in successive weeks. Analysis was similar to Trial D. Results In Trial A, the tarpaulin roofs did not reduce the numbers of fed or unfed C. brevitarsis and C. fulvus collected from the cattle. Significantly more fed and unfed C. actoni were collected from the cattle in the covered pens (Table I). In Trial B, the high roof reduced the numbers of C. brevitarsis collected from the cattle (Table II). In Trial C, the lowest number of C. brevitarsis was collected in the pens with walls and roof, with fewer in the high walls and roof pens than the low walls and roof pens (Table II). Fewer C. brevitarsis were collected in pens with walls only than in pens with roof only. The pens with walls or roof only collected more C. brevitarsis than the uncovered pens but comparisons with the uncovered pens may be complicated by the different pen sizes. The light traps may have been less effective in the larger uncovered pens as they may not have been as close to the cattle at all times. Table I Trial A (Northern Territory) Mean numbers of Culicoides per vacuum sample from penned cattle Treatment Pen with roof Uncovered pen C. actoni C. brevitarsis C. fulvus Fed Unfed Fed Unfed Fed Unfed 61.4 a a 1.4 a 6.8 a 0.9 a 14.7 a 32.8 b b 1.2 a 6.4 a 1.0 a 15.5 a Within each column, means with different superscripts are significantly different (P<0.05) In Trial D, the pens with slatted walls and tarpaulin roof reduced both blood-fed and total numbers of C. brevitarsis and C. wadai collected (Table III). The pens with slatted walls only reduced the numbers of C. wadai, both blood-fed and total, but not the numbers of C. brevitarsis. The pens with walls and roof reduced total C. wadai numbers more than walls alone. Table II Trials B and C (New South Wales) Back-transformed mean numbers of Culicoides brevitarsis per collection from penned cattle Trial Sampling method Treatment C. brevitarsis B Vacuum Pen with high (3 m) 3.1 a sampler tarpaulin roof Uncovered pen 14.9 b C Light traps 1) Small pen with high 4.6 c roof and walls (>2 m) 2) Small pen with low 13.7 d roof and walls (<2 m) 3) Small pen with low b walls only (<2 m) 4) Small pen with low a roof only (<2 m) 5) Large uncovered pen 37.4 e Within each trial only, means with different superscripts are significantly different (P<0.05) In Trial E, both Flyaway and fenvalerate reduced both blood-fed and total numbers of C. brevitarsis and C. wadai collected (Table III). Flyaway and fenvalerate were equally effective. The efficacy of both chemicals for both species did not vary significantly between the assessment periods, 1 h-4 h, 25 h-28 h and 49 h-52 h post treatment. Discussion The effect of roofs on C. brevitarsis varied between trials. There are many factors, which could interact with the response of C. brevitarsis and other species to shelters to produce this variability. For example, C. brevitarsis is crepuscular but its activity is also affected by temperature (2). If its response to 322 Veterinaria Italiana, 40 (4), 2004

277 shelters differed before and after dusk, different areas and times of year could give different results. Table III Trials D and E (Queensland) Back-transformed mean numbers of Culicoides per vacuum sample from penned cattle Trial Treatment C. brevitarsis C. wadai Fed Total Fed Total D Pen with walls 1.6 b 13.2 b 0.1 a 1.5 b Pen with walls 0.4 a 2.6 a 0.04 a 0.27 a and roof Uncovered pen 2.3 b 21.2 b 0.5 b 3.4 c E Fenvalerate 0.74 a a 0.06 a 1.12 a Flyaway 1.71 a a 0.16 a 1.87 a Untreated b b 1.65 b b Within each column of each trial only, means with different superscripts are significantly different (P<0.05) Culicoides actoni appears to be endophagic and so even more complete shelters may not afford protection against this species. C. fulvus was undeterred by roofs but may, like C. brevitarsis, be deterred by a more complete shelter. C. wadai can be deterred and it is likely that any shelter that is effective for C. brevitarsis will also be effective for C. wadai. Improvised shelters consisting of only roofs or walls currently appear unlikely to give cattle reliable protection against all the Australian BT vectors. An unmodified cattle transport compartment typically consists of walls only and could not be relied on to reduce the risk of BT transmission in any area where C. brevitarsis is present. C. brevitarsis occurs wherever C wadai occurs so the effectiveness of walls against C. wadai is unfortunately of no practical value. Shelters with walls and roofs appear to offer useful protection for cattle against C. brevitarsis and C. wadai. The addition of tarpaulin roofs to transport compartments could be a useful risk reduction strategy if it does not compromise the welfare of the cattle. Chemical treatments offer reliable protection for cattle. With the addition of this trial against C. wadai, to those mentioned above, fenvalerate has proved to be effective against all BT vectors in Australia. Flyaway gives useful protection against C. brevitarsis and C. wadai but has not yet been tested against C. actoni and C. fulvus although another permethrin product was effective against these latter species (7). Although neither shelter nor chemical treatments alone can currently entirely eliminate the risk of BT transmission to cattle during transport, they could be a valuable addition to other risk reduction strategies, such as uninterrupted travel to enable safe transport of cattle through areas of BT risk. Acknowledgements Biosecurity Australia provided funding for much of this work. Harry McKenzie, Lorraine Spohr, Idris Barchia, Diana Pinch, Bill Albrecht and Angela Reid provided invaluable assistance. References 1. Bishop A.L., McKenzie H.J., Spohr L.J. & Barchia I.M. (1994) Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in different farm habitats. Aust. J. Zool., 42, Bishop A.L., McKenzie H.J., Spohr L.J. & Barchia I.M. (1995) Daily activity of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) in the Hunter Valley, NSW. Gen. Appl. Entomol., 26, Bishop A.L., McKenzie H.J., Spohr L.J. & Barchia I.M. (2001) In vitro testing of chemicals for repellency against Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae). Gen. Appl. Entomol., 30, Cheah T.S. & Rajamanickam C. (1991) Occurrence of Culicoides spp. (Diptera: Ceratopogonidae) in sheep sheds and their relevance to bluetongue in peninsular Malaysia. Trop. Anim. Hlth Prod., 23, Doherty W.M., Johnson, S.J. & Reid A.E. (2001) Suppression of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) on cattle in Queensland with deltamethrin and cypermethrin. Gen. Appl. Entomol., 30, Meiswinkel R., Baylis, M. & Labuschagne K. (2000) Stabling and the protection of horses from Culicoides bolitinos (Diptera: Ceratopogonidae), a recently identified vector of African horse sickness. Bull. Entomol. Res., 90, Melville L.F., Hunt N.T., Bellis G.A. & Pinch D. (2001) Evaluation of chemical treatments to prevent Culicoides spp. feeding on cattle in the Northern Territory. Gen. Appl. Entomol., 30, Standfast H.A. & Dyce A.L. (1972). Arthropods biting cattle during an epizootic of ephemeral fever in Aust. Vet. J., 48, Standfast H.A., Dyce A.L. & Muller M.J. (1985). Vectors of bluetongue virus in Australia. In Bluetongue and related orbiviruses (T.L. Barber, M.M. Jochim & B.I. Osburn, eds). Proc. First International Symposium, Monterey, California, January A.R. Liss, Inc., New York, Veterinaria Italiana, 40 (4),

278 Vet. Ital., 40 (3), Temporal activity of biting midges (Diptera: Ceratopogonidae) on cattle near Darwin, Northern Territory, Australia G.A. Bellis (1), L.F. Melville (2), N.T. Hunt (2) & M.N. Hearnden (2) (1) Australian Quarantine and Inspection Service Northern Australia Quarantine Strategy, GPO Box 3000 Darwin, NT 0801 Australia (2) Northern Territory Department of Business, Industry and Resource Development, GPO Box 3000 Darwin, NT 0801 Australia Summary The activity of nine species of biting midges aspirated from cattle was recorded in the late afternoon, evening and early morning at a site near Darwin, Northern Territory, between March and June in 1999 and There were no significant differences between the temporal activity patterns for nulliparous and parous females of any species. Nulliparous females dominated collections of all species except Culicoides marksi. C. actoni and Forcipomyia (Lasiohelea) sp., were mostly active during daylight hours while C. peregrinus, C. bundyensis and C. brevipalpis, were nocturnal. Differences in the peak activity of C. brevitarsis were noted between years and occurred slightly earlier than that observed at other sites. C. fulvus, C. marksi and C. oxystoma were generally crepuscular but differed in the length and peak period of activity. C. actoni was four times more active in the evening than in the morning while C. marksi and C. peregrinus, were respectively 2.6 and 3.4 times more active in the morning than in the evening. Numbers of the other six species were not significantly different in the evening and morning. All nine species were collected at least once from cattle shortly after dawn. Keywords Australia Biting midges Bluetongue Cattle Culicoides Lasiohelea Onchocerca Temporal activity Vector. Introduction Biting midges (Diptera: Ceratopogonidae) are recognised pests of bovids in northern Australia and elsewhere, chiefly through their ability to transmit diseases and parasites (1, 2, 11, 13). Species often differ in their temporal activity on hosts and this information can be important for those interested in collecting insects for laboratory studies, sampling the biting fauna or protecting valuable stock from attack. When collecting biting midges from cattle in Malaysia, Buckley (2) concentrated on the period between 7.30 am and 9.30 am and in the two hours prior to sunset as he had observed maximum activity at these times. He noted that midges were extremely scarce during the heat of the day. His nocturnal collections were conducted between 11 pm and 2 am. However, midges collected during these hours were thought to be attracted to the light used to facilitate collecting rather than being attracted to the animals to feed. Of the species he collected, only Culicoides actoni Smith (listed as C. pungens (14), C. oxystoma Kieffer and C. peregrinus Kieffer are also present in the coastal Northern Territory of Australia. In south-east Queensland, C. brevitarsis Kieffer exhibited very low activity during daylight hours which sharply increased at sunset, peaked 30 min thereafter and gradually reduced to zero over the next 6 h, followed by a minor period of activity at sunrise (3). Standfast and Dyce (12) also noted a peak in activity of this species at dusk. Beveridge et al. (1) collected C. marksi Lee and Reye, C. brevitarsis, C. bundyensis Lee and Reye C. actoni and two species of Forcipomyia (Lasiohelea) from cattle before sunset and after sunrise but made no attempt to collect at night. 324 Veterinaria Italiana, 40 (3), 2004

279 Pathogens are usually transmitted biologically by biting midges. Consequently, the epidemiologically important part of the population is the older females that have had the opportunity to take a blood meal and become infective for the pathogen. Protecting valuable stock from infection therefore relies on protecting against attack by old females. If protection is to be achieved by reducing exposure to attack during peak midge activity, the activity of old female midges needs clarification. Few studies of the relative activity of nulliparous and parous female midges have so far been published although the proportion of parous female C. sonorensis Wirth and Jones (as C. variipennis sonorensis) collected in a CO 2 - baited trap was reported to be higher in the middle of the night, prompting the suggestion that crepuscular oviposition patterns may explain the difference in temporal activity between parous and nulliparous midges (10). The authors report on the temporal activity of nulliparous and parous biting midges collected from cattle on the sub-coastal plains of the Northern Territory. and 75th quartiles used to describe dispersal around the median (Figs 1, 2 and 3). Parous females Nulliparous females Total females Figure 1 Temporal activity with respect to sunrise of nine bitingmidge species on cattle near Darwin, Northern Territory, March, April, May and June 1999 Lines delineate the 25th to 75th quartiles around the median Materials and methods Insect collections The method used to collect, preserve and sort midges on cattle are fully described elsewhere (8, 9). Briefly, midges were collected from a group of ten cattle which were not treated with insecticides and which were housed in open pens while collections were being made. Midges were aspirated using a commercially available garden leaf blower (Makita RBL250) modified to allow insect collections into a gauze bag. Midges were aspirated from the back-line and flanks for one five-minute period per half hour on each day collections were made. Collections made between 30 March and 3 June 1999 began as early as 3 h prior to sunset and continued as late as 3.75 h after sunset and recommenced as early as 3 h prior to sunrise continuing as late as 2.5 h after sunrise. Collections made between 12 March and 14 June 2001 began as early as 1.75 h prior to sunset and continued until as late as 2.75 h after sunset. Collections were sorted into species and parity was assessed based on abdominal pigmentation (5). Statistical analyses The time between the collection of a sample and the closest dawn or dusk time was calculated for each sample. The distribution of activity times for most species were highly skewed so the measure of peak activity was characterised by the median, with 25th Parous females Nulliparous females Total females Figure 2 Temporal activity with respect to sunset of nine biting midge species on cattle near Darwin, Northern Territory, March, April, May and June 1999 Lines delineate the 25th to 75th quartiles around the median To compare numbers of biting nulliparous and parous females of each species trapped at dawn (early morning) and dusk (afternoon and evening), the sample data were analysed using a generalised linear model (6). The model used tested the main effects of time of day (dawn or dusk), and parity type (nulliparous and parous females) and the interaction of time and type. Models were fitted using a Poisson error distribution with a log link. Veterinaria Italiana, 40 (3),

280 All statistical tests were calculated at the 5% significance level (α=0.05) using S-Plus statistical software (7). generally active earlier than C. fulvus. C. marksi had a much narrower period of activity than either C. oxystoma or C. fulvus. Despite the significant nocturnal activity of C. oxystoma in the evening, peak morning activity of this species was not until well after dawn. Evening activity of C. brevitarsis peaked before dusk and was significantly earlier in 2001 than in 1999 (median test, Chi-square= , p<0.0001). Figure 3 Temporal activity with respect to sunset of nine biting midge species on cattle near Darwin, Northern Territory, March, April, May and June 2001 Lines delineate the 25th to 75th quartiles around the median Results Parous females Nulliparous females Total females A total of and biting midges comprising 18 species were collected from cattle in 1999 and 2001, respectively. The most dominant species that attacked cattle were C. actoni and C. peregrinus, comprising 27% and 60%, respectively of the total in 1999 and 48% and 25%, respectively, of the total in Sufficient data to measure temporal activity was obtained for species and these are presented in Figures 1, 2 and 3. The parity of C. brevipalpis Delfinado and an as yet unidentified species of Forcipomyia (Lasiohelea) could not be reliably assessed using abdominal pigmentation so data on these species is presented as total females and were excluded from results comparing activity of parous and nulliparous females. There were notable differences in the periodicity of these nine species on cattle. Three of these species, C. brevipalpis, C. bundyensis and C. peregrinus, were nocturnal although they were also collected after sunrise. C. actoni and F. (Lasiohelea) sp. were mostly diurnal with significant numbers collected after sunrise and both exhibited a sharp drop in activity following sunset. C. fulvus Sen and Das Gupta was crepuscular with significant activity extending for at least 1 h after dusk and 2 h prior to dawn. C. marksi and C. oxystoma were also crepuscular but were Numbers of C. actoni were about four times higher at dusk than those at dawn. Numbers of C. marksi and C. peregrinus were respectively about 2.6 and 3.4 times higher in the morning than in the evening. No significant differences were observed between morning and evening totals for the other species (Table I). Nulliparous females dominated collections of all species excepting C. marksi (Table II). There were no significant differences in the temporal pattern of activity of nulliparous and parous females of any of the species studied (Table III). Table I Mean (standard error) and associated p-values of female biting midges aspirated from cattle during morning or afternoon and evening periods at Beatrice Hill, Northern territory, between March and June 1999 Species Morning Afternoon and evening p-value Culicoides actoni 18.3 (9.29) 76.3 (15.46) C. brevitarsis 0.8 (0.17) 1.8 (0.43) C. bundyensis 6.5 (1.75) 6.4 (1.49) C. fulvus 4.8 (1.51) 2.9 (0.60) C. marksi 1.6 (0.47) 0.6 (0.12) C. oxystoma 23.7 (15.48) 16.1 (3.44) C. peregrinus 291 (107.9) 265 (70.2) Table II Mean (standard error) and associated p-values of nulliparous and parous female biting midges aspirated from cattle at Beatrice Hill Northern territory, between March and June 1999 Species Nulliparous females Parous females p-value Culicoides actoni 96.6 (21.89) 26.0 (7.76) C. brevitarsis 2.4 (0.61) 0.7 (0.19) C. bundyensis 10.1 (2.28) 2.7 (0.19) C. fulvus 5.5 (1.12) 1.2 (0.24) < C. marksi 0.9 (0.19) 0.9 (0.23) C. oxystoma 28.4 (9.09) 7.7 (1.92) C. peregrinus 265 (70.2) 12.2 (2,86) < Veterinaria Italiana, 40 (3), 2004

281 Table III Interaction means (standard errors) and associated p-values for collection period (morning or afternoon and evening) of nulliparous and parous biting midge species aspirated from cattle at Beatrice Hill, Northern Territory, between March and June 1999 Species Nulliparous females Afternoon and Morning evening Parous females Afternoon and Morning evening p-value Culicoides actoni 34.2 (18.11) (28.33) 2.5 (1.09) 34.1 (10.24) C. brevitarsis 1 (0.27) 2.9 (0.8) 0.6 (0.21) 0.7 (0.25) C. bundyensis 9 (3.24) 10.5 (2.86) 4 (1.17) 2.3 (0.51) C. fulvus 7.9 (2.85) 4.7 (1.14) 1.8 (0.66) 1 (0.24) C. marksi 1.2 (0.47) 0.8 (0.2) 2.1 (0.8) 0.5 (0.12) C. oxystoma 45.2 (30.61) 22.6 (6.35) 2.3 (1.26) 9.6 (2.51) C. peregrinus 550 (202.93) (60.1) 32.5 (9.4) 5.2 (1.28) Discussion Temporal activity is clearly a critical factor when assessing the biting midge fauna attacking hosts. For example, sampling only before sunset would overestimate the importance of C. actoni, C. marksi, F. (Lasiohelea) sp. and possibly also C. brevitarsis while underestimating C. bundyensis, C. brevipalpis and C. peregrinus and vice versa. The low activity of C. actoni, F. (Lasiohelea) sp. and to some extent C. brevitarsis after sunset, may also affect the ability of some sampling devices, for example light traps to representatively sample populations of these species. All of these species were still active after dusk and consequently still detectable in light traps. Indeed, Melville et al. (9) found that C. brevitarsis dominated light-trap collections made in the same time and place that the present study was conducted. Temporal activity of C. actoni, C. oxystoma and C. peregrinus in Darwin was generally similar to that observed for these species in Malaysia (2). Debenham (4) noted that although few studies into the temporal activity of Australian species of F. (Lasiohelea) had been undertaken, the collection times of those specimens she studied indicated a diurnal, sometimes crepuscular, activity pattern. The difference in the activity pattern of C. brevitarsis between 1999 and 2001 was unexpected. The predusk peak observed in both years also differs from the post-dusk peak observed in this species in southeast Queensland (3). Early morning activity was more important relative to dusk activity in Darwin, which also contrasts with the predominantly evening activity observed in south-east Queensland. Environmental factors may contribute to the differences between the Darwin and south-east Queensland activity patterns although these are less likely to explain the difference between the 1999 and 2001 results, as these were both obtained at the same site during the same time of the year. The trigger for these changes in the activity pattern remains unclear. Protection of stock from attack may be achieved by reducing exposure to potentially infective vectors during periods of peak activity. As no significant differences were observed between nulliparous and parous females of the seven species studied, this assessment is somewhat simplified. For those diseases with only one known vector, for example Onchocerca sweetae (11), protection may only be required during peak activity of the vector, C. bundyensis, i.e. after sunset and before dawn. For those diseases transmitted by a suite of vectors, for example bluetongue virus (13), the period over which all of the vectors are active in the area of concern needs to be taken into account. Acknowledgements Eric Cox, Bradley Hunt, Lindsay Melville and Jason Stevens assisted with the insect collections at Beatrice Hill Farm. Bradley Hunt and Lindsay Melville pre-sorted the insect collections. Lindsay Melville entered all the data and Alan Bishop provided advice on the design and analysis of the experiment. This work was partly funded by Biosecurity Australia. References 1. Beveridge I., Kummerow E.L., Wilkinson P. & Copeman D.B. (1981). An investigation of biting midges in relation to their potential as vectors of bovine onchocerciasis in north Queensland. J. Aust. Entomol. Soc., 20, Veterinaria Italiana, 40 (3),

282 2. Buckley J.J.C (1938). On Culicoides as a vector of Onchocerca gibsoni (Cleland and Johnson 1910). J. Helminthol., 16 (3), Campbell M.M. & Kettle D.S. (1979). Abundance and temporal and spatial distribution of Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae) on cattle in south-east Queensland. Aust. J. Zool., 27, Debenham M.L. (1983). Australasian species of the blood-feeding Forcipomyia subgenera Lasiohelea and Dacnoforcipomyia (Diptera: Ceratopogonidae). Aust. J. Zool., Supplementary Series No. 95, Dyce A.L. (1969). The recognition of nulliparous and parous Culicoides (Diptera: Ceratopogonidae) without dissection. J. Aust. Entomol. Soc., 8, McCullagh P. & Nelder J.A. (1989). Generalized linear models, Second Ed. Chapman and Hall, London, 511 pp. 7. MathSoft (1998). S-Plus 6 for Windows. Guide to Statistics. Insightful Corporation. MathSoft, Inc., Seattle, 1014 pp. 8. Melville L., Hunt N., Bellis G. & Pinch D. (2001). Evaluation of chemical treatments to prevent Culicoides spp. (Diptera: Ceratopogonidae) feeding on cattle in the Northern Territory. Gen. Appl. Entomol., 30, Melville L., Hunt N., Bellis G., Hearnden M. & Pinch D. (2002). An assessment of Culicoides spp. attacking cattle under cover. Unpublished report to Biosecurity Australia. 10. Mullens B.A. (1995). Flight activity and response to carbon dioxide of Culicoides variipennis sonorensis (Diptera: Ceratopoginidae) in southern California. J. Med. Entomol., 32, Spratt D.M., Dyce A.L. & Standfast H.A. (1978). Onchocerca sweetae (Nematoda: Filarioidae): notes on the intermediate host. J. Helminthol., 52, Standfast H.A. & Dyce A.L. (1972). Potential vectors of arboviruses of cattle and buffalo in Australia. Aust. Vet. J., 48, Standfast H.A., Muller M.J. & Dyce A.L. (1992). An overview of bluetongue virus vector biology and ecology in the Oriental and Australasian regions of the Western Pacific. In Bluetongue, African horse sickness and related orbiviruses (T.E. Walton & B.I. Osburn, eds). Proc. Second International Symposium, Paris, June CRC Press, Boca Raton, Wirth W.W. & Hubert A.A (1989). The Culicoides of South-East Asia (Diptera: Ceratopogonidae). Mem. Am. Entomol. Inst., 44, Veterinaria Italiana, 40 (3), 2004

283 Vet. Ital., 40 (3), Epidemiology and vectors Field disinfestation trials against Culicoides in north-west Sardinia G. Satta (1), M. Goffredo (2), S. Sanna (1), L. Vento (1), G.P. Cubeddu (1) & E. Mascherpa (3) (1) Istituto Zooprofilattico Sperimentale della Sardegna G. Pegreffi, Via Duca degli Abruzzi 8, Sassari, Italy (2) Istituto Zooprofilattico Sperimentale Abruzzi e Molise G. Caporale, Via Campo Boario, Teramo, Italy (3) Industria Chimica Fine, Palazzo Pignano, Cremona, Italy Summary Bluetongue (BT) first affected Sardinia in August 2000, spreading rapidly across the island causing more than outbreaks and significant economic damage. Culicoides imicola Kieffer (Diptera: Ceratopogonidae) was the main vector of the disease and was also found to be the most abundant Culicoides species on Sardinia. During 2002, a field trial was conducted to evaluate the efficacy of an insecticide on local Culicoides populations in north-western Sardinia. A synthetic pyrethroid derivative (Mycrocip, ICF, Cremona, Italy) was used on two farms where outbreaks of BT had been reported; a third farm was used as control. The same treatment was repeated after 15 days. For the collection of Culicoides, two blacklight traps were placed on each farm and operated every second day for two weeks before and after insecticide treatment. Insect collections and data analyses were performed in accordance with the protocols of the Italian National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche). For each collection, the total number of insects, Culicoides spp. and C. imicola was determined. A slight decrease in the number of Culicoides collected on treated farms was recorded for only a few days after treatment. Mycrocip played a secondary role in suppressing insect numbers, but did not reduce the number of Culicoides. Indeed, periodic variations of Culicoides population sizes correlated with significant changes in weather conditions that prevailed, including oscillating temperatures, winds and relative humidity. Keywords Bluetongue Culicoides Culicoides imicola Cypermethrin Disinfestation Italy Sardinia. Introduction The appearance of bluetongue (BT) in Sardinia in August 2000 drastically reduced sheep populations and incurred extensive economic losses. Culicoides imicola Kieffer, the principal vector of the disease, was discovered to be abundant on the island in 2000 (3) and thus it would be of benefit if its numbers (adult and/or larval) could be reduced. The scientific literature on the insecticidal control of Culicoides is scarce and, furthermore, the synthetic derivatives of pyrethrum utilised to date have not given encouraging results (1, 2, 6). The micro-encapsulated pyrethroid-based product, Mycrocip, produced by ICF (Industria Chimica Fine, Cremona, Italy), is of low toxicity and is long acting. The efficacy of Mycrocip against Culicoides was evaluated in the field, through treatment trials conducted on two farms in the north-west of Sardinia. Material and methods Study sites The sites investigated were as follows: 1) Station 1 (Olmedo): north-west Sardinia, half way between Sassari and Alghero; an area of vegetable gardens and vineyards, 60 m above sea level (asl) (Fig. 1) 2) Station 2 (Bortigiadas): north-west Sardinia, between Sassari and Tempio Pausania, an agricultural area, with several cork oak plantations, 400 m asl (Fig. 2) Stations 1 and 2 are family-managed sheep and cattle farms 3) Control Station (Bonassai): a livestock reproduction institute, m from Station 1 (Fig. 3). Veterinaria Italiana, 40 (3),

284 Figure 1 Station 1 (Olmedo) Figure 4 Mycrocip spraying Stations 1 and 2 were sampled weekly from September 2000 to July 2003 (Figs 5 and 6). Insect collection Onderstepoort blacklight suction traps of the type described by Venter and Meiswinkel (7) were used to collect insects. Blacklight is 8-10 times more attractive for Culicoides than white light (8). Figure 2 Station 2 (Bortigiadas) Figure 3 Control station (Bonassai) Insecticidal treatment Two treatment trials using Mycrocip were conducted at Stations 1 and 2 fifteen days apart, the first on 27 June and the second on 12 July. The composition of 100 g Mycrocip was as follows: cypermethrin (11 g), a synthetic pyrethroid that acts by contact/ingestion, displaying rapid action and having prolonged effect; esbiothrin (1 g), with rapid neurotoxic action; piperonyl butoxide (11 g), a synergist of pyrethrins; and coformulants (77 g). Mycrocip was distributed over one hectare around each station at 1% concentration at 20 atmospheric pressures using a vaporiser mounted on a vehicle (Fig. 4). Meteorological data (temperature, humidity and wind) were provided by the regional meteorological office of Sardinia (SAR: Servizio Agrometeorologico Regionale per la Sardegna) for Stations 1 and 3. For Station 2, the data were obtained from the nearest meteorological station located in Luras. Collections and the analysis thereof were performed in accordance with the protocols of National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche) (4). Two blacklight traps (lettered A and B) were placed on each farm; they were operated every 2 or 3 days, commencing about two weeks before, and ending about two weeks after, disinfestation. The number of total insects, total Culicoides and total C. imicola was recorded for each collection. For the analysis of results, we considered three time frames: the first from 9-14 June to 27 June, the second from 28 June to 12 July and the third from 13 to July. The average total number of insects, Culicoides and C. imicola collected in the second and in the third periods, were compared to the average of those collected during the first period. Results and discussion Station 1 After the first disinfestation a reduction in total insect numbers was recorded in Trap A ( 7.3%) and in Trap B ( 51.7%). However, the number of Culicoides spp. increased by 23.1% in Trap A while a decrease of 16.7% was noted in Trap B. After the second disinfestation, total insects decreased in Trap A ( 38.5%) and B ( 42.9%), while Culicoides increased in Trap A (1.5%) and B (0.2%) (Figs 7 and 8). In addition, C. imicola gradually increased from 21.5% to 34.8% in Trap A and from 48 to 629% in Trap B in the second and third periods, respectively (Table I). 330 Veterinaria Italiana, 40 (3), 2004

285 Number of insects (log) Sep Feb Jul-01 Total Culicoides 01-Dec May Oct Mar-03 Date Culicoides imicola Number of insects (log) Jun Jun Jun Jun-02 2-Jul-02 7-Jul Jul Jul Jul Jul-02 Date Total insects Total Culicoides Culicoides imicola Figure 5 Station 1 (Olmedo): Seasonal abundance of insects (September 2000-July 2003) Figure 7 Station 1 (Olmedo): Trap A Number of insects (log) Number of insects (log) Sep Feb Jul-01 Total Culicoides 01-Dec May Oct Mar-03 Date Culicoides imicola 0 12-Jun Jun Jun Jun-02 2-Jul-02 7-Jul Jul Jul Jul Jul-02 Date Total insects Total Culicoides Culicoides imicola Figure 6 Station 2 (Bortigiadas): Seasonal abundance of insects (September 2000-July 2003) Figure 8 Station 1 (Olmedo): Trap B Table I Station 1: Traps A and B Trap Period Total insects Mean Total Culicoides Mean C. imicola Mean A First Second % % % Third % % % B First Second % % % Third % % % Veterinaria Italiana, 40 (3),

286 Station 2 After the first disinfestation, an increase in total insect numbers was found in Trap A (11.5%) and a reduction in Trap B ( 22.3%). Culicoides increased by 37.7% and 18.9% in the two traps, respectively. After the second disinfestation, total insect numbers increased in Trap A (10.0%) but were reduced in Trap B ( 30.3%); Culicoides increased by 185.0% and 161.7%, respectively (Figs 9 and 10). Culicoides imicola sharply increased in numbers from 301.5% to % in Trap A and from 89.0% to % in Trap B, in the second and third periods, respectively (Table II). Number of insects (log) Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul-02 Date Total insects Total Culicoides Culicoides imicola Figure 9 Station 2 (Bortigiadas): Trap A Number of insects (log) Jun Jun Jun Jun Jun-02 4-Jul-02 9-Jul Jul Jul Jul-02 Date Total insects Total Culicoides Culicoides imicola Figure 10 Station 2 (Bortigiadas): Trap B Control station Total insect numbers in Trap A decreased by 26.1% in the second period, but increased by 4.5% in the third period. Culicoides increased from 78.4% to 166.7% and C. imicola from 143.5% to 517.4% in the two periods, respectively (Figs 11 and 12). Total insect numbers in Trap B decreased by 50.9% in the second period and by 73.1% in the third period, whereas both Culicoides and C. imicola increased by 108% and 460%, and by 0.6% and 8.3%, in the two periods, respectively (Table III). Number of insects (log) Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Jul-02 Date Total insects Total Culicoides Culicoides imicola Figure 11 Control station (Bonassai): Trap A Number of insects (log) Jun Jun Jun Jun Jul Jul Jul Jul Jul-02 Date Total insects Total Culicoides Culicoides imicola Figure 12 Control station (Bonassai): Trap B 29-Jul Veterinaria Italiana, 40 (3), 2004

287 Table II Station 2: Traps A and B Trap Period Total insects Mean Total Culicoides Mean C. imicola Mean A First Second % % % Third % % % B First Second % % % Third % % % Table III Station 3: Traps A and B Trap Period Total insects Mean Total Culicoides Mean C. imicola Mean A First Second % % % Third % % % B First Second % % % Third % % % No significant differences were detected when the data from Traps A and B for each station were compared, nor were differences noted when the data from the two disinfestation sites were compared against those from the control station (Fig. 13). The variation in insect and Culicoides numbers observed correlate with marked oscillations in local climatic conditions and so are not the result of insecticidal disinfestation (Figs 14 and 15). Generally, an increase in minimum temperatures and a decrease in windspeed corresponded to an increase in the number of insects collected; conversely, a sudden decrease in minimum temperatures, associated with an increase in windspeed, led to a decrease in the total number of insects collected. For example, at Station 1 on 5 July, the temperature dropped to 10.7 C and resulted in a marked decrease in insect numbers. At Station 2, a slackening in the windspeed Figure 13 Variations observed at study stations 1 and 2 compared to control station for each of the three-period timeframes (1 = 9-27 June; 2 = 28 June-12 July; 3 = July) The mean number of midges captured per day refers to the first period assumed equal to 100%; disinfestations were applied on 27 June (period 1) and on 12 July (period 2) Veterinaria Italiana, 40 (3),

288 Relative humidity (%) Relative humidity (%) Jun 14-Jun 20-Jun 25-Jun 29-Jun 3-Jul 9-Jul 16-Jul 23-Jul 28-Jul 9-Jun 14-Jun 20-Jun 25-Jun 29-Jun 3-Jul 9-Jul 16-Jul 23-Jul 28-Jul Temperature ( C) Temperature ( C) Jun 14-Jun 20-Jun 25-Jun Windspeed (m/sec) 29-Jun 3-Jul 9-Jul 16-Jul 23-Jul 28-Jul 9-Jun 14-Jun 20-Jun 25-Jun 29-Jun 3-Jul 9-Jul 16-Jul 23-Jul 28-Jul Windspeed (m/sec) Jun 13-Jun 16-Jun 20-Jun 23-Jun 26-Jun 29-Jun 2-Jul 5-Jul 9-Jul 14-Jul 18-Jul 23-Jul 26-Jul 9-Jun 14-Jun 20-Jun 25-Jun 29-Jun 3-Jul 9-Jul 16-Jul 23-Jul 28-Jul Date Date Figure 14 Figure 15 Station 1 (Olmedo): meteorological data collected Station 2 (Luras): meteorological data collected during the field trials during the field trials on 2 July accompanied by a sudden increase in temperature (22.2 C), and an increase in relative humidity (74%), was associated with a peak in insect numbers (20 822) collected in Trap B. Conclusions Mycrocip plays a secondary role in the environmental control of insects. Collection results for farms treated with this product were not significantly different from those of the control station. Variations in total insect numbers appeared to be associated with prevailing weather conditions rather than with the efficacy of the insecticide used. References 1. Braverman Y. & Chizov-Ginzburg A. (1997). Repellency of synthetic and plant-derived preparations for Culicoides imicola. Med Vet. Entomol., 11 (4), l. 2. Braverman Y. & Chizov-Ginzburg A. (1998). Duration of repellency of various synthetic and plantderived preparations for Culicoides imicola, the vector of African horse sickness virus. Arch Virol. [Suppl.], 14, Goffredo M., Satta G., Torina A., Federico G., Scaramozzino P., Cafiero M.A., Lelli R. & Meiswinkel R. (2001). The 2000 bluetongue virus (BTV) outbreak in Italy: distribution and abundance 334 Veterinaria Italiana, 40 (3), 2004

289 of the principal vector Culicoides imicola Kieffer. In Proc. Tenth International Symposium of the American Association of Veterinary Laboratory Diagnosticians (AAVLD), Salsomaggiore, Parma, 4-7 July. AAVLD, Ames, Goffredo M. & Meiswinkel R. (2004). Entomological surveillance of bluetongue in Italy: methods of capture, catch analysis and identification of Culicoides biting midges. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Goffredo M., Conte A.M., Cocciolito R. & Meiswinkel R. (2003). The distribution and abundance of Culicoides imicola in Italy. Vet. Ital., 39 (47), Mullens B.A., Gerry A.C. & Velten R.K. (2001). Failure of a permethrin treatment regime to protect cattle against bluetongue virus. J. Med. Entomol., 38 (5), Venter G.J. & Meiswinkel R. (1994). The virtual absence of Culicoides imicola (Diptera: Ceratopogonidae) in the colder, high-lying area of the eastern Orange Free State, South Africa and its implications for the transmission of arboviruses. Onderstepoort J. Vet. Res., 61, Wieser-Schimpt L., Foil L.D. & Holbrook R.F. (1990). Comparison of New Jersey light traps for collection of adult Culicoides variipennis (Diptera: Ceratopogonidae). J. Am. Mosquito Control Assoc., 6, Veterinaria Italiana, 40 (3),

290 Vet. Ital., 40 (3), Susceptibility and repellency of Culicoides imicola and Culex pipiens to lambda-cyhalothrin Y. Braverman (1), A. Chizov-Ginzburg (1), H. Pener (2) & A. Wilamowski (2) (1) Kimron Veterinary Institute, PO Box 12, Beit Dagan 50250, Israel (2) Laboratory of Entomology, Ministry of Health, PO Box 34410, Jerusalem 94467, Israel Summary The basic efficacy of lambda-cyhalothrin was tested in the laboratory against newly colonised adult Culex pipiens and field-collected Culicoides imicola. C. imicola was found to be more susceptible (LD 50 =0.0098%) than Cx. pipiens (LD=0.0233%); the efficacy against both species was definitely higher than that of cyhalothrin. Lambda-cyhalothrin showed slight repellency for C. imicola during the first hour post application. Keywords Bluetongue Culex pipiens Culicoides imicola Lambda-cyhalothrin Toxicity. Introduction Culex pipiens is the most common mosquito species in Israel, where it breeds abundantly in all the known water courses throughout the country (16, 19). The species is generally considered as endophilic and anthropophilic, but not exclusively so. Birds and mammals (especially farm animals) are known to be highly attractive to females of Cx. pipiens (5, 9). Cx. pipiens is also one of the most efficient vectors of mosquito-borne arboviruses; it is implicated in the transmission of mosquito-borne mammal-associated viruses such as Rift Valley fever (20) and bovine ephemeral fever (6), and also in the transmission of mosquito-borne bird-associated viruses, especially flaviviruses such as West Nile (17) and turkey meningo-encephalitis (7, 13). Its vectorial role in pathogen transmission gives Cx. pipiens great economic importance in animal health. The biting midge Culicoides imicola Kieffer is the most important vector of livestock viral pathogens (bluetongue, Akabane and African horse sickness) in Africa, the Middle East and southern Europe (4, 21). Bovine ephemeral fever virus was also isolated from this species (14). Recently Israel turkey meningoencephalitis virus was detected from specimens of C. imicola (13). The economic loss caused by these diseases is substantial; bluetongue alone causes an estimated worldwide loss of three billion dollars annually (3) and is estimated to be increasing due to the spread of the virus into Europe. In addition, C. imicola is the major agent that causes allergic recurrent summer dermatitis in livestock in Israel (8, 26), which reduces the productivity and commercial value of the animals. Efficient control of these two species is necessary in order to prevent further transmission of pathogens and to prevent allergic dermatitis caused by their bites. While control of Cx. pipiens is generally targeted against the larval stages and adultciding is practised only as a complementary treatment, the control and deterrence of Culicoides is based on adulticiding and repelling methods with compounds that are animal and environmentally safe. The only studies performed on the efficacy of insecticides against adult C. imicola were conducted in Israel. In 1994, the pyrethroid cyhalothrin was tested in the laboratory and found to be effective against C. imicola. An extensive repellency study on natural and pyrethroid preparations was performed recently (12, 22). In recent years, an isomer of cyhalothrin, lambdacyhalothrin has been widely employed in the control of blood-sucking arthropods in medical and in veterinary entomology (2, 23). In addition, because of its long residual effect, lambda-cyhalothrin is used to impregnate bed nets for antimalarial protection (18). The present study was undertaken to test the efficacy of lambda-cyhalothrin against Cx. pipiens and its toxicity and repellency against, C. imicola. Its residual 336 Veterinaria Italiana, 40 (3), 2004

291 effect could provide an effective tool in the control of both species, especially in and around animal shelters. Materials and methods Susceptibility tests were performed in the laboratory on a suction light trapped field population of mainly nulliparous C. imicola, which constituted the bulk of the catch (11) and also on 3- to 5-day-old newly colonised females (50%) and males (50%) of Cx. pipiens. Both populations were collected at Beit Dagan in Israel. The colony of mosquitoes was started from larvae whereas C. imicola adults were collected with three suction light traps (15). In the laboratory, the tested insects were acclimatised for about an hour prior to testing. Test procedures were essentially those described by the World Health Organization (WHO) (25) for routine susceptibility tests of adult biting midges and mosquitoes. The insects were exposed to lambda-cyhalothrinimpregnated papers for 1 h in test kit tubes, held horizontally. Mortality was recorded after a 24 h recovery period, during which 10% sugar solution was provided on cotton wool pads. The test kit tubes were kept in an insectary throughout the experiment and were maintained at 28 C±2 C and 80±2% relative humidity. Test papers were prepared by impregnating Whatman no. 1 filter paper (13.5 cm 14.5 cm) with 2 ml of a solution containing a known concentration of 125 mg/ml technical grade lambdacyhalothrin (Syngenta) in reagent grade acetone. Control papers were impregnated with 2 ml of acetone only. The tests were replicated three times for Cx. pipiens and seven times for C. imicola. In each test, three groups of 30 adults of both sexes of Cx. pipiens and females of C. imicola were tested for each concentration. The results for each concentration in valid tests (control 10%) were pooled and evaluated by probit regression analysis (Spss 8.0 for Windows). In vitro repellency tests with C. imicola were conducted in the field from 6 pm until 5 am, using four blacklight suction traps, where each trap was used for a different treatment, as described previously (11). The experiment was performed using 5% lambda-cyhalothrin and preparations containing 4% and 24% of cyhalothrin as well as a control (water). Results and discussion The results of the measurement of the susceptibility of both species to lambda-cyhalothrin are summarised in Table I and Figure 1. Table I LD 50 and LD 90 values (%) following 1-h exposure to lambda-cyhalothrin (in brackets: 95% confidence limits) Species LD 50 LD 90 Culex pipiens Culicoides imicola ( ) ( ) ( ) ( ) While adult C. imicola were more susceptible than adult Cx. pipiens (LD 50 for Cx. pipiens = %, for C. imicola = %), the efficacy of lambdacyhalothrin was quite high for both species. Lambdacyhalothrin was found to be nearly 10 times as effective as cyhalothrin against Culicoides (10), and the same has been found for Cx. pipiens (internal reports, Entomology Laboratory, Ministry of Health, Israel). Consequently, control of these two vectors could be achieved with a much smaller quantity of insecticide than is currently being used, which would be beneficial both environmentally and economically. It was observed that during the toxicity test, lambdacyhalothrin was not found to be repellent to either of the species tested as they landed on the filter paper. Probit Culex pipiens Culicoides imicola Log of dose (percentage ai) Figure 1 Baseline susceptibility to lambda-cyhalothrin Dose mortality curve Veterinaria Italiana, 40 (3),

292 Lambda-cyhalothrin is an insecticide with contact and stomach action and repellent properties (24). According to Artimev et al. (1) lambda-cyhalothrin did not repel Cx. pipiens. Therefore, we did not conduct special tests for repellency to this species. No information could be found on its repellency to Culicoides spp., therefore we conducted a repellency test for C. imicola; the results show that its repellency for C. imicola was poor and lasted for up to 1 h and for the two preparations of cyhalothrin, no repellency was detected (Table II). The residual activity of the lambda-cyhalothrin was not tested in the present study. However, since the compound is used against malaria-transmitting mosquitoes, both as a space spray and in the impregnation of bed nets (18), it can be safely assumed that it would also have a long-lasting effect against C. imicola and Cx. pipiens. Table II Repellency of 5% lambda-cyhalothrin and cyhalothrin to Culicoides imicola; number of trapped C. imicola Treatments Hours post-application Control (water) Lambdacyhalothrin 5% Sylotox-4 (4% cyhalothrin) Sylotox-20 (24% cyhalothrin) References 1. Artimev M.M., Sorokin N.N., Aliev, A.I., Stepanova A.N., Demyanova E.V., Bakiev R.A., Chabanenko A.A. & Labzin V.V. (1991). Testing the insecticides Icon and Ficam against mosquitoes in the south of the USSR. Med. Parazitol., 1, Barros A.T., Alison M.W. & Foil L.D. (1999). Evaluation of a yearly insecticidal ear tag rotation for control of pyrethroid-resistant horn flies (Diptera: Muscidae). Vet. Parasitol., 82, Bath G.O. (1989). Bluetongue. In Proc. 2nd International Congress for sheep veterinarians, Massey University, Palmerston North, February. New Zealand Veterinary Association, Wellington, Braverman Y. (1994). Nematocera (Ceratopogonidae, Psychodidae, Simuliidae, Culicidae) and control methods. Rev. Sci. Tech. Off. Int. Épiz., 13, Braverman Y. & Rubina M. (1976). Light trapping of biting insects in poultry houses in Israel. Isr. J. Zool., 25, Braverman Y. (2001). The vectors of bovine ephemeral fever, Akabane and bluetongue viruses in Israel. In Proc. 13th Symposium of Dairy Cattle Science, Zichron Yaakov, Israel, February. Israel Ministry of Agriculture, Beit Dagan, Braverman Y., Rubina M. & Frish K. (1981). Pathogens of veterinary importance isolated from mosquitoes and biting midges in Israel. Ins. Sci. Applic., 2, Braverman Y., Ungar-Waron H., Frish K., Adler H., Danieli Y., Baker K.P. & Quinn P.J. (1983). Epidemiological and immunological studies of sweet itch in horses in Israel. Vet. Rec., 112, Braverman Y., Kitron U. & Killick-Kendrick R. (1991). Attractiveness of vertebrate hosts to Culex pipiens (Diptera: Culicidae) and other mosquitoes in Israel. J. Med. Entomol., 28, Braverman Y., Wilamowski A. & Chizov- Ginzburg A. (1995). Susceptibility of Culicoides imicola to cyhalothrin. Med. Vet. Entomol., 9, Braverman Y. & Chizov-Ginzburg A. (1997). Repellency of synthetic and plant derived preparations for Culicoides imicola. Med. Vet. Entomol., 11, Braverman Y., Wegis M. & Mullens B.A. (2000). Response of Culicoides sonorensis (Diptera: Ceratopogonidae) to 1-Octen-3-ol and three plantderived repellent formulations in the field. J. Am. Mosq. Control Assoc., 7, Braverman Y., Rechtman S., Frish A. & Braverman R. (2003). Dynamics of biting activity of C. imicola Kieffer (Diptera: Ceratopogonidae) during the year. Isr. J. Vet Med., 58, Davies J.A. & Walker A.R. (1974). The isolation of ephemeral fever virus from cattle and Culicoides midges in Kenya. Vet. Rec., 20, DuToit R.M. (1944). The transmission of bluetongue and horse-sickness by Culicoides. Onderstepoort J. Vet. Sci. Anim. Ind., 19, Kitron U. & Pener H. (1986). Distribution of mosquitoes (Diptera: Culicidae) in northern Israel: a historical perspective. II. Culicine mosquitoes. J. Med. Entomol., 23, Lundstrom S.O. (1999). Mosquito-borne viruses in Western Europe: a review. J. Vect. Ecol., 24, Marbiah N.T., Petersen E., David K., Magbity E., Lines J. & Bradely D.J. (1998). A controlled trial of lambda-cyhalothrin-impregnated bed nets and/or 338 Veterinaria Italiana, 40 (3), 2004

293 dapsone/pyrimethamine for malaria control in Sierra Leone. Am. J. Trop. Med. Hyg., 58, Margalit J. & Tahori A.S. (1974). An annotated list of mosquitoes in Israel. Isr. J. Entomol., 9, Meegan J.M., Khalil G.M., Hoogstraal H. & Adham F.K. (1980). Experimental transmission and field isolation studies implicating Culex pipiens as vector of Rift Valley fever virus in Egypt. Am. J. Trop. Med. Hyg., 29, Mellor P.S. (1993). Culicoides do vectors respect international boundaries? Br. Vet. J., 49, Mullens, B.A., Velten R.K., Gerry A.G., Braverman Y. & Endris R.G. (2000). Feeding and survival of Culicoides sonorensis on cattle treated with permethrin or primiphos-methyl. Med. Vet. Entomol., 14, Talbert A., Nyange A. & Molteni F. (1998). Spraying tick-infested houses with lambdacyhalothrin reduces the incidence of tick-borne relapsing fever in children under five years old. Trans. R. Soc. Trop. Med. Hyg., 92, Tomlin C.D.S. (2000). The pesticide manual, 12th Ed. British Crop Protection Council, Farnham, Surrey, World Health Organization (WHO) (1981). Instructions for determining the susceptibility or resistance of adult blackflies, sandflies and biting midges to insecticides. WHO, Geneva, Mimeographed document. WHO/VBC/ Yeruham I., Braverman Y. & Orgad U. (1993). Field observations in Israel on hypersensitivity in cattle, sheep and donkeys caused by Culicoides. Aust. Vet. J., 70, Veterinaria Italiana, 40 (3),

294 Vet. Ital., 40 (3), Christopher Columbus and Culicoides: was C. jamaicensis Edwards, 1922 introduced into the Mediterranean 500 years ago and later re-named C. paolae Boorman 1996? R. Meiswinkel (1), K. Labuschagne (2) & M. Goffredo (1) (1) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, via Campo Boario, Teramo, Italy (2) Agricultural Research Council (ARC)-Onderstepoort Veterinary Institute (OVI), Private Bag X05, Onderstepoort 0110, South Africa Summary The biting midge, Culicoides paolae Boorman, described from specimens collected in the extreme south of Italy in 1996, belongs in the subgenus Drymodesmyia. This subgenus was erected by Vargas in 1960 for the so-called Copiosus species group, an assemblage of 22 species endemic to the tropical regions of the New World and, where known, breed in vegetative materials including the decaying leaves (cladodes) and fruits of Central American cacti. The Mexican peoples have utilised these cacti for over years; one of these, Opuntia ficus-indica Linnaeus, was brought to Europe by Christopher Columbus following his voyages of discovery. As a taxon C. paolae is very similar to the Central American C. jamaicensis Edwards, 1922 raising the possibility that it (or a closely related species of Drymodesmyia) was introduced into the Mediterranean Region at the time of Columbus, but was (perplexingly) discovered only 500 years later and named C. paolae. The comparison of Sardinian specimens of C. paolae with Panamanian material of C. jamaicensis (housed in the Natural History Museum in London) confirmed the two species to be very similar but unusual differences were noted around the precise distribution of the sensilla coeloconica on the female flagellum. Until it is understood whether these differences represent either intra- or interspecific variation, the question of the possible synonymy of C. paolae must be held in abeyance. Keywords Culicoides Culicoides jamaicensis Culicoides paolae Taxonomy. Introduction In 1994 an apparently new species of Culicoides was captured at a horse stable in Pellaro, southern Italy. It was described subsequently and named C. paolae Boorman, 1996 (after its discoverer Paola Scaramozzino). Based upon similarities in the wing pattern C. paolae was first thought to belong in the Old World Schultzei Complex (= subgenus Remmia Glukhova, 1977) but closer scrutiny confirmed it to differ in many other taxonomic features. Despite these differences, the superficial resemblance between C. paolae and species of the Schultzei Complex was emphasised. This led to it being labelled a potential vector of livestock orbiviruses, firstly because it had been collected around horses and, secondly, because epizootic haemorrhagic disease of deer virus (EHDV) had previously been isolated in the Sudan from C. kingi Austen, 1912, a species of the Schultzei Complex. Following the incursion of bluetongue (BT) virus (BTV) into Italy in August 2000, a national survey was implemented and Culicoides collected countrywide. Onderstepoort blacklight traps were deployed throughout the islands of Sardinia and Sicily and on the southern peninsula of mainland Italy. These soon revealed C. paolae to be widespread, and also that it could, on occasion, be captured in 100s (but never in 1 000s). In 2001, it was noticed adventitiously that the wing of C. paolae closely resembled that of C. jamaicensis Edwards, However, this resemblance was 340 Veterinaria Italiana, 40 (3), 2004

295 initially ascribed to congruence as many species of world Culicoides share similar wing patterns. Also, the fact that C. jamaicensis was a Central American (New World) species seemed to weigh too heavily against its dispersal across such a wide expanse of ocean into the Mediterranean Basin (Old World). However, doubts persisted, firstly, because perusal of the published descriptions of the two species indicated that they were very similar indeed, and, secondly, because C. paolae seemed unrelated taxonomically to any other Old World species of Culicoides. These two facts heightened the likelihood of C. paolae being alien to the Mediterranean, and so it was decided to investigate the congruence in greater depth. a) Female wing b) Thoracic pattern c) Palpus Materials and methods Males and females of C. paolae captured in light traps on the island of Sardinia were slide-mounted. These were then compared in detail to the original descriptions (and re-descriptions) of C. jamaicensis (Fig. 1) and C. paolae; two slide-mounted specimens of C. jamaicensis from Panama, stored in the Natural History Museum in London, were also examined. Information on the biology of C. jamaicensis (and of related species) was also collated. All the taxonomic data were also compared against those obtained from the study of hundreds of afrotropical specimens of eight species of the Schultzei Complex. Two attempts were made in the field (in Italy) to harvest the immature stages of C. paolae. d) Tibial comb e) Spermathecae f) Male paramers Results and discussion The taxonomic data on C. jamaicensis from various studies (1, 6, 9, 10) were compared with those presented in the original description of C. paolae (3) and with those gleaned from six specimens (three males and three females) collected in Sardinia. Based upon these published data, the two species seemed inseparable, and suggested C. paolae to be a junior homonym of C. jamaicensis. However, published descriptions of world Culicoides nearly always lack important species-specific details, and for this reason it was decided to also study New World material of identified C. jamaicensis. Two female specimens from two localities in Panama were located in the holdings of the Natural History Museum in London, and were examined in detail. At first, there seemed little doubt that C. paolae and C. jamaicensis were one and the same species. However, upon closer examination, the Panamanian specimens displayed an unusual conformation in the distribution of the sensilla coeloconica on the female flagellum in that they were found to occur on both faces of each of flagellomeres IV-VII. For example, on all four flagella examined, the four coeloconica g) Male genitalia (paramers removed) Figure 1 Culicoides jamaicens Edwards Adapted from Wirth and Blanton (9) found on flagellomere IV were always split into two distinct groups (of two sensilla each), and were positioned (in direct apposition to each other) on the dorsad and ventrad faces of this flagellomere. There are many species of world Culicoides that have multiple sensilla coeloconica on one or more of the Veterinaria Italiana, 40 (3),

296 basal flagellomeres but even where these may number up to 16 they are always found on one face of the flagellomere and are tightly grouped. Thus, it is highly unusual to find a species of Culicoides in which the coeloconica are split into two widely separate groups on a single flagellomere. This splitting of the coeloconica into two groups was not seen in the three Sardinian females (i.e. six antennae) of C. paolae examined. At this stage, it is not possible to decide the taxonomic importance of this difference. The study of further material of the subgenus Drymodesmyia from across a wider New World range is required to answer this question. We thus refrain from declaring C. paolae a junior homonym of C. jamaicensis. In 1960, Vargas created the subgenus Drymodesmyia for those species of Culicoides that belong to the Copiosus group, which included C. jamaicensis (8). This fairly large subgenus of 22 species (4) is endemic to the New World, where some of its member species have been found to breed in the rotting parts of Central American cacti (10). No species of Drymodesmyia has ever before been reported from anywhere in the Palaearctic Region (nor from Africa), and so would support the contention that C. paolae is an introduced species and thus alien to the Old World (which includes the Mediterranean Basin). This, in turn, suggests that it may therefore be a synonym of a previously described New World species. However, it is possible also that still other species of Drymodesmyia remain to be discovered in Central America and that one of them could well prove to be C. paolae. In such an event, C. paolae would remain a valid taxon with the unusual distinction of having been described from well beyond its faunal home range. The clarification of this issue requires the morphological study of all species of Drymodesmyia and, ideally, should be coupled to the sequencing of targeted gene regions. In regard to the question of C. paolae having been introduced into the Mediterranean, it is pertinent to recall that another Central American species of Drymodesmyia was unexpectedly discovered in Australia (5). The species is C. loughnani Edwards, 1922, and most likely arrived there after boatloads of parasite-laden rotting cacti stems had been introduced from Central America to Australia in the 1920s as part of a biological control effort against the spread of jointed cactus (Opuntia sp.). Culicoides loughnani was subsequently found to breed in the rot pockets that formed in cacti stems in Australia just as the closely related C. jamaicensis has been found to do in Mexico. Ad hoc efforts to rear C. paolae from Opuntia in Italy have thus far failed and are discussed further below. The prickly pear Opuntia ficus-indica Linnaeus apparently took its name from its alleged morphological similarity to the Mediterranean fig and from its geographical origin (the West Indies). It was introduced by Christopher Columbus into Spain in around 1500 to be cultivated in the gardens of the nobles (2); its strangely odd form led Oviedo to refer to it as the monster among trees. It subsequently was spread throughout the hotter areas of the Mediterranean Basin by sailors who used it as a vegetable against scurvy. In these new locations in Europe, the spined and spineless forms of Opuntia were described subsequently (and erroneously) by botanists as new species. Apparently the domestication of O. ficus-indica dates back some years to the ancient Mexicans who referred to it as the sweet song plant due to the sound made through the leaves by the blowing wind; it is also known as the bone fixing tree as the leaves (more correctly cladodes) are used in poultices to treat bone fractures. The distribution of C. paolae in Italy has yet to be mapped thoroughly but current indications are that it, like Opuntia, is restricted to hotter climes, being found widely on the islands of Sardinia, Sicily and Malta (7), and on the southern third of peninsular Italy (which includes Pellaro the type locality of C. paolae). Whilst a number of species of Drymodesmyia have been shown to breed in cacti in Central America, two attempts to breed C. paolae from the ripe fruits, and from the rotting cladodes, of Opuntia in Italy, have failed. Thus the reputed association between the insect and this plant in the Mediterranean has still to be demonstrated. In the original description of C. paolae, it was noted that because of its resemblance to species of the Schultzei Complex, and because it had been captured in abundance at a horse stable, it deserved consideration as a potential vector of orbiviruses to livestock. However, the adult female of C. paolae possesses three remarkable features that would seem to mitigate against this supposition, as follows: a) the third palpal segment, which bears the hostseeking sensory pit, is extremely inflated b) all basal flagellomeres III-X bear two long and two short sensilla trichodea that are not slender but inflated c) multiple sensilla coeloconica occur on each of flagellomeres III-XV. Although the host preferences of C. paolae are unknown, these three features suggest it to be ornithophilic in its bloodsucking habits. If this is the case, the original specimens of C. paolae captured at Pellaro may have been feeding on birds (or chickens) roosting in the vicinity of the light trap and not upon the stabled horses. 342 Veterinaria Italiana, 40 (3), 2004

297 Finally, C. paolae is taxonomically quite dissimilar from eight species of the Schultzei Complex (= subgenus Remmia) that were studied. They also have little in common in terms of biology. Indeed, based upon the morphological data, their evolutionary links are, at best, tenuous, as the subgenus Drymodesmyia is restricted to the New World and Remmia to the Old World. Also, and contrary to a widespread belief amongst culicoidologists, the subgenus Remmia is not a synonym of the New World subgenus Oecacta. Conclusions This story of C. paolae (and of C. loughnani in Australia) demonstrates clearly that species of Culicoides can be introduced by man into new localities and from across wide expanses of ocean. However, their successful establishment would depend upon a suitable larval habitat being available. In the case of C. paolae (and C. loughnani), it would seem that they could establish themselves because the larval host plants (cacti) were introduced simultaneously. Their establishment, and subsequent maintenance, would require at least one adaptive step, i.e. a switch to sucking blood from new hosts. In the case of these two taxa, their hosts are likely avian. It is also possible that species of Drymodesmyia are autogenous (i.e. do not require a blood-meal to lay their first batch of eggs) and, if so, would aid further in their ability to survive and persist in their new homes. The fact that hundreds of specimens of C. paolae can be captured in a given locale, and over a considerable area in the central Mediterranean, would seem to attest to the apparent success of C. paolae, but does make it difficult to explain why the presence of this species has been discovered only very recently. Is this due simply to a paucity of studies on the biting midge fauna of the Mediterranean or is there another explanation? If indeed C. paolae was introduced from the West Indies some 500 years ago, it would be interesting to establish the degree of genetic drift that has occurred since. Such a study may provide molecular markers for dating the arrival and movement of taxa into new regions. A pertinent example is that of the BT vector C. imicola, which is believed by some to have arrived recently in the Mediterranean and is spreading rapidly northwards into Europe. On the taxonomic level this case of possible mistaken identity is not easily resolved. As noted above, small (but unusual) differences were found between C. paolae from the Mediterranean and C. jamaicensis from Panama. If these represent intraspecific variation, they would be highly unusual for the genus Culicoides. On another level, it is possible also that the Panamanian specimens in the National History Museum have been misidentified, which is not unlikely when many species complexes remain entirely hidden from view because inter- and intraspecific variation is consistently being confused. The onus is upon taxonomists to explore these variations in greater detail, through larger series of specimens collected over a wider geographic range. It is an inescapable fact that many species of Culicoides remain undiscovered because often their distinctness as genetic entities is not reflected by glaringly obvious changes in the phenotype. The elucidation of the true identity of C. paolae will require that it be more intensively compared against each of the 22 species currently deemed to comprise the subgenus Drymodesmyia, and that such a study be conducted on both morphological and molecular levels. The wing pattern of C. paolae resembles that found in species of the Schultzei Complex; all share an hourglass-shaped pale spot in the centre of wing cell R5. Given the fact that the viruses of EHD and BT had previously been isolated from Remmia in Africa, C. paolae was implicated as a potential new vector. This deduction, based upon the tenuous taxonomic link between C. paolae and species of Remmia, is weakened further by the apparent vast differences in their respective biologies (which includes the possibility that C. paolae feeds preferentially on birds). However, these evidences should not be construed as mitigating unequivocally against the vector potential of C. paolae; rather they serve to illustrate that too little is known about the host preferences and breeding habitats of the 100 or more species of Culicoides that occur across the Mediterranean Basin. It is important for us to appreciate that our current reliance upon the light trap as the sole surveillance tool, though being of great value in determining the seasonal and geographic distribution and adult densities of Culicoides vectors, will contribute little towards elucidating the oft unusual life-cycles of these small blood-sucking insects. References 1. Aitken T.H.G., Wirth W.W., Williams R.W., Davies J.B. & Tikasingh E.S. (1975). A review of the bloodsucking midges of Trinidad and Tobago, West Indies (Diptera: Ceratopogonidae). J. Entomol. (B), 44, Barbera G. & Inglese P. (2001). Ficodindia. L Epos, Ed. C. di Biagio & C. Cortimiglia, Palermo, Sicily. 3. Boorman J., Mellor P.S. & Scaramozzino P. (1996). A new species of Culicoides (Diptera, Ceratopogonidae) from southern Italy. Parassitologia, 38, Veterinaria Italiana, 40 (3),

298 4. Borkent A. & Spinelli G.R. (2000). Catalog of the New World biting midges south of the United States of America (Diptera: Ceratopogonidae). Contrib. Entomol. Int., 4, Dyce A.L. (1969). Biting midges (Diptera: Ceratopogonidae) reared from rotting cactus in Australia. Mosquito News, 29, Edwards F.W. (1922). On some Malayan and other species of Culicoides, with a note on the genus Lasiohelea. Bull. Entomol. Res., 13, Goffredo M., Buttigieg M., Meiswinkel R., Delécolle J.-C. & Chircop S. (2004). Entomological surveillance for bluetongue on Malta: first report of Culicoides imicola Kieffer. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Vargas L. (1960). The subgenera of Culicoides of the Americas (Diptera: Ceratopogonidae). Rev. Biol. Trop., 8, Wirth W.W. & Blanton F.S. (1959). Biting midges of the genus Culicoides from Panama (Diptera: Heleidae). Proc. US Natl Mus., 109, Wirth W.W. & Hubert A.A. (1960). Ceratopogonidae (Diptera) reared from cacti, with a review of the copiosus group of Culicoides. Ann. Entomol. Soc. Am., 53, Veterinaria Italiana, 40 (3), 2004

299 Vet. Ital., 40 (3), Epidemiology and vectors Adult characters defining and separating the Imicola and Orientalis species complexes of the subgenus Avaritia Fox, 1955 (Culicoides, Diptera: Ceratopogonidae) R. Meiswinkel Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Campo Boario, Teramo, Italy Research affiliate: Agricultural Research Council (ARC)-Onderstepoort Veterinary Institute (OVI), Private Bag X05, Onderstepoort 0110, South Africa Summary Thirty-six subgenera comprise the biting midge genus Culicoides Latreille, One of these, the relatively small subgenus Avaritia Fox, 1955, is the most important as it contains nearly half of the 30 world species of Culicoides known to play a greater or lesser role in the transmission of orbiviral diseases to livestock pantropically. These diseases include bluetongue (BT), African horse sickness (AHS) and epizootic haemorrhagic disease of deer (EHD). The subgenus Avaritia is distributed globally and the 70 species described have been subdivided into seven species groups and four subgroups. These 11 informal categories, variously labelled as either subgroups, groups or complexes, are reduced here to seven in number (six species complexes and one subgroup) and although they appear natural, they are nearly all poorly defined. In this study two of these, namely the Imicola and Orientalis species complexes, are re-evaluated to determine which morphological characters define them more precisely, and so may help to establish their monophyly in the future. The two complexes are separable on eight discrete adult characters (two in the female, six in the male); these characters, and three secondary ones, are discussed and illustrated. The Imicola and Orientalis Complexes together embrace 23 species; these species are assigned to their respective complexes and according to recent nomenclatural adjustments. The taxonomy of vector Culicoides worldwide remains superficial; to improve the situation it is recommended that the traditional morphological method be integrated with the modern molecular approach. Keywords Culicoides Imicola Complex Orbivirus vector Orientalis Complex Subgenus Avaritia Taxonomy. Introduction The blood-sucking Dipteran genus Culicoides was established by Latreille in It has been divided since into 36 subgenera, which have been listed by Borkent and Wirth (4). One of these is the subgenus Avaritia. It was created by Fox in 1955 (8), and the Holarctic species C. obsoletus (Meigen) designated as subgenotype. Of the 109 names available for species of Avaritia, only 63 were considered valid a decade ago and 46 considered synonyms (3). The number of Avaritia species has grown since to about 70, approximately 5% of the total Culicoides world fauna of species. The employment of Avaritia by taxonomists worldwide has been chequered; most authors, until quite recently, preferred to assign species to groups and subgroups only, and often without reference to the subgenus (5, 9, 11, 12, 14). This dichotomous approach may have found cause in the fact that no other species of Avaritia outside of the Holarctic possess the complicated, almost flamboyantly aberrant, genitalia found in the subgenotype C. obsoletus and its true congeners C. gornostaevae Mirzaeva, 1984, C. montanus Shakirzjanova, 1962, C. sanguisuga (Coquillett), 1901, C. scoticus Downes and Kettle, 1952 and C. sinanoensis Tokunaga, Veterinaria Italiana, 40 (3),

300 More recently, there has been a shift towards employing Avaritia more consistently and correctly (2, 6, 7, 10, 15, 18, 19, 20, 21, 22, 25, 27); others have gone a step further by acknowledging that distinct lineages comprise the subgenus (1, 3, 11, 22, 27, 28). Accordingly, seven species groups and four subgroups have been proposed. In the former category are listed the Obsoletus, Imicola (= Pallidipennis), Andicola, Actoni, Pusillus, Orientalis and Montanus groups (3, 5, 9, 14, 28, 29), and in the latter category the Grahamii, Trifasciellus, Imicola and Pseudopallidipennis subgroups (11, 23). In regard to the former category, all, except the Montanus group (3), appear to be valid. However, they are either not delineated (Andicola and Pusillus groups), have been established in the female only (Actoni and Orientalis groups), or are too broadly defined (Obsoletus, Imicola and Orientalis groups). As a result, their usage, especially regarding the last trio, is confused. To address this, characters that define, and separate, the Imicola and Orientalis Complexes, are presented here. It is hoped that this will create the base for establishing their monophyly in future, and for defining the remaining species complexes. Materials and methods Approximately slide-mounted males and females of 12 species (some undescribed) of the Imicola Complex were examined; these are the specimens listed in previous publications (19, 20, 21, 22, 23). Fifty males and females of 11 species of the Orientalis Complex (collected in Thailand and the Philippines) were examined; these were identified using either keys, illustrations or descriptions in the works cited. In the female, the two characters of most taxonomic weight are compared in Figure 1 (ad), and those for the male compared in Figure 2 (ah). The illustrations of the male genitalia shown in Figure 2 (a-d) are of C. orientalis Macfie, 1932, collected in Thailand, whilst Figure 2 (e-h) are of C. imicola (South Africa). Observations and illustrations were made from slide-mounted material in which the characters were symmetrically displayed, and not in any way distorted by compression during coverslipping. Three additional discriminatory characters (of secondary weight) are also considered because of their prominence in the literature: wing pattern, the sensilla coeloconica distribution on the female flagellomere and the spines on the hind tibial comb. As to what constitutes a taxonomic character, the methods outlined by Mayr et al. were followed (17), namely: A character in systematics may be defined as any feature which may be used to distinguish one taxon from another. Recent nomenclatural adjustments (4) are adopted as reflected in Table I. Orientalis Complex a) c) b) d) Imicola Complex Figure 1 Female abdomen a) and c) dorsal pigmented terga b) and d) sclerotised plates surrounding gonopore Results Principal characters defining and separating the Imicola and Orientalis Complexes Abdomen : dorsal pigmented terga As shown in Figure 1a, species of the Orientalis Complex always have large, rectangular, darkly pigmented terga dorsally on abdominal segments II- VII; in the Imicola Complex they are reduced in size especially on segments III-V where they appear almost round in shape (Fig. 1c). This feature is the most reliable for assigning a specimen to its respective complex when viewed under the dissecting microscope. Unfortunately, the terga are only clearly visible in nulliparous females as the burgundy pigmentation laid down in the abdomens of older parous and gravid females tends to obscure them. Similarly, during slide-mounting, these terga fade in material treated in potassium hydroxide (KOH), and are further obscured in those females whose abdomens are mounted ventral side up (for better examination of the spermathecae and of the sclerotised plates surrounding the gonopore). 346 Veterinaria Italiana, 40 (3), 2004

301 Abdomen : sclerotised plates surrounding gonopore As illustrated in Figure 1 (b and d), the precise form of the pigmented plates embracing the gonopore differs between the two complexes. They are simpler in the Imicola Complex, i.e. lack the forefinger and thumb-like projections that partially encircle the gonopore opening as seen in all Orientalis Complex species. Their precise delineation depends upon specimens being carefully prepared, and in which the distal segments of the abdomen have not been telescoped. Antenna : sensilla coeloconica on flagellomeres All species of the Orientalis Complex have sensilla coeloconica distributed on flagellomeres III, XI-XV. Nine of the 12 species of the Imicola Complex differ in having fewer coeloconica, i.e. on flagellomeres III, XII-XV only. However, the three species of the Pseudopallidipennis subgroup (within the Imicola Complex) have coeloconica on III, XI-XV, while a fourth species, C. nudipalpis, has this distribution in approximately 50% of specimens. It is fair to say that some 80% of the 70 species in the subgenus Avaritia found worldwide will share one of these two coeloconica distribution patterns, irrespective of species complex, and so they have limited value as a distinguishing character. Legs : hind tibial comb The Trifasciellus subgroup (= the Orientalis Complex) was erected on the fact that the first spine of the hind tibial comb is noticeably longer and thicker than the four adjoining spines (11). However, in this study, the first spine was found to be equally long for both species complexes. Whilst it does appear to be more darkly pigmented and robust in the Orientalis Complex, the degree of pigmentation can be affected by the particular clearing and slidemounting protocol used. Genitalia : aedeagus In the Orientalis Complex, the infuscated peg, which is a prolongation of the distal process that projects anteriorly into the aedeagus, narrows to a sharp or slightly blunt point (Fig. 2b), and is rather smoothly pigmented. In the Imicola Complex, the anterior end of the peg differs in that it usually expands into an irregular, almost amorphous shape (Fig. 2f), and is not smoothly pigmented, but granular. Whilst this seemingly insignificant character is highly diagnostic for the Imicola Complex, the form of the peg seen in the Orientalis Complex is more likely to reappear in other species complexes yet to be defined. Orientalis Complex a) e) b) f) c) g) d) h) Figure 2 Male genitalia a) and e) parameres b) and f) aedeagus c) and g) gonocoxite d) and h) tergum nine Imicola Complex Veterinaria Italiana, 40 (3),

302 Table I Reassignment (by faunal region) of 23 Old World species of Culicoides of the subgenus Avaritia to the Orientalis and Imicola species complexes; nomenclatural adjustments follow Borkent and Wirth (4) Orientalis Complex Oriental/Australasian regions dumdumi Sen and Das Gupta, 1959: 628. India flavipunctatus Kitaoka, 1975: 199. Nansei Islands (Japan) fulvus Sen and Das Gupta, 1959: 628. India hui Wirth and Hubert, 1961: 16. Taiwan jacobsoni Macfie, 1934: 215. Indonesia buckleyi Macfie, 1937: 117. Malaysia kitaokai Tokunaga, 1955: 6. Japan unisetiferus Tokunaga, 1959: 236. Papua New Guinea obscurus Tokunaga and Murachi, 1959: 347. Belau (USA) pungens de Meijere, 1909: 200. Sumatra orientalis Macfie, 1932: 490. Malaysia nayabazari Das Gupta, 1963: 35. India tainanus Kieffer, 1916: 114. Taiwan maculatus (Shiraki), 1913: 294. Taiwan kii Tokunaga, 1937: 284. Japan sigaensis Tokunaga, 1937: 322. Japan kyotoensis Tokunaga, 1937: 329. Japan suborientalis Tokunaga, 1951: 106. Indonesia Afrotropical region brosseti Vattier and Adam, 1966: 297. Gabon dubitatus Kremer, Rebholtz-Hirtzel and Delécolle, 1976: 233. Angola trifasciellus Goetghebuer, 1935: 175. Zaire Imicola Complex Oriental/Palaearctic/Australasian regions brevitarsis Kieffer, 1917: 187. Australia robertsi Lee and Reye, 1953: 386. Australia (Queensland) radicitus Delfinado, 1961: 657. Philippines superfulvus Das Gupta, 1962: 253. India nudipalpis Delfinado, 1961: 655. Philippines imicola Kieffer, 1913: 11. Kenya iraqensis Khalaf, 1957: 343. Iraq minutus Sen and Das Gupta, 1959: 622. India pseudoturgidus Das Gupta, 1962: 537. India Afrotropical region bolitinos Meiswinkel, 1989: 30. South Africa imicola Kieffer, 1913: 11. Kenya pallidipennis Carter, Ingram and Macfie, 1920: 265. Ghana kwagga (R. Meiswinkel, 1995, unpublished MSc thesis) loxodontis Meiswinkel, 1992: 147. South Africa miombo Meiswinkel, 1991: 161. Malaŵi pseudopallidipennis Clastrier, 1958: 197. Senegal tuttifrutti Meiswinkel, Cornet and Dyce, 2003: 42. South Africa Genitalia : aedeagus In the Orientalis Complex, the infuscated peg is connected to the lateral converging sclerotised arms of the aedeagus by a lightly pigmented membranous arch (Fig. 2b). This arch is better developed in some species than in others but is absent in all species of the Imicola Complex. Genitalia : parameres In the Imicola Complex, all species, without exception, have the tip of the parameres erect, sharp and simple (Fig. 2e). In the Orientalis Complex, the tips of the parameres are always sinuous, are most often limp and recurved, and are usually finely to conspicuously feathered (Fig. 2a). (In some species, this tip has been described as bare but this requires confirmation throughout the Orientalis Complex as its precise observation depends upon high magnification [ ], and upon material being well prepared.) Genitalia : tergum nine Figure 2d shows that in the Orientalis Complex the apicolateral processes or, more correctly, the flanged processes, are broad, subtlely triangular in shape, and most importantly, the apex of the process is positioned either laterally, sublaterally or submedianally. Furthermore, these flanges are always narrowly and fairly abruptly separated medianally by a deep and clearly pigmented subtriangular excision (Fig. 2d). In the Imicola Complex, these flanged processes are not as broadly developed, and arise always on the lateral corners of the tergum (Fig. 2h). The processes are thus always broadly separated by a gently concave posterior margin that is very saddle-like ( sway-backed ) in shape and is not infuscated but can, on occasion, be longitudinally striated medially; the posterior margin is also never excised medianally (Fig. 2h). Genitalia : tergum nine In the Orientalis Complex, the tergum is trapezoidal in shape, gradually narrowing posteriorly (Fig. 2d); in the Imicola Complex, it is almost square as wide anteriorly as posteriorly, and noticeably waisted medially (Fig. 2h). This character is best observed in specimens where the genitalia have not been compressed out of their natural shape by 348 Veterinaria Italiana, 40 (3), 2004

303 coverslipping. In compressed specimens, the anterior half of the tergum, in being three-dimensional where it is fused to the narrow strip that forms sternum 9, tends to splay outwards, i.e. will become artefactually trapezoidal, and so will give a false impression of its true shape. Genitalia : gonocoxite Figure 2c shows the basal half of the gonocoxite of most species of the Orientalis Complex to be broader than the apical half, and the ventral root to arise at a shallower angle (45-75 ); it shows too that the dorsal root is almost straight, i.e. it does not curl around the base of the paramere. In the Imicola Complex (Fig. 2g) the gonocoxite is nearly parallelsided for its entire length, and the dorsal and ventral roots are distinctly bowed (almost like a curled forefinger and thumb). Furthermore, the ventral root arises almost at right angles (75-90 ) from the gonocoxite. In all species of the Imicola Complex the shorter dorsal root is rather broad and always curls noticeably around the base of the paramere. Secondary characters Wing pattern Three features deserve mention, as follows: 1. As a general rule, the pale spot found at the tip of the second radial cell in all species of the Orientalis Complex combines with those found at the base of cell m1, and medianally in cells m2 and m4, to form a broadly pale and straight line that bisects the wing longitudinally. In the Imicola Complex, the small, but prominent, pale spot in the base of cell m1 is positioned more proximally and thus a subtle zigzag of pale areas bisects the wing. 2. In the Orientalis Complex, the short, slanting vein that forms the proximal boundary of the second radial cell is always thickened where it meets the costa; this thickening does not occur in the Imicola Complex. 3. Most species of the Orientalis Complex have the anal angle dark; in the Imicola Complex, the anal angle is pale in 10 of the 12 species. Discussion Five points emerge, as follows: 1. This study reveals that the earlier definitions (9, 29) of the Imicola and the Orientalis Complexes are too broad and will have to be amended in the future. This can be done only following the reevaluation of other species complexes within the subgenus Avaritia. Nevertheless, it is now clear that the Imicola and the Orientalis Complexes are restricted to the Old World. The distribution of the latter is more tropical, whilst the former has radiated outside of the equatorial forest block in the adjoining lower rainfall subtropical woodlands, savannah grasslands and semi-desert regions. The added revelation that closely related species (e.g. imicola/nudipalpis; bolitinos/brevitarsis) occur in separate biogeographic faunas does indicate that discrete lineages have developed within Avaritia over millions of years, but that current species complex definitions have been too weak to expose these patterns of radiation and vicariance. 2. Twenty three species of Avaritia are now assigned to either the Imicola or the Orientalis Complex. Only eight of the 19 species previously assigned by Wirth and Hubert to the Orientalis Complex are retained (29); but this number is now expanded to 11 by the inclusion of the three afrotropical species formerly placed in the Trifasciellus subgroup. Of the remaining 11 species of Wirth and Hubert, seven belong to as yet undefined complexes, and four to the Imicola Complex. Of the 12 species comprising the predominantly afrotropical Imicola Complex, three await description, and so only nine species are listed in Table I; the systematics of the Orientalis Complex is also fluid as at least 11 species occur in South-East Asia, three more than the number currently recognised (R. Meiswinkel, personal observation). 3. The subgroup category is being employed at different hierarchical levels. For example the Pseudopallidipennis subgroup was created for a trio of allied species that reside within the Imicola Complex (23), whereas other workers have employed the subgroup category at the higher group or complex level (11). Furthermore, they used also the Imicola group in lieu of Avaritia, and subdivided it into the Grahamii, Trifasciellus and Imicola subgroups (which were not defined) (11). This hierarchical discordance obscures the fact that their first two subgroups are synonymous with the South-East Asian Actoni and Orientalis groups of Wirth and Hubert (29). To align their studies with those being conducted in other faunal regions, the three subgroups of Itoua and Cornet (11) are raised here to the species complex (or Complex ) level. 4. This realignment and synonymy lowers to seven (six species complexes and one subgroup) the number of categories currently comprising the subgenus Avaritia worldwide: these are the Obsoletus, Imicola, Orientalis, Grahamii, Andicola and Pusillus Complexes, and the Pseudopallidipennis subgroup. The existence of a Veterinaria Italiana, 40 (3),

304 further two species complexes, i.e. the Gulbenkiani and Suzukii Complexes, has recently been mooted (24). 5. It is proposed that in future studies the term Complex or species complex should be used in preference to group. The term Complex is not employed here as defined earlier (13) wherein complexes were applied at a higher hierarchical level, i.e. as an additional category between species groups and subgenera. The term species complex is employed here to group closely related terminal taxa (presumably recently evolved) and united phylogenetically in sharing one or more synapomorphic features. However, if one were to apply this definition strictly, then the recently created Pseudopallidipennis subgroup (23), a distinct clade within the Imicola Complex, merits also consideration as a species complex ; the rational splitting of subgenera into species complexes clearly requires further phylogenetic study. Conclusion A future cladistic analysis will probably confirm the monophyly of the Imicola and the Orientalis Complexes within the subgenus Avaritia. However, a complete redefinition of these two complexes, and of the subgenus Avaritia, must await the further evaluation of the distinguishing characters presented here across the entire world fauna. The admission that due to poor preparation of slide-mounted specimens certain characters were not determined (29) to a great extent sums up problems surrounding the systematics of world Culicoides. The capture of insufficient numbers of specimens of both sexes, the lack of attention to detail in regard to their preparation, description and illustration, and the employment of a descriptive format that is too superficial and stylised, is hampering taxonomic progress. As Culicoides transmit a number of orbiviral diseases that threaten the livestock industry pantropically there is a constant need to identify species with precision. A firm taxonomic foundation is also essential to advancing knowledge on other fronts ranging from basic biology to the development of predictive risk maps using satellite imagery. Due to this lack of depth, many taxonomic decisions are today still being taken at the species complex, and not at the species, level (e.g. the Obsoletus Complex). In other instances, the existence of vector complexes is only being revealed gradually (e.g. the Imicola Complex); alternatively, where such complexes have been recognised to exist for some time, their taxonomy has long remained in a state of flux (e.g. the Variipennis Complex). From these examples, it is abundantly clear that taxonomic studies on the genus Culicoides need to be refined. As some vector complexes have eluded the best taxonomic efforts (e.g. the Obsoletus Complex), it would seem the way forward is to now integrate traditional morphological systematics with a modern molecular approach. This integrated approach has recently been applied effectively on the Imicola Complex (16, 26), and so promises to resolve the kinds of difficult taxonomic issues that continue to plague vector systematics. One of the most pressing is the resolution of the taxonomy of the Obsoletus and Pulicaris Complexes following their recent incrimination in the transmission of bluetongue in south-eastern Europe, and in areas hitherto unaffected by this economically devastating disease. At this moment these vectors are assumed to be C. obsoletus sensu stricto, C. scoticus sensu stricto (both subgenus Avaritia) and C. pulicaris sensu stricto, (subgenus Culicoides) but these identifications remain to be unequivocally proven. References 1. Boorman J. (1988). Taxonomic problems in Culicoides of southwest Asia, in particular of the Arabian Peninsula. In Biosystematics of Haematophagous insects (M.W. Service, ed.). Systematics Association, Special Volume No. 37, Clarendon Press, Oxford, Boorman J. (1989). Culicoides (Diptera: Ceratopogonidae) of the Arabian Peninsula with notes on their medical and veterinary importance. Fauna Saudi Arabia, 10, Boorman J. (1991). A review of Culicoides subgenus Avaritia species (Insecta, Diptera: Ceratopogonidae), vectors of viruses of sheep, cattle and horses, with particular reference to Culicoides imicola in Europe and the Mediterranean region. Report prepared for the Overseas Development Administration, The Natural History Museum, London, 54 pp. 4. Borkent A. & Wirth W.W. (1997). World species of biting midges (Diptera: Ceratopogonidae). Bull. Am. Natl Hist. Mus., 23, Campbell J.A. & Pelham-Clinton E.C. (1960). A taxonomic review of the British species of Culicoides Latreille (Diptera, Ceratopogonidae). Proc. R. Soc. Edin., Ser. B (Biol.), 67, Dyce A.L. (1983). Reviews of published records of Culicoides species in subgenus Avaritia (Diptera: Ceratopogonidae) from the New Guinea subregion. Int. J. Entomol., 25, Dyce A.L. & Wirth W.W. (1983). Reappraisal of some Indian Culicoides species in the subgenus Avaritia (Diptera: Ceratopogonidae). Int. J. Entomol., 25, Veterinaria Italiana, 40 (3), 2004

305 8. Fox I. (1955). New advances in the study of the genus Culicoides [original in Spanish]. Bol. Lab. Clin. Luis Razetti, 43-44, Glick J.I. (1990). Culicoides biting midges (Diptera: Ceratopogonidae) of Kenya. J. Med. Entomol., 27, Glukhova V.M. (1989). Bloodsucking midges of the genera Culicoides and Forcipomyia (Ceratopogonidae) [original in Russian]. Fauna USSR, 139, 3 (5a), 408 pp. 11. Itoua A. & Cornet M. (1986). Les Ceratopogonidae (Diptera) du Mayombe congolais 3. Revue taxonomique des espèces du genre Culicoides Latreille, Cah. ORSTOM., Sér. Entomol. Méd. Parasitol., 24, Jamnback H. & Wirth W.W. (1963). The species of Culicoides related to obsoletus in eastern North America. Ann. Entomol. Soc. Am., 56, Khalaf K.T. (1954). The speciation of the genus Culicoides (Diptera: Heleidae). Ann. Entomol. Soc. Am., 47, Khamala C.P.M. & Kettle D.S. (1971). The Culicoides Latreille (Diptera: Ceratopogonidae) of East Africa. Trans. R. Entomol. Soc., London, 123, Kitaoka S. (1985). Japanese Culicoides (Diptera: Ceratopogonidae) and keys for the species. I. Bull. Nat. Inst. Anim. Hlth, 87, Linton Y.-M., Mordue A.J., Cruickshank R.H., Meiswinkel R., Mellor P.S. & Dallas J.F. (2002). Phylogenetic analysis of the mitochondrial cytochrome oxidase subunit I gene of five species of the Culicoides imicola species complex. Med. Vet. Entomol., 16, Mayr E., Linsely E.G. & Usinger R.L. (1953). Methods and principals of systematic zoology. McGraw-Hill, New York. 18. Meiswinkel R. (1987). Afrotropical Culicoides: a redescription of C. (Avaritia) kanagai Khamala & Kettle, 1971, reared from elephant dung in the Kruger National Park, South Africa. Onderstepoort J. Vet. Res., 54, Meiswinkel R. (1989). Afrotropical Culicoides: a redescription of C. (Avaritia) imicola Kieffer, 1913 (Diptera: Ceratopogonidae) with description of the closely allied C. (A.) bolitinos sp. nov. reared from the dung of the African buffalo, blue wildebeest and cattle in South Africa. Onderstepoort J. Vet. Res., 56, Meiswinkel R. (1991). Afrotropical Culicoides: C. (Avaritia) miombo sp. nov. a widespread species closely allied to C. (A.) imicola Kieffer, 1913 (Diptera: Ceratopogonidae). Onderstepoort J. Vet. Res., 58, Meiswinkel R. (1992). Afrotropical Culicoides: C. (Avaritia) loxodontis sp. nov. a member of the Imicola group (Diptera: Ceratopogonidae) associated with the African elephant in the Kruger National Park, South Africa. Onderstepoort J. Vet. Res., 59, Meiswinkel R. (1995). Afrotropical Culicoides: biosystematics of the Imicola group, subgenus Avaritia (Diptera: Ceratopogonidae), with special reference to the epidemiology of African horse sickness. MSc (Agric.) dissertation, University of Pretoria, South Africa, iv +359 pp (unpublished). 23. Meiswinkel R. & Linton Y.-M. (2003). Afrotropical Culicoides (Diptera: Ceratopogonidae): Morphological and molecular description of a novel fruit-inhabiting member of the Imicola Complex, with a redescription of its sister species C. (Avaritia) pseudopallidipennis Clastrier. Cimbebasia, 19, Meiswinkel R., Gomulski L.M., Delécolle J.-C., Goffredo M. & Gasperi G. (2004). The taxonomy of Culicoides vector complexes: unfinished business. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Mirzaeva A.G. (1984). A review of biting midges of the subgenus Avaritia Fox (Diptera: Ceratopogonidae, genus Culicoides Latr.) from Siberia [original in Russian, English summary]. Entomol. Oboz., 63, [English translation in Entomol. Rev., 63 (3), 28-43]. 26. Sebastiani F., Meiswinkel R., Gomulski L.M., Guglielmino C.R., Mellor P.S., Malacrida A.R. & Gasperi G. (2001). Molecular differentiation of the Old World Culicoides imicola species complex (Diptera: Ceratopogonidae), inferred using random amplified polymorphic DNA markers. Molec. Ecol., 10, Wirth W.W., Dyce A.L. & Peterson B.V. (1985). An atlas of wing photographs, with a summary of the numerical characters of the Nearctic species of Culicoides (Diptera: Ceratopogonidae). Contrib. Am. Entomol. Inst., 22, Wirth W.W., Dyce A.L. & Spinelli G.R. (1988). An atlas of wing photographs, with a summary of the numerical characters of the Nearctic species of Culicoides (Diptera: Ceratopogonidae). Contrib. Am. Entomol. Inst., 25, Wirth W.W. & Hubert A.A. (1989). The Culicoides of South-East Asia (Diptera: Ceratopogonidae). Mem. Am. Entomol. Inst., 44, Veterinaria Italiana, 40 (3),

306 Vet. Ital., 40 (3), Molecular taxonomy and population structure of a Culicoides midge vector D.V. Nolan, J.F. Dallas & A.J. Mordue (Luntz) School of Biological Sciences, University of Aberdeen, United Kingdom Correspondence: D.V. Nolan, Zoology Building, Tillydrone Avenue, Aberdeen AB24 2TZ, Scotland, United Kingdom Summary The biting midge Culicoides imicola Kieffer (Diptera: Ceratopogonidae) is the major Old World vector of the arboviruses that cause African horse sickness (AHS) and bluetongue (BT). Recently, the incidence and geographical scales of AHS and BT outbreaks in the Mediterranean Basin have increased, with serotype distribution in the BT outbreaks being geographically structured. The authors review molecular approaches for assessing the contribution of cryptic species and population subdivision in C. imicola to BT serotype structure in this region. No evidence was found for cryptic species. In contrast, evidence was found for marked matrilineal subdivision between the eastern and western parts of the Mediterranean Basin. This pattern is comparable to the geographic structure of BT serotypes, suggesting that subdivision in the insect vector potentially constrains serotype spread. The authors are presently testing this hypothesis. Keywords African horse sickness Bluetongue Culicoides imicola Cytochrome oxidase subunit I Europe Mediterranean Basin Mitochondrial DNA Molecular population genetics. Introduction The biting midge Culicoides imicola Kieffer, 1913 (Diptera: Ceratopogonidae) is the most important Old World vector of the arboviruses that cause African horse sickness (AHS) and bluetongue (BT) (16). AHS is endemic in sub-saharan Africa and BT is endemic in tropical latitudes worldwide. Before 1980, outbreaks of both diseases in the Mediterranean Basin were localised and sporadic. This pattern has changed recently, with AHS outbreaks in Iberia since the mid-1980s lasting longer (14, 15) and the largest epidemic of BT ever recorded in the Mediterranean Basin between 1998 and 2001 (1). The most likely causes of these epidemiological changes are increases in the range and abundance of C. imicola (2, 15, 20, 21), and these changes might be responses to climate change (27). Although primarily an Afro-Asiatic species, C. imicola is reported widely throughout the Mediterranean Basin: from Algeria, Cyprus, Egypt, Israel, the islands of Lesbos and Rhodes, Morocco, Portugal, Spain, Tunisia and Turkey (16), and most recently from Italy (7) and the Balearic Islands (18). Moreover, risk maps for predicting the distribution of C. imicola from climatic variables (10, 28) and vegetation indices (4) suggest that C. imicola is even more widespread in this region, and that its apparent absence potentially reflects a lack of sampling. Taxonomic status in the North American BT virus vector Culicoides variipennis (Coquillett) has crucial implications for epidemiology and vector control. According to analyses of male morphology and genetic distance, Culicoides variipennis consists of three subspecies, only one of which, C. v. sonorensis Wirth and Jones, is an efficient BT virus vector (24, 25). Subdivision of insect vector populations has important implications for control measures and predicting disease spread (26). Despite this, the phylogenetic status and genetic structure of C. imicola in Mediterranean countries were unknown prior to recent work by our group at the University of Aberdeen. 352 Veterinaria Italiana, 40 (3), 2004

307 Identification of insect species: a DNA barcoding system? It is now recognised that coherence in insect systematics will ultimately depend on having a large database of comparable DNA sequence data. Four genes presently stand out as well-surveyed and informative across a wide taxonomic range: the cytochrome oxidase subunit I (COI) and 16S genes of mitochondrial DNA (mtdna), and the nuclear 18S and EF-1α genes (9). Indeed, a DNA-based identification system based on COI can potentially provide a consistent means to resolve species diversity using DNA barcoding (12). When fully developed, this approach is expected to provide a reliable, cost-effective and accessible solution to current problems of species identification. Molecular taxonomy of the Imicola Complex in South Africa Species identification in Culicoides is often problematic, owing to their small size and environment-dependent morphology, the prevalence of taxa with very similar morphologies and the expression of diagnostic traits in adult males only. Analyses of morphological characters of adult insects suggest that the Imicola complex in South Africa comprises at least ten species. Eight species are confirmed (C. imicola sensu stricto; C. brevitarsis Kieffer; C. pseudopallidipennis Clastrier; C. nudipalpis Delfinado; C. bolitinos Meiswinkel; C. miombo Meiswinkel; C. loxodontis Meiswinkel, C. tuttifrutti Meiswinkel, Cornet and Dyce,) and two are provisional (C. kwagga Meiswinkel, and C. sp. #103 Meiswinkel). These species have distinct larval habitats, including rich damp clay soil (C. imicola), marshy areas (C. miombo), elephant dung (C. loxodontis), horse, rhinoceros and zebra dung (C. kwagga), buffalo and cattle dung (C. bolitinos), rotting fruits (C. tuttifrutti) and banana stumps (C. pseudopallidipennis). According to analysis of random amplification of polymorphic DNA (RAPD) markers, these species are genetically different but not phenetically distinct (22). The evolution of vector competence and transitions in larval habitat in the Imicola complex are best elucidated using phylogenetic approaches. The phylogenetic status of four confirmed species (C. imicola s.s., C. bolitinos, C. loxodontis and C. tuttifrutti) and one provisional (C. kwagga) species of this complex was assessed, using a phylogenetic analysis of COI (13). The Culicoides spp. sequences were aligned against the corresponding COI sequence of Anopheles gambiae, and base differences among the sequences were identified relative to the IMI 16 sequence of C. imicola. All the base substitutions within species were synonymous, and several base differences between species were evident (Fig. 1). A neighbour-joining tree (Fig. 2a) yielded consistently high bootstrap support for each separate species but consistently low support for clades containing more than one species. All parsimony analyses yielded the same group of four equally parsimonious trees. One of these trees had the same topology as the NJ tree, which also had the highest likelihood. The other trees differed in the placement of C. bolitinos, C. loxodontis or C. tuttifrutti. A maximum likelihood tree (Fig. 2b) yielded high quartet puzzling support for all species except C. imicola as distinct groups, and for the clade containing C. bolitinos, C. loxodontis and C. tuttifrutti. Thus, the usefulness of COI was demonstrated for species identification in the Imicola Complex, and showed that five of its members are phylogenetically distinct. Species identification and subdivision in Culicoides imicola The species status and matrilineal subdivision in C. imicola in the Mediterranean Basin has been characterised using COI sequences (11). Partial COI sequences (472 bp without primers) were obtained from 19 sites in Portugal, Rhodes and Israel (Table I). Phylogenetic trees were constructed from the COI sequences of C. imicola in the Mediterranean Basin and South Africa and of four other species of the Imicola Complex from southern Africa. The maximum likelihood tree (Fig. 3) contained five separate clades. Each clade represented a single species in the Imicola Complex, and each was supported by high (73-100) bootstrap values. All the C. imicola sequences grouped in the same clade. Phylogenetic substructure within the C. imicola clade was absent, according to the maximum likelihood bootstrap values of less than 60. Thus, midges of the morphospecies C. imicola from Portugal, Rhodes, Israel and South Africa belong to the same phylogenetic clade of COI, and are phylogenetically distinct from four other species of the Imicola Complex. These results are consistent with all C. imicola from the study areas being potentially competent AHS and BT vectors. The samples from Portugal had zero values of haplotype and nucleotide diversity, being monomorphic for haplotype IMICOI 01. The samples from Israel had a haplotype diversity of (SD 0.112) and a nucleotide diversity of (SD ). One haplotype (IMICOI 03) was shared between Israel and Rhodes, and the other Veterinaria Italiana, 40 (3),

308 eight haplotypes were unique to each country. According to an analysis of molecular variance, genetic differentiation between the samples from Portugal and Israel was highly significant (permutation test, P<0.001). The proportion of molecular variance within countries was 0.20 and between countries The corresponding value of Φ ST was 0.80, and this value was significantly different from zero (permutation test, P<0.001). Thus, the level of matrilineal subdivision in C. imicola between the eastern and western ends of the Mediterranean Basin is marked and is comparable to the highest levels known for Diptera. Whether this result reflects genetic drift associated with dependence on fragmented habitat, or a recent 80 IMI16 ATTAATATTAGGGGCTCCTGATATAGCTTTTCCTCGAATAAATAATATAAGTTTTTGAATATTACCGCCATCTATTACTC IMI13... BOL14 T...A...C...A...T..G..A...AT KWA2...A...C...A...C.T..T..T...T LOX1 T...C..A...C...T.G...G...T..T..A...AT LOX10 T...A...C...T.G...G...T..T..A...AT TUT1 T...A...A...C..T..A...T GAM T...A..A...A...GC.T..A..T..AT.A..A. 160 IMI16 TTCTTTTATTAAGTAGATTAGTAGAAAATGGAGCAGGAACAGGATGAACTGTTTATCCTCCATTATCAGCAAATGTTTCT IMI13... BOL14.A...A...C..G...G..A...C...T...TT.T... KWA2..AT.AA...A...G...G...T...AT... LOX1...A.T..A...TC.G...A...A..T...TTT...A... LOX10...A.T..A...TC...A...A..T...TTT...A... TUT1..A...G...A...C...T...T...A... GAM...T.AA.T.CT...TA...C..G..T...TC...TT.TGGAA..G IMI16 CATGCTGGAGCTTCAGTTGATTTAGCTATTTTCTCTTTACATTTAGCTGGGATTAGTTCAATTTTAGGTGCTGTAAATTT IMI13... BOL14...T...T...C...A...G... KWA2...G...T...T...A...T...G..G... LOX1...A..T..A...A...A.A...T..A...A..T...A... LOX10...A..T..A...A...A.A...T..A...A..T...A... TUT1...T...T...A...T...A..A...A..T...A... GAM...A...A...T...C.T...A..A...TC...T...A..A IMI16 TATTACAACAATTATTAATATACGTCCTGAAGGAATAACTATGGATCGAATACCTTTATTTGTTTGATCAGTATTTATTA IMI13... BOL14...T..T...A..A.T...T.A...G..AC.A. KWA2...T...A...AT...T.A...T.T...T...AG.G. LOX1...T..T...A..CATT...T..T.A...A...AG.A. LOX10...T..T...A..CATT...T..T.A...A...AG.A. TUT1...T..T...A..AAT...T..T.A...G...A... GAM...G.A...GT..CC...T..AT.A...A...G...G IMI16 CAGCTATTTTATTACTTTTATCTTTACCTGTGTTAGCAGGGGCTATTACTATATTATTAACAGATCGGAATATTAATACT IMI13... BOL14...GT.A...A...A..T...A...G..T...A..C... KWA2...T.A...A..G...A...T..A...A... LOX1...A...C.T...T.A...AC...A...T..A...A...T...A...A... LOX10...A...C.T...T.A...A...A...T..A...A...T...A...A... TUT1...A...T.G...A...T...T..T...A...T...A... GAM...AG.A...T.A...A...A..A...A...T...A...T.A...A IMI16 TCCTTTTTTGACCCAGCAGGAGGAGGAGATCCTATTTTATACCAACATTTATTTTGATTTTTTGGGCATCCA IMI13...C... BOL14..A...T...G...A...T..G...T... KWA2...T...G...G..C..A...T...T...T LOX1...T...T..T...T...A...T...T...T LOX10..T...T..T...T...A...T...T...T TUT1...A...T..T...G..G...A...T...T... GAM...T..C...T...T...A...T...C...C...T Figure 1 Alignment of nucleotide sequences (5 3 ) of 472 bp (without primers) of the mitochondrial DNA cytochrome oxidase subunit I gene of Culicoides imicola (IMI), C. bolitinos (BOL), C. kwagga (KWA), C. loxodontis (LOX), C. tuttifrutti (TUT) from South Africa and Anopheles gambiae (GAM) Identical sequences are omitted, dots indicate identical bases, and species-diagnostic bases are shown in bold Redrawn from Linton et al. (13) 354 Veterinaria Italiana, 40 (3), 2004

309 a) Neighbour-joining tree constructed from Tamura-Nei distances with γ-distributed rates Numbers on the nodes represent NJ (above) and parsimony (below) bootstrap proportions ( replicates each) b) Maximum likelihood tree constructed using a GTR substitution model with γ-distributed rates Numbers on the nodes represent quartet puzzling values ( replicates). The full species names corresponding to the abbreviations are given in Figure 1 Figure 2 Phylogenetic relationships among partial nucleotide sequences of the mitochondrial DNA cytochrome oxidase subunit I gene of five members of the Imicola Complex Redrawn from Linton et al. (13) Veterinaria Italiana, 40 (3),

310 Table I Locations and years of sampling of Culicoides imicola, number of sequences and COI haplotypes obtained at each site Country Site Latitude and longitude Year Site code No. IMICOI haplotypes Portugal Alcaíns N 7 26 W 2001 PO Avis N 7 54 W 2001 PO Castelo de Vide N 7 28 W 2001 PO Mourão N 7 18 W 2001 PO Pedrogão N 8 19 W 2001 PO Barca D Alva N 6 57 W 2001 PO Greece Dimilia, Rhodes N E 2001 GR1 5 02, 03 Israel Beit Dagan N E 1996 IBET 3 04, 08, IS1 2 04, 05 Newe Ya ar N E 1996 INY 4 04 Karmiyya N E 2001 IS Kefar Malal N E 2001 IS Kefar Daniyyel N E 2001 IS Kefar Silver N E 2001 IS Mishmar Hasharon N E 2001 IS6 2 04, 05 Regba N E 2001 IS Ma ale Hahamisha N E 2001 IS Merom Golan N E 2001 IS Nahalal N E 2001 IS Yotvata N E 2001 IS , 07 South Africa* Onderstepoort S E 1996 IMI 4 10, 11 * Sequences were reported in Linton et al. (13) Redrawn from Dallas et al. (11) colonisation of the Mediterranean Basin by two or more genetically distinct sources, is presently unclear. Seasonal airstreams are thought to mediate the movement of arbovirus-infected C. brevitarsis Kieffer in Australia (5, 19) and the immigration of BT virus (BTV)-infected C. imicola into Portugal (23) and Israel (6). Our result suggests that matrilineal gene flow between the latter areas in the past has been limited. The matrilineal subdivision in C. imicola is comparable to the geographic structure of BTV serotypes in outbreaks around the Mediterranean during BTV serotypes 4, 9 and 16 spread westwards from Israel and Turkey to Greece and Bulgaria, while BTV-2 spread northwards from Tunisia through Sardinia, Corsica, and Sicily to mainland Italy (17). Assuming that the C. imicola populations in these areas are equally competent BTV vectors, we hypothesise that the matrilineal subdivision reflects true population subdivision in C. imicola, and that this structure constrained the recent BT outbreaks in these areas to different geographical sources and routes of spread. This hypothesis has important implications for epidemiological monitoring and vaccination programmes for BT, and we are presently testing it using a phylogeographic analysis of a comprehensive sampling of C. imicola from this region. Future work Outbreaks of BT have occurred in areas where C. imicola, although looked for, has never been recorded (as far north as 44 N in Serbia and Bosnia- Herzegovina, north-west Greece, Bulgaria and west European Turkey) (3, 17). In these areas, novel Culicoides vectors, possibly members of the widespread Obsoletus and Pulicaris Complexes, are potential vectors (17). Moreover, in 2002, in Sicily, where outbreaks of BT occurred in the absence of C. imicola, bluetongue viral RNA was detected in wild-caught, non-blood-engorged, parous C. pulicaris, further suggesting that C. pulicaris can be a fully competent BT vector (8). Therefore, we plan to extend the molecular approaches described above to these novel vectors. Analysis of specimens from mainland Italy, and of larger samples from the areas in the above studies, is ongoing. The utility of COI to detect genetic subdivision within countries such as Portugal is limited, however. We are therefore analysing mtdna and nuclear genes more polymorphic than COI for a comprehensive characterisation of genetic subdivision in C. imicola in the Mediterranean Basin. The results of these molecular studies will be combined with GIS data and used to construct a risk map for bluetongue in Europe. This map will be 356 Veterinaria Italiana, 40 (3), 2004

311 Figure 3 Phylogenetic relationships among COI haplotypes of C. imicola from Portugal, Rhodes, Israel and South Africa, and of four other species of the Imicola Complex Maximum likelihood tree. Numbers are bootstrap percentages. Values above the nodes are for the GTR+I+γ model and values below the nodes are for the GTR+γ model. The details corresponding to the code of each C. imicola sample are given in Table I The species names corresponding to the abbreviations used for the other species are given in Figure 1 Redrawn from Dallas et al. (13) Veterinaria Italiana, 40 (3),

312 used to predict outbreaks and inform agricultural and veterinary practice in the participating countries and the EU as a whole. The combination of data on subdivision in vector species and on BT serotype distribution is expected to provide critical information in developing risk assessment relating to emerging diseases and will become more important as vector ecologists realise the full potential of population genetics and its application to medical veterinary entomology. Acknowledgements Part of this work was financed by the European Commission (Contract No. QLK2-CT ). References 1. Baylis M. (2002). The re-emergence of bluetongue. Vet. J., 164, Baylis M., El-Hasnaoui H., Bouayoune H., Touti J. & Mellor P.S. (1997). The spatial and seasonal distribution of African horse sickness and its potential Culicoides vectors in Morocco. Med. Vet. Entomol., 11, Baylis M. & Mellor P.S. (2001). Bluetongue around the Mediterranean in Vet. Rec., 149, Baylis M., Mellor P.S., Wittmann E.J. & Rogers D.J. (2001). Prediction of areas around the Mediterranean at risk of bluetongue by modelling the distribution of its vector using satellite imaging. Vet. Rec., 149, Bishop A.L., Barchia I.M. & Spohr L.J. (2000). Models for the dispersal in Australia of the arbovirus vector, Culicoides brevitarsis Kieffer (Diptera: Ceratopogonidae). Prev. Vet. Med., 47, Braverman Y. & Chechik F. (1996). Air streams and the introduction of animal diseases borne on Culicoides (Diptera, Ceratopogonidae) into Israel. Rev. Sci. Tech. Off. Int. Épiz., 15, Calistri P., Goffredo M., Caporale V. & Meiswinkel R. (2003). The distribution of Culicoides imicola in Italy: Application and evaluation of current Mediterranean models based on climate. J. Vet. Med. B, 50, Caracappa S., Torina A., Guercio A., Vitale F., Calabro A., Purpari G., Ferrantelli V., Vitale M. & Mellor P.S. (2003). Identification of a novel bluetongue virus vector species of Culicoides in Sicily. Vet. Rec., 153, Caterino M.S., Cho S. & Sperling F.A.H. (2000). The current state of insect molecular systematics: a thriving Tower of Babel. Ann. Rev. Entomol., 45, Conte A., Giovannini A., Savini L., Goffredo M., Calistri P. & Meiswinkel R. (2003). The effect of climate on the presence of Culicoides imicola in Italy. J. Vet. Med. B., 50, Dallas J.F., Cruickshank R.H., Linton Y.-M., Nolan D.V., Patakakis M., Braverman Y., Capela R., Pena I., Meiswinkel R., Ortega M.D., Baylis M., Mellor P.S. & Mordue (Luntz) A.J. (2003). Phylogenetic status and matrilineal structure of the biting midge, Culicoides imicola, in Portugal, Rhodes and Israel. Med. Vet. Entomol., 17 (4), Hebert P.D.N., Cywinska A., Ball S.L. & DeWaard J.R. (2003). Biological identifications through DNA barcodes. P. Roy. Soc. Lond. B-Bio., 270, Linton Y.M., Mordue A.J., Cruickshank R.H., Meiswinkel R., Mellor P.S. & Dallas J.F. (2002). Phylogenetic analysis of the mitochondrial cytochrome oxidase subunit I gene of five species of the Culicoides imicola species complex. Med. Vet. Entomol., 16, Mellor P.S. (1994). Epizootiology and vectors of African horse sickness virus. Comp. Immunol. Microbiol. Infect. Dis., 17, Mellor P.S & Boorman J. (1995). The transmission and geographical spread of African horse sickness and bluetongue viruses. Ann. Trop. Med. Parasitol., 89, Mellor P.S., Boorman J. & Baylis M. (2000). Culicoides biting midges: Their role as arbovirus vectors. Ann. Rev. Entomol., 45, Mellor P.S. & Wittmann E.J. (2002). Bluetongue virus in the Mediterranean Basin Vet. J., 164, Miranda M.A., Borras D., Rincon C. & Alemany A. (2003). Presence in the Balearic islands (Spain) of the midges Culicoides imicola and Culicoides obsoletus group. Med. Vet. Entomol., 17, Murray M.D. (1995). Influences of vector biology on transmission of arboviruses and outbreaks of disease the Culicoides brevitarsis model. Vet. Microbiol., 46, Rawlings P. & Mellor P.S. (1994). African horse sickness and the overwintering of Culicoides spp in the Iberian Peninsula. Rev. Sci. Tech. Off. Int. Épiz., 13, Rawlings P., Capela R., Pro M.J., Ortega M.D., Pena I., Rubio C., Gasca A. & Mellor P.S. (1998). The relationship between climate and the distribution of Culicoides imicola in Iberia. Arch. Virol., Sebastiani F., Meiswinkel R., Gomulski L.M., Guglielmino C.R., Mellor P.S., Malacrida A.R. & Gasperi G. (2001). Molecular differentiation of the Old World Culicoides imicola species complex (Diptera, Ceratopogonidae), inferred using random amplified polymorphic DNA markers. Molec. Ecol., 10, Sellers R.F., Pedgley D.E. & Tucker M.R. (1979). Possible windborne spread of bluetongue to Portugal, June-July J. Hyg., Camb., 81, Tabachnick W.J. (1992). Genetic differentiation among populations of Culicoides variipennis (Diptera, 358 Veterinaria Italiana, 40 (3), 2004

313 Ceratopogonidae), the North American vector of bluetongue virus. Ann. Entomol. Soc. Am., 85, Tabachnick W.J. (1996). Culicoides variipennis and bluetongue virus epidemiology in the United States. Ann. Rev. Entomol., 41, Tabachnick W.J. & Black W.C. (1995). Making a case for molecular population genetic studies of arthropod vectors. Parasitol. Today, 11, Wittmann E.J. & Baylis M. (2000). Climate change: effects on Culicoides-transmitted viruses and implications for the UK. Vet. J., 160, Wittmann E.J, Mellor P.S. & Baylis M. (2001). Using climate data to map the potential distribution of Culicoides imicola (Diptera: Ceratopogonidae) in Europe. Rev. Sci. Tech. Off. Int. Épiz., 20, Veterinaria Italiana, 40 (3),

314 Vet. Ital., 40 (3), Use of a Montecarlo simulation model for the re-planning of bluetongue surveillance in Italy P. Calistri, A. Giovannini, A. Conte & V. Caporale Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy Summary Since August 2000, Italy has lost almost sheep to the largest incursion of bluetongue (BT) yet to affect Europe. The national BT surveillance system includes serological and entomological programmes. The main objective of the serological programme is the early detection of BT virus (BTV) circulation by periodical testing of more than sentinel cattle distributed across Italy. The sentinel network has been in force since October 2001 and since its inception has provided accurate and timely detection of viral circulation. However, the repeated testing of such a large number of animals required significant effort from the Veterinary Services and was costly for farmers. Consequently, a Montecarlo simulation model was developed to simulate different sentinel system scenarios and the results from each were compared. The model was validated using data derived from the serological surveillance activities in Sardinia from October 2001 to December Keywords Bluetongue Italy Montecarlo simulation model Sardinia Sentinel system Simulation Surveillance. Introduction Since August 2000, Italy has lost almost sheep to the largest incursion of bluetongue (BT) yet to affect Europe (1, 2). A vaccination programme was enforced in 2002, in conjunction with strict animal movement controls to reduce the spread of infection (3, 4). A surveillance system was implemented that required the periodical testing of more than sentinel animals and included a number of other surveillance activities (5). The sentinel network has been in place since October 2001 and has proved reliable for detecting viral circulation. However, the periodical testing of more than sentinel animals across Italy in 2002 required significant effort from the Veterinary Services and was costly for farmers. A strong demand to reduce sentinel animal testing came from representatives of the local Veterinary Services and farmers. To evaluate and refine the Italian serological surveillance system, a model was developed simulating different sentinel system scenarios and the results that could be expected from each. Materials and methods A simulation model of a serological surveillance system based on periodical testing of sentinel animals was implemented, with the aim of simulating the following: a) expected number of animals to seroconvert and to be detected upon subsequent testing b) expected number of herds in which at least one animal seroconverts to BTV and for the animal to be detected upon subsequent testing. The model neither takes into account the sensitivity and/or specificity of the diagnostic tests used, nor the effects of multiple testing. Assumptions The model is applicable to any geographic area, but the probability of the final event (i.e. seroconversion) must be constant and continuous geographically and temporally. In other words, the geographic unit of reference and the time period considered in the model must be selected with care, so as to apply the 360 Veterinaria Italiana, 40 (3), 2004

315 proposed model on a homogeneous epidemiological population correctly. Description of the model The output of the model is the number of sentinel animals that seroconvert to BT in a specific geographic area, during a specified time unit (week, fortnight, month, etc.). The input variables are as follows: incidence of sentinel herds that seroconvert (Pi) per time unit and in a specific geographic area (a sentinel herd was considered to have seroconverted when at least one of its sentinel animals seroconvert) number of sentinel herds tested (N) per time unit and in a specific geographic area number of sentinel animals tested in each herd (n) per time unit and in a specific geographic area incidence of animals that seroconvert in the same sentinel herd (Ph) per time unit and in a specific geographic area. The model includes the following calculations: expected number of herds that seroconvert expected number of sentinel animals that seroconvert. The expected number of sentinel herds that seroconvert, per time unit and in a specific geographic area, is calculated according to a Poisson probability distribution, as follows: ( λ t) X A = Poisson * Where: X A = expected number of sentinel herds that seroconvert, in a given time period and in a specific geographic area t = time period considered λ = mean number of herds that seroconvert per time unit and in a specific geographic area. This value is calculated according to the following formula: λ = Binomial, ( N Pi) Where: N = number of tested sentinel herds per time unit in a specific geographic area Pi = incidence of sentinel herds that have seroconverted per time unit and in a specific geographic area. The expected number of sentinel animals that will seroconvert in each herd that has seroconverted (as calculated in Step 1 of the model) is calculated as follows: C = Binomial n, Ph Where: ( ) n = number of sentinel animals tested in each herd Ph = incidence of animals that seroconvert in the same sentinel herd. Therefore, total number of sentinel animals that seroconvert is calculated by adding the expected number of animals that seroconvert in each herd (C): Expected number of animals that will seroconvert per time unit and in a specific geographic area n = C n i= 1. Validation of the model The model was validated using data stored in the bluetongue national database and that had been collected as part of the serological surveillance plan implemented in Sardinia from October 2001 to December Monthly data, from the four Provinces in Sardinia (Cagliari, Nuoro, Oristano and Sassari) were analysed separately. For the validation of the model, Pi was calculated based on the following: the incidence of diseased sheep and goat flocks for each province and each month, taking into account also the number of vaccinated flocks that were not susceptible to the disease the incidence of infection among bovine herds for each province and each month, taking into account the number of positive sentinel herds. The final estimation was calculated assuming that the incidence of positive sentinel herds was equal to the real incidence of infection in the overall susceptible bovine herds. Furthermore, in the calculation of the C value, n was calculated as the mean number of sentinel animals tested from each herd during each month, and Ph was considered as the mean value of the incidence of animals that seroconverted in the same herd each month. The model was implemented using Palisade software (6). The output of the model (1 000 iterations with Latin hypercube sampling) was compared with the real observed number of animals that seroconverted in each province each month. Veterinaria Italiana, 40 (3),

316 Using the model to re-plan the sentinel system Two scenarios were considered. In the first scenario, the number of sentinel herds tested in the population varied. In the second, the number of sentinel animals tested in each herd was varied, leaving unchanged the number of sentinel herds. As a first step, the expected number of herds that would seroconvert was simulated (1 000 iterations with a Latin hypercube sampling) for each of the different scenarios and the number of herds tested (10, 20, 30, 100 herds tested). For each scenario, the following variables were established: Pi = incidence of infected herds = 5% Ph = incidence of infection within the infected sentinel herd = 25%. The chosen value of 25% for the incidence of infection within the herd was inferred from the observed values (Table I). Secondly, the expected number of animals that would seroconvert was simulated (1 000 iterations with Latin hypercube sampling) for each of the different scenarios and number of animals tested for each sentinel herd (4, 5, 6, 12 animals tested). For each scenario, an incidence of 5% of infected herds, 25% incidence of infection within herds and 100 herds tested were considered. Results The results of model validation indicate that the model predicted the number of animals that would seroconvert correctly. The comparison between the observed number of animals that seroconverted in each Sardinian Province from October 2001 to December 2002 and the output of the model outputs is shown in Figures 1, 2, 3 and 4. To verify the possibility of refining the sentinel system, different scenarios of herds and animals tested were evaluated. The model outcomes indicate that the reliability of the sentinel system is highly sensitive to the variation in the number of herds tested. The probability of detecting at least one animal that has seroconverted decreases greatly when the No. of animals that seroconverted / / /2 2002/4 2002/6 2002/8 2002/ /12 Date 95% 5% 50% Observed Figure 1 Comparison between the number of animals that seroconverted in the Cagliari Province and the 5th, 50th and 95th percentile of the distribution of the values predicted by the model, October 2001-December 2002 No. of animals that seroconverted / / /2 2002/4 2002/6 2002/8 2002/ /12 Date 95% 5% 50% Observed Figure 2 Comparison between the number of animals that seroconverted in the Nuoro Province and the 5th, 50th and 95th percentile of the distribution of the values predicted by the model, October 2001-December 2002 number of herds tested is lowered (Fig. 5). Varying the number of animals tested in each sentinel herd has little effect on the probability of detecting at least one animal that has seroconverted. This means that if either 4 or 12 sentinel animals were tested in each herd, the sentinel system would maintain the same Table I The mean monthly incidence of infection with bluetongue virus serotypes 2 and 9 in infected sentinel herds in Italy Serotype Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Mean (minimum-maximum) BTV % 21.4% 17.9% 25.4% 27.3% 25.5% 27.0% 30.9% 24.8% 23.1% 26.5% 27.7% 25.0% (17.9%-30.9%) BTV % 20.1% 20.9% 26.5% 35.2% 33.6% 28.9% 26.2% 18.1% 19.8% 22.4% 22.7% 25.1% (18.1%-35.2%) 362 Veterinaria Italiana, 40 (3), 2004

317 No. of animals that seroconverted / /1 2002/3 2002/5 Date 2002/7 2002/9 2002/11 95% 5% 50% Observed Figure 3 Comparison between the number of animals that seroconverted in the Oristano Province and the 5th, 50th and 95th percentile of the distribution of the values predicted by the model, October 2001-December 2002 ability to detect infection if all other variables remain unchanged (Fig. 6). For example, the probability of detecting at least one animal that has seroconverted is if four animals are tested in each herd and when twelve animals are tested. Discussion The current Italian sentinel network is probably the most intensive serological BT surveillance system in No. of animals that seroconverted / / /2 2002/4 2002/6 2002/8 2002/ /12 Date 95% 5% 50% Observed Figure 4 Comparison between the number of animals that seroconverted in Sassari Province and the 5th, 50th and 95th percentile of the distribution of the values predicted by the model, October 2001-December 2002 Europe. This network has demonstrated its ability to detect virus both accurately and timeously. However, the repeated testing of a large number of sentinel animals, as prescribed in the surveillance plan, required significant efforts from the Veterinary Services and a heavy commitment from farmers. Therefore, a simulation model was developed to evaluate different sentinel system scenarios in order to develop a new streamlined serological surveillance system. Probability >= x No. of animals that seroconverted Figure 5 Model outcome: inverse cumulative distribution of probability of having one or more animals to seroconvert Consideration is made of different scenarios of tested herds (10, 20, ) and the following variables: 5% incidence of infected herds 25% of incidence of infection within herds 12 sentinel animals tested for each herd Veterinaria Italiana, 40 (3),

318 Probability >= x No. of animals that seroconverted Figure 6 Model outcome: inverse cumulative distribution of probability of having one or more animals that seroconvert Consideration is made of different scenarios of tested animal for each herd (4, 5, 6 12) and the following variables: 5% incidence of infected herds 25% incidence of infection within herds 100 sentinel herds tested The simulation model that was developed indicates a significant reduction in the effectiveness of the system to occur when the number of sentinel herds tested is decreased. However, the variation in the number of animals tested within each sentinel herd does not significantly reduce the accuracy of the system. The outcomes of the model are consistent with the expected epidemiological patterns in a vector-borne infection like bluetongue. Consequently, the most effective system for detecting infection would be to use sentinel herds dispersed to form a fine grid mesh. On the other hand, the clustering of infection in positive herds allows the selection of a relatively small numbers of animals to be tested within each herd. The simulation study, therefore, indicates that the number of sentinel animals within herds could be reduced by one third with significant advantages for livestock owners who can then maintain a lower number of unvaccinated animals in their herds resulting in significant savings for them and for the Veterinary Services. The simulation model described here was designed and validated using data derived from the Italian serological surveillance plan. Results, therefore, are applicable only to the situation in Italy and cannot be extrapolated to others. However, the mathematical approach used in the present model could prove useful in developing simulation models in countries other than Italy. References 1. Calistri P., Giovannini A., Conte A., Nannini D., Santucci U., Patta C., Rolesu S. & Caporale V. (2004). Bluetongue in Italy: Part I. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds).proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Giovannini A., Calistri P., Nannini D., Paladini C., Santucci U. & Caporale V. (2004). Bluetongue in Italy: Part II. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds).proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Giovannini A., MacDiarmid S., Calistri P., Conte A., Savini L., Nannini D. & Weber S. (2003). The use of risk assessment to decide the control strategy for bluetongue in Italian ruminant populations. J. Risk Anal., 24 (6), Giovannini A., Conte A., Calistri P., Di Francesco C. & Caporale V. (2004). Risk analysis on the introduction into free territories of vaccinated animals from restricted zones. In Bluetongue, Part II (N.J. MacLachlan & J.E. Pearson, eds).proc. Third International Symposium, Taormina, October Vet. Ital., 40 (4), Giovannini A., Paladini C., Calistri P., Conte A., Colangeli P., Santucci U., Nannini D. & Caporale V. (2004). Surveillance system of bluetongue in Italy. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Vose D. (2000). Risk analysis: a quantitative guide, 2nd Ed. John Wiley & Sons, Chichester, 418 pp. 364 Veterinaria Italiana, 40 (3), 2004

319 Vet. Ital., 40 (3), Epidemiology and vectors Data management and analysis systems for bluetongue virus zoning in Australia A.R. Cameron AusVet Animal Health Services, 140 Falls Road, Wentworth Falls, NSW 2782, Australia Summary Bluetongue virus (BTV) monitoring data in Australia is managed using a secure web-enabled centralised database. Scientists across the country submit virological and entomological data from sentinel and other sites using the Internet. Automated reporting and mapping systems make this data immediately available to all users. This system underpins the process used for defining zone boundaries. Immediate access to monitoring results allows the zones to be redefined as soon as any evidence of expansion of the area of BTV activity is detected. The method used to define zone boundaries, incorporating detailed information on vector and virus activity, property boundaries and subdivision boundaries, geography and climate, is described. Keywords Australia Bluetongue virus Free zone Geographic information system Information system Internet Mapping Surveillance zone Zones. In many countries where bluetongue virus (BTV) is present, environmental and other factors influence the distribution of the virus and its vectors, resulting in free and infected areas, which can change over time. The Office International des Épizooties (OIE) has introduced the concept of zones to facilitate safe trade from those areas of a country that are free from the virus. Substantiation of zone boundaries and zone status requires effective monitoring systems to detect changes in the distribution of BTV, in response to changing climatic or environmental conditions. Since the 1970s, Australia has monitored virus and vector distribution using a combination of sentinel animals for serosurveillance and virus isolation, and vector collections at sentinel and other sites, supplemented with cross-sectional surveys. Specimens from the field are analysed in a national network of laboratories. Novel approaches are required to rapidly collect, analyse, interpret and disseminate the large quantity of data generated by the monitoring programme from laboratories in different parts of a very large country with a wide range of environments. This paper describes the systems developed in Australia to manage BTV monitoring data, and to rapidly generate up-to-date zone maps in response to shifts in the distribution of the virus. Data sources Four main sources are used to provide monitoring data as follows: 1) Data on the seroconversion of animals within sentinel herds. This is supplemented with serotyping and viral isolation in key areas. There is an extensive network of sentinel sites covering all of Australia. Specimens are collected at intervals of between two weeks (intensively monitored areas with maximum variability in BTV distribution) and six months (areas with a long-term history of constant status). All animals are tested to demonstrate BTV seronegative status before being enrolled in sentinel herds. 2) Data on serological status of animals derived from periodic cross sectional surveys. These surveys involve only young animals (usually between six and 18 months) known to have been born and reared on the property of origin. 3) Vector trapping with the identification and quantification of Culicoides species. Vector traps are located at each sentinel herd site and numerous other strategically located sites. 4) Background textual information. The practical field expertise of scientists with many years of BTV experience is captured in the form of Veterinaria Italiana, 40 (3),

320 quarterly reports. The aim of these reports is to document factors influencing the monitoring or distribution of BTV, which otherwise would not be captured in purely numerical data. For instance, a cyclone in the north-west of the country may destroy all vector traps, as well as temporarily removing all vectors from a region. Data collection and management Historically, local data management by participating laboratories was based on a variety of spreadsheet and database software. In order to minimise disruption to laboratory staff, these systems have been retained. A centralised Internet-based real-time database is used for data storage, management and analysis (1). Data is submitted using a secure web interface, with access limited to authorised scientists to avoid data corruption. A block of local data in standardised format (e.g. from a spreadsheet) is simply copied and pasted onto the web page, and submitted to the central server. The data is parsed, checked for completeness and errors, and merged into the central database (checking for duplicates and updating previously submitted data). This system allows large volumes of detailed monitoring data to be integrated into the national database with only a few keystrokes, as well as allowing laboratory staff to update any previously submitted data based on subsequent tests. The core data held by the database includes the following: Virology Date of sampling Location of sampling Number of animals sampled Number of animals at risk of seroconversion (i.e. negative prior to the sampling period) Number of animals seroconverted (i.e. previously negative and now positive) Summary analysis of viral or vector activity at monitoring sites Records of data submissions by laboratory to monitor programme participation. As analysis is based on the on-line database, all reports immediately reflect the most recently available data. A web-enabled geographic information system (GIS) provides real-time automated mapping of the monitoring data, and allows users to view sentinel herd or vector trapping results along with the current BTV zone boundaries, roads, railways, towns, rivers etc (Fig. 1). In addition, several levels of administrative boundaries are available, including individual property boundaries. Users can zoom in to determine if a specified property is inside or outside a zoning boundary, to an accuracy of about 5 m (Fig. 2). This system displays the location of monitoring sites and uses colour coding to indicate if positive results have been observed (either seroconversions or vector species detected). The maps also allow interactive querying, so that the full monitoring details and history of a particular site can be accessed by clicking on that site (Fig. 3). This system provides immediate, up-to-date visual assessment of BTV activity in relation to the current zone boundaries, as well as giving access to the underlying data. It is used extensively to facilitate communication between programme scientists around the country during live telephone conferences. It is also used as a mechanism to communicate proposed changes to zone boundaries for approval by local experts. Entomology Date of trapping Location of trapping Species identified Number of Culicoides of that species identified. The website also provides a range of data management and reporting options, including: Creation of new monitoring site records, identified by geographic coordinates Export of raw data to spreadsheet format for download and analysis Figure 1 Interactive real-time bluetongue virus monitoring map interface, showing zoning boundaries (at the time of writing) and sentinel herd monitoring results 366 Veterinaria Italiana, 40 (3), 2004

321 possible, zone boundaries follow clearly defined administrative subdivisions or property boundaries. An entire property must lie within the free zone in order to be classified as part of that zone. In some cases, the boundaries follow geographical barriers to vector spread, such as mountain ranges. Using desktop GIS tools, it is usually possible to define new zone boundaries and update them on the website in less than a day. This process is illustrated in Figures 4 to 7. An automated distribution list is employed to inform registered users of zone updates. Figure 2 Detail of map in Figure 1 zoomed in to show the boundary between the free zone and the surveillance zone in south-eastern Queensland Blue lines indicate individual property boundaries Bluetongue seroconversions at monitoring site No serconversions Figure 4 Step 1 of the zone boundary definition A live link is established to the central database and monitoring site results mapped as points Figure 3 Clicking on a point on the map produces a pop-up window showing the full monitoring results from the specified site Zoning Zone boundaries are determined on the basis of monitoring results, geographical and environmental factors, and the outputs of a temporo-spatial model of vector and virus distribution (2). Three zones are defined: the free zone, the surveillance zone and the zone of possible transmission. Boundaries are determined such that the boundary between the zone of possible transmission and the surveillance zone is at least 50 km from the nearest identified BTV activity and makes biological sense. The surveillance zone is at least a further 50 km wide, providing a separation of at least 100 km between the free zone and any area of known activity. Where Figure 5 Step 2 of the zone boundary definition Buffers are drawn around each positive site at distances of 50 km (minimum distance to surveillance zone) and 100 km (minimum distance to free zone) Veterinaria Italiana, 40 (3),

322 zone requires immediate re-drafting of the zone boundaries. Decreases in the distribution of BTV resulting in demonstrable lack of activity in an area for at least two years also result in changes to zone boundaries to expand the free zone. These reviews occur after a meeting of experts who consider all the monitoring results and other relevant factors. The system described demonstrates how advanced information management tools can be used to maximise the effectiveness of a BTV monitoring programme. The system supports the definition of timely and accurate zone boundaries, providing a high degree of confidence as well as providing a strong scientific basis to negotiation of health protocols to support trade. The current Australian BTV zones are displayed at Figure 6 Step 3 of the zone boundary definition Properties and administrative subdivisions intersecting the buffers are identified Acknowledgements The system described in this paper is managed by Animal Health Australia, with collaborative funding from the Australian State and Commonwealth governments, and the Australian livestock industries. The attendance of the author at the Symposium was supported by the Australian cattle, sheep and live export industries. References Figure 7 Step 4 of the zone boundary definition Zone boundaries defined to enclose all selected properties, taking into account geographical features, model outputs and climatic information 1. Cameron A.R. (2000). Development of an internetbased information system for monitoring veterinary arboviruses and their vectors. In Proc. IXth Symposium of the International Society for Veterinary Epidemiology and Economics (ISVEE), Breckenridge, 6-11 August. ISVEE, Fort Collins. 2. Cameron A.R. (2000). Development of a temporospatial disease distribution model for arboviruses and their vectors in Australia. In Proc. IXth Symposium of the International Society for Veterinary Epidemiology and Economics (ISVEE), Breckenridge, 6-11 August. ISVEE, Fort Collins. The protocols governing changes to zoning boundaries are clearly defined. Any evidence of BTV activity within less than 100 km of the current free 368 Veterinaria Italiana, 40 (3), 2004

323 Vet. Ital., 40 (3), Epidemiology and vectors Surveillance system of bluetongue in Italy A. Giovannini (1), C. Paladini (1), P. Calistri (1), A. Conte (1), P. Colangeli (1), U. Santucci (2), D. Nannini (1) & V. Caporale (1) (1) Istituto Zooprofilattico Sperimentale dell Abruzzo e del Molise G. Caporale, Via Campo Boario, Teramo, Italy (2) Ministero della Salute, Direzione Generale della Sanità Pubblica Veterinaria, Alimenti e Nutrizione, Piazzale Marconi, Rome, Italy Summary The authors provided details of the bluetongue surveillance and the Internet-based information systems that were implemented in Italy. The systems were structured with the aim of gathering and spreading information and data to support decision-making, management of control activities and provide an early warning system. Information and data generated by the surveillance system enabled the detailed analysis of bluetongue epidemiology, vector distribution and vector population dynamics. This information and data also allowed the analysis of risk factors associated with vector spread and animal movements, which resulted in and increased the flexibility and the efficiency of the enforcement of control measures. Keywords Bluetongue Epidemiological surveillance Information system Information technology Italy. During August 2000, Italy experienced the largest bluetongue (BT) epidemic in Europe (2). At the time, control of BT was covered by European Union (EU) Directive 92/119/EEC (12), together with some other Office International des Épizooties (OIE) List A diseases (such as foot and mouth disease, rinderpest, sheep pox, swine vesicular disease etc.), and EU Directive 92/35/EEC (11) had already defined specific measures against African horse sickness. Directive 92/119/EEC outlined direct control measures and the demarcation of a 3-km radius protection zone and a 10-km radius surveillance zone, around infected farms. Direct control measures included the slaughter of all susceptible animals on farms and the possible extension of such measures to neighbouring farms suspected of being exposed. In November 2000, the EU, after reviewing the provisions of Directive 92/35/EEC (11), issued EU Directive 2000/75/EC (13) defining specific rules for the control and eradication of BT. In particular, Directive 2000/75/EC established the following: a) a 100-km radius protection zone around the outbreaks or any farm on which virus circulation was confirmed b) establishment of a 50-km radius surveillance zone around the protection zone c) slaughter of all animals deemed necessary to prevent the spread of the epidemic and, the destruction of the carcasses of those animals d) implementation of serological and entomological surveillance programmes in the protection and surveillance zones e) prohibition of animal movement from protection and surveillance zones. To complement these measures, the Directive foresaw the possibility of performing vaccination in the protection zone. Within the context of its general principles, Directive 2000/75/EC allows for the regulation of specific aspects by Commission Decisions (13). If relevant information on the spread of infection and risk factors were considered, this Directive allowed some flexibility in planning and enforcing control measures. Therefore, in Italy, a control strategy was adopted, combining direct and indirect control measures with an intense surveillance programme. The aim of the control measures was to reduce virus circulation in susceptible animal populations. The objective of the surveillance system to collect and analyse the Veterinaria Italiana, 40 (3),

324 information needed to: a) classify the territory in relation to the presence/absence of virus circulation b) evaluate risk factors associated with the spread of vectors and animal movements c) establish an early warning system. The authors describe the BT surveillance programme and the Internet-based information system implemented in Italy. Analysis of the systems Surveillance system The system was based upon regular: a) recording of all suspected and confirmed clinical cases of BT b) recording of results of periodical testing of sentinel animals c) reporting on monitoring of the spread of vectors and their seasonal dynamics d) recording of all diagnostic results e) recording progress of vaccination campaigns. The system also records results of specific ad hoc monitoring, such as the evaluation of the immune status of vaccinated populations, risks posed by transhumant animals (18) and verification if BT is present when testing results were inconclusive. The following laboratory tests were conducted exclusively at National Reference Centre for Exotic Diseases (CESME: Centro Studi Malattie Esotiche) in Teramo: virus isolation, serotyping of viral isolates, entomological identification of insects, polymerase chain reaction (PCR) testing of blood, tissues and insects, and virus neutralisation (VN) test confirmation of positive competitive enzyme-linked immunosorbent assays (c-elisa) from local Istituti Zooprofilattici Sperimentali (IZS) laboratories. The reason why samples were sent to the laboratory was recorded in the information system and included: trace-back of animals from protection and surveillance zones, follow-up testing of a previously positive farm, confirmation of suspected clinical cases, testing of sentinel animals, surveys of infected or adjoining areas, surveys of vaccinated animals and evaluation of vaccine-related problems. Data was obtained from local Veterinary Services and from laboratories that supplied the data for the national information system. Information generated was circulated over the Internet and provided a common source to facilitate decision-making processes and the management of activities at central, regional and local level and also enabled verification of the consistency and efficiency of control activities in relation to national objectives. Clinical surveillance When clinical signs that were suspected as being BT were observed, a clinical examination on all susceptible animals present on the farm was performed and samples were collected for laboratory testing. During farm visits, the following information was collected and recorded: national farm identification code, postal address and geographic co-ordinates, date of onset of symptoms, number of susceptible animals present on the farm by age (<6 months of age, >6 months of age), number of diseased animals and number of dead, slaughtered or destroyed animals. Specimens sent to the laboratory were accompanied by the following information: national farm identification code, individual identification code of animal sampled, type of test requested (virological and/or serological) and the reason for sampling. Weekly follow-up clinical visits were performed on the farm to monitor the evolution of disease and to update the information system. Clinical visits were discontinued when either the suspicion of BT was ruled out or when there was no further evidence of virus circulation. When the presence of BT was confirmed, the clinical visits were extended to all ovine flocks within a radius of 20 km of a confirmed case with clinical disease or 4 km of a confirmed case with subclinical infection. During the clinical visits, the following data were collected and submitted: national farm identification code, geographic co-ordinates, altitude of the farm, date of and reason for the visit, total number of animals, number of animals examined and presence or absence of clinical signs of BT. Serological surveillance In the summer of 2001, in order to detect or exclude the presence of BT, a serological surveillance system based on sentinel animals was implemented in the protection and surveillance zones, and in areas at risk of infection (Fig. 1). In October 2001, the system was extended nationwide (Fig. 2). Italy was divided into grids of square units of either 400 km 2 or km 2 according to the occurrence or risk of introducing infection, respectively. To detect a 5% infection rate with a 95% confidence level in each 400 km 2 unit, a sample of 58 bovine animals was selected from 5 to 8 farms. To detect a 2% infection rate with a 95% confidence level in each unit of km 2, a sample of 148 bovine animals was selected from 8 to 12 farms (Fig. 3). If cattle were not present in the area, sheep were selected as sentinel animals. Sentinels were bled regularly with variable frequency, depending on the season and infection rate in the area. Blood samples were 370 Veterinaria Italiana, 40 (3), 2004

325 collected at least once every 30 days in the protection and surveillance zones and, from May-December, samples were collected every fortnight. In infectionfree zones, sentinels were examined every 30 days from May to December and every 60 days from January to April. Blood samples were tested for BT using the c-elisa by the local IZS laboratory and positive results were confirmed by the VN test at the CESME. >Foreseen density level +10% Foreseen density level ±10% <Foreseen density level 10% Area 1 Area 2 Area 3 Area 4 Area 5 Figure 1 Five subdivisions of Italy, according to five different risk levels, May 2001 Figure 2 Subdivision of Italy into two areas, September 2001 Area A: squares of 400 km 2 Area B: squares of km 2 Figure 3 Density of sentinel animals per km 2 Each province was classified on the basis of the number of sentinel animals per km 2 as predicted by the serological surveillance plan Entomological surveillance From August 2000 until October 2001, an entomological surveillance programme was implemented in the protection and surveillance zones to map the distribution of vectors, with particular reference to Culicoides imicola. Blacklight traps were moved around the study areas to define the distribution of C. imicola. Permanent blacklight traps were operated from June to October 2001 in selected sites in various parts of the country to evaluate the effect of soil type on C. imicola presence. Since October 2001, entomological surveillance was extended nationwide (Fig. 2). Blacklight traps, were positioned in fixed locations in all provinces (Fig. 4) and operated weekly to monitor Culicoides population dynamics (17). Blacklight traps were also operated on a temporary basis in cases of suspected or confirmed virus circulation and whenever a more specific understanding of vector distribution was required. For each insect catch, the following data were collected: national farm identification code, geographic co-ordinates, altitude of the farm, animal species present, date of capture, minimum and maximum temperatures during the night of the capture. Captures were examined to determine the total number of insects, the total number of Culicoides and the total number of C. imicola. Veterinaria Italiana, 40 (3),

326 Progress of the vaccination campaign Vaccination in infected and in at-risk zones commenced in the late autumn of 2001 (15). Since vaccination levels of susceptible populations in Italy were linked to approval of animal movement, progress of vaccination was recorded in the national BT information system. Data recorded were: Local Health Unit where vaccination was performed, total number of susceptible animals vaccinated, vaccine serotype/s used, vaccine batch number, national farm identification code, total number of susceptible animals on the farm, number of eligible animals, number of vaccinated animals by species and vaccination date. Geo-coded permanent traps Not geo-coded permanent traps Figure 4 Geographical distribution of permanent traps in Italy, 30 July 2003 Ad hoc monitoring Ad hoc monitoring was conducted to verify: a) antibody prevalence in vaccinated populations b) the risk linked to animal transhumance c) the extent of virus circulation in certain zones where single seroconversions or unexpected positive results were found. Antibody prevalence in vaccinated populations was evaluated by serological examination of 35 randomly selected vaccinated animals per grid cell. The risk linked to animal transhumance was evaluated through specific plans (18), which required entomological and virological examination of insects collected in both the departure and the arrival pastures and serological examination of nonvaccinated animals in transhumant flocks. The extent of BT virus (BTV) circulation in some zones, where single seroconversions or unexpected positive results were detected, was investigated by either PCR testing or serology, or by both PCR and serology of 58 randomly selected animals in the municipalities located within a 20 km radius around the suspect farm. In general, PCR was performed in vaccinated populations, while serology was used in unvaccinated ones. In either case, test results were recorded together with the national farm identification code, geographic co-ordinates of the farm, individual identification code for each animal, date and type of sample taken. Side-effects of vaccine Possible undesirable side effects of the vaccine (deaths, abortions, stillbirths) were monitored by: a) sampling animals in flocks where problems arose and testing dead animals and/or foetuses for the presence of vaccine virus b) collecting information concerning type and incidence of disease observed, vaccine used, dates of vaccination, etc. Samples were submitted to the laboratory for differential diagnosis and if BTV was isolated, the virus was identified (vaccine or field virus). Information system To ensure effective control of data and the link with surveillance activities, all laboratory test results were recorded in the local data bank of the IZS which was responsible for testing in the zone in which they were located. Data recorded in the local data banks, according to a pre-defined record layout, were transmitted weekly by from the local IZS to the National Information System (NIS) for BT at the CESME (Fig. 5). Clinical surveillance, which included suspected and confirmed outbreaks, municipal summaries of the clinical visits and vaccination data, were sent to the NIS each week by local Veterinary Services, either directly or through the regional services or through the local IZS. To facilitate data transmission, specific software was prepared by the CESME. Data were processed with the objective of providing a common base for all stakeholders that would provide the following: a) an early warning system b) a decision support tool to verify and plan activities by regional and national government bodies c) a tool to manage routine activities. 372 Veterinaria Italiana, 40 (3), 2004

327 To this end, most information was made available online on the CESME website (izs.it/bluetongue/ bluehome.html) and updated weekly (Figs 6 and 7). Only information concerning ad hoc monitoring and vaccine side-effects were not available online on a weekly basis, as this data required in-depth analysis. The ad hoc monitoring reports were, however, made available on the website in the form of specific reports ( documents at izs.it/bluetongue/documenti /bt_index.html or new, izs.it/bluetongue/vaccini. pdf). Information on the disease and on the relevant legislation was also available on the website (Fig. 6). SN Test CESME Flow of samples Data flow Figure 5 Flow chart of samples and data Figure 6 Contents of the National Information System for bluetongue Veterinaria Italiana, 40 (3),

328 Figure 7 Web pages accessible by password from the National Information System for bluetongue: details of data available Early warning system and decision-making support The list of infected municipalities (derived from the notification of outbreaks or from serological surveillance) was available in the form of a table or map. Moreover, the table of infected municipalities was divided into the lists of newly infected municipalities and of municipalities that had regained free status (Figs 6 and 8). This category of information can be used both for early warning to highlight the trend of the epidemic and for management and decision support for applying animal movement restriction measures or vaccination. Maps showing the location of sentinel animals that had seroconverted were provided and the BTV serotype(s) implicated was indicated on the map. The C. imicola population dynamics in the main permanent traps in each region was also described in a series of graphs. These latter items of information were also provided in the form of Microsoft PowerPoint presentations of the evolution of the epidemic during the previous 100 days; this was updated every Monday (Figs 6 and 9). Finally, the monthly mean of the daily minimum and maximum temperatures nationwide were also available on the website (Fig. 10). These were derived from analyses of 10-year data provided by the Italian Military Air Force and collected at more than 100 weather stations uniformly distributed across Italy (Fig. 11). Figure 8 List of new infected municipalities, 4 August 2003 The number and location of all outbreaks by year or by seasonal epidemic was available both in tabular form (Figs 7 and 12) and as a dynamic map (Fig. 13). These data may be aggregated at municipality or 374 Veterinaria Italiana, 40 (3), 2004

329 Local Health Unit level. The tables also provided data on the following: a) number of outbreaks b) total number of sheep/goat flocks c) total number of animals in infected flocks d) number of diseased, dead, slaughtered and destroyed animals in outbreaks. Dynamic maps were also available showing the percentage of the population that had been vaccinated by animal species or by serotype (Fig. 14). Results of entomological surveillance were mapped by seasonal epidemic displaying the presence/ absence of C. imicola by municipality (Fig. 13). Data on the serological and entomological surveillance Figure 9 Maps showing the location of serologically positive animals, bluetongue serotypes identified and the presence or absence of Culicoides imicola Veterinaria Italiana, 40 (3),

330 Services, such as entomological surveillance, vaccination and health status of each area. Other management support data were also available, such as: detailed data on each holding investigated by the surveillance system nationwide, results of all diagnostic, serological, virological and entomological activities performed in a specific holding (Figs 7 and 15). In the case of an infected holding, the date of onset of clinical signs, dates of clinical visits and total number of animals, in addition to the number of diseased, dead, slaughtered, and destroyed animals, were available. Figure 10 Mean value of minimum and maximum monthly temperatures for January 2003 Ancillary information Ancillary information (Fig. 6) includes the following: a) EU legislation (directives and decisions) and all the national regulations on BT b) software to record and despatch outbreak data, clinical visits, vaccination management and serology c) forms and procedures for data input, including record layouts for data transmission to the NIS d) a detailed description of the epidemiology, symptoms, pathology, laboratory diagnosis, prophylaxis and control of BT, in addition to a gallery of pictures and a comprehensive bibliography. Rules for access to information Access was regulated by password according to the type of user (Fig. 16). In general, official veterinarians and laboratories could access all data concerning their territory, while the Ministry of Health and the CESME had access to all information. A public section was also available and included a compendium of the information described above. Weather station Figure 11 Distribution of weather stations of the Italian Air Force Meteorological Service performed during the previous 100 days were provided as Microsoft PowerPoint presentations (Figs 6 and 9). Other data useful to the decisionmaking process and available in the website concerned clinical visits (Fig. 7), subdivided by reason for the visit (animals introduced from protection/surveillance zone; surveillance in protection/surveillance zone; visits to and around suspected outbreak; monitoring of risk zones) and aggregated by the municipality or Local Health Unit. This information provided support for the management of daily activities of local Veterinary Discussion The strict enforcement of Directive 2000/75/EC (13) caused very serious problems to the animal production sector of at least one third of Italy. After the first month of the epidemic, Decision 2001/138/EC (4) instituted protection and surveillance zones according to the criteria established by Directive 2000/75/EC (13) and movement restrictions were enforced in one-third of Italy. The movement of susceptible animals was prohibited from the regions of Sicily, Sardinia and southern Italy into infection-free regions. Consequently, it became virtually impossible to either fatten or cull cattle as the production system was organised to perform this type of activity in the Po Valley which was in the free zone. Moreover, 376 Veterinaria Italiana, 40 (3), 2004

331 Figure 12 Number and location of all outbreaks by seasonal epidemic, year and municipality Veterinaria Italiana, 40 (3),

332 Figure 13 Results of entomological surveillance 378 Veterinaria Italiana, 40 (3), 2004

333 Figure 14 Example of the information made available by Government Maps show the percentage of the vaccinated population by species and/or by serotype Veterinaria Italiana, 40 (3),

334 Figure 15 Detailed data on each holding investigated by the surveillance system 380 Veterinaria Italiana, 40 (3), 2004

335 Figure 16 Sections of the information system accessible by password only animal transhumance from the winter pastures in Apulia (surveillance zone) to the summer pastures in Abruzzo and Molise (free zones) was also blocked. The very nature and ecology of the BTV means that it cannot be eliminated from infected areas rapidly. The animal movement limitations imposed on BT susceptible species, on two-thirds of the Italian territory, was to result in either a complete reorganisation of the structure of the Italian bovine and ovine animal production sector or closing it down. A reorganisation of this magnitude appeared unlikely given not only the lack of both economic and human resources and the chronic shortage of adequate infrastructures, but the most important factor was the short time available. Things had been made much more difficult by the refusal of the EU Veterinaria Italiana, 40 (3),

336 to allow the Italian government to compensate farmers for indirect losses due to the impossibility to access markets because of the animal movement restrictions. The problem was aggravated by the spread of the disease up the northern Tyrrhenian coast (Latium and Tuscany) (18). If these restrictions on animal movements had been enforced indefinitely or until BTV had been eradicated, which was very unlikely, the result would probably have been an irreversible decline of the agricultural sector with significant attendant social problems. Since the commencement of the outbreak, authorities in Italy selected a strategy based on risk assessment and management with the objective of slowing, as far as possible, spread of infection and to develop a plan to alleviate the inconvenience caused by movement restrictions. After the first months of the epidemic, a number of factors highlighted the necessity for an in-depth understanding of the epidemiology of BT infection in Italy. Some of the elements that have characterised the seriousness of BT in Italy are the severity of the epidemic pattern in Sardinia, its spread to southern Italy (2), the lack of knowledge of vector biology and the lack of knowledge of BT epidemiology in countries like Italy that are located at the northern border of BT distribution. It has been well established that the circulation of BT is associated with a specific geographical area rather than on individual farms and, consequently, animal movement restrictions and trace-back activities were not applied to individual farms but to all the farms located in an infected zone. Therefore, management of control activities required regular oversight and co-ordination by Veterinary Services at the local, regional and central levels; efficiency of communication was of paramount importance. Information channels also had to be accessible to veterinary authorities in infection-free zones as they required punctual information on the distribution of infection in order to be able to evaluate the risks related to the introduction of animals from other regions and to implement the relevant safeguards. Italy was the first country in the world in which BT caused a wide-scale epidemic in a previously infection-free susceptible population. The rapid implementation of a system that was able to support an organisation as complex as that needed to control BT in Italy, required the collection of massive quantities of data which required very intensive field activity. Response to the epidemic included, among other things, the periodical testing (in most cases fortnightly) of more than sentinel animals (1) and the placing of about 250 permanent insect traps nationwide. Information and data produced by this surveillance system constitute the information base of the early warning system for BT in Italy. The information system was implemented online in an Internet environment. The advantages of this Internet system were that it was widely accessible by all parties concerned, that information was disseminated immediately and nationwide and that the privacy of sensitive data was controlled by a password system. Data and information generated by the surveillance system have contributed to the knowledge on BT epidemiology by providing a better understanding of the distribution and dynamics of its vectors (3, 17), the ability to monitor the spread of infection (2, 15), and the possibility to evaluate the risk factors linked to the spread of vectors and to animal movements. This knowledge has probably contributed to the flexibility that now characterises the EU Directives on BT. The information generated by the Italian surveillance system probably contributed to the amendments to the original European Union decision 2001/138/EC (4). The following are EU decisions that were developed: a) Decision 2001/783/EC of 11 September 2001 (5) recognised the validity of two OIE standards (Terrestrial animal health code) (19), namely: Article for animal movements from infected zones and Article for the definition of seasonally free areas; furthermore, the same Decision reduced the radius of the zone from which slaughter animals cannot be sent to free zones from 100 to 20 km, provided a surveillance system was in place b) commencing in January 2002, three Decisions, namely: 2002/35/EC (6), 2002/189/EC (7), 2002/543/EC (8), excluded from the surveillance zone some Italian provinces in which the surveillance system had documented the absence of virus circulation c) Decision 2003/14/EC (9) taken in January 2003 that allowed the shipment of slaughter animals from infected to free areas, provided that the province of origin had a vaccination coverage of susceptible populations of at least 80% and a risk assessment had been performed d) Decision 2003/218/EC (10) taken in March 2003 that introduced the concept of risk and subdivided the territories into areas of higher and lower epidemiological risks. Furthermore, the decision allows the shipment of live animals from the lower risk areas where viral circulation has not been detected to the entire territory of the EU and the movement of slaughter animals from lower risk areas, even with active infection, and from higher risk areas, where viral circulation has not been detected, to free areas in the national territory. The latter was 382 Veterinaria Italiana, 40 (3), 2004

337 allowed only if the animals had been vaccinated at least more than 30 days previously, belonged to a herd in which all the animals were vaccinated and transport occurred during daylight hours only. According to the decision 2003/218/EC, the Member State, can demarcate epidemiologically relevant areas of origin autonomously; in other words, on the basis of surveillance results, it can modify the radius of the protection zone to more than or less than 20 km and it can modify the risk level of a zone from low risk to high risk. It has also been possible, through a series of risk assessments to define the national BT control strategy (14), the risk linked to animal movement from restriction zones, in relation to presence or absence of documented viral circulation and to the level of vaccination in susceptible populations (16), and the sensitivity of the serological surveillance system (1). Finally, the online information system has grown progressively with the evolution of the epidemic. The system was tailored to the evolution of knowledge on BT, control measures and legislation. The initial design of the central core of the website followed top-down logic. Progressive modifications, most of them generated in emergency situations, have led to a loss of the original design of the information system, increasing data complexity and sometimes making it difficult to retrieve specific information. For this reason, the system has been completely re-designed and the prototype is currently undergoing its validation phase. References 1. Calistri P., Giovannini A., Conte A. & Caporale V. (2004). Use of a Montecarlo simulation model for the re-planning of bluetongue surveillance in Italy. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Calistri P., Giovannini A., Conte A., Nannini D., Santucci U., Patta C., Rolesu S. & Caporale V. (2004). Bluetongue in Italy: Part I. In Bluetongue (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Conte A., Giovannini A., Savini L., Goffredo M., Calistri P., Meiswinkel R. (2003). The effect of climate on the presence of Culicoides imicola in Italy. J. Vet. Med. B, 50: European Commission (2001). Commission Decision of 9 February 2001 establishing protection and surveillance zones in the Community in relation with bluetongue (2001/138/EC). Off. J., L 050, European Commission (2001). Commission Decision of 9 November 2001 on protection and surveillance zones in relation to bluetongue, and on rules applicable to movements of animals in and from those zones (2001/783/EC). Off. J., L 293, European Commission (2002). Commission Decision of 16 January 2002 amending Decision 2001/783/EC as regards the protection and surveillance zones in relation to bluetongue in Italy (2002/35/EC). Off. J., L 015, European Commission (2002). Commission Decision of 5 March 2002 amending Decision 2001/783/EC as regards the protection and surveillance zones in relation to bluetongue in Italy (2002/189/EC). Off. J., L 063, European Commission (2002). Commission Decision of 4 July 2002 amending Decision 2001/783/EC as regards the protection and surveillance zones in relation to bluetongue in Italy (2002/543/EC). Off. J., L 176, European Commission (2003). Commission Decision of 10 January 2003 amending Decision 2001/783/EC as regards the bluetongue protection and surveillance zones and conditions for movements of animals for immediate slaughter (2003/14/EC). Off. J., L 007, European Commission (2003). Commission Decision of 27 March 2003 on protection and surveillance zones in relation to bluetongue, and on rules applicable to movements of animals in and from those zones and repealing Decision 2001/783/EC (2003/218/EC). Off. J., L 082, European Council (1992). Council Directive 92/35/EEC of 29 April 1992 laying down control rules and measures to combat African horse sickness. Off. J., L 157, European Council (1992). Council Directive 92/119/EEC of 17 December 1992 introducing general Community measures for the control of certain animal diseases and specific measures relating to swine vesicular disease. Off. J., L 062, European Council (2000). Council Directive 2000/75/EC of 20 November 2000 laying down specific provisions for the control and eradication of bluetongue. Off. J., L 327, Giovannini A., MacDiarmid S., Calistri P., Conte A., Savini L., Nannini D. & Weber S. (2003). The use of risk assessment to decide the control strategy for bluetongue in Italian ruminant populations. J. Risk Anal., 24 (6), Giovannini A., Calistri P., Nannini D., Paladini C., Santucci U., Patta C. & Caporale V. (2004). Bluetongue in Italy: Part II. In Bluetongue, Part I (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (3), Giovannini A., Conte A., Calistri P., Di Francesco C. & Caporale V. (2004). Risk analysis on the Veterinaria Italiana, 40 (3),

338 introduction into free territories of vaccinated animals from restricted zones. In Bluetongue, Part II (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (4), Goffredo M., Conte A.M., Cocciolito R. & Meiswinkel R. (2003). The distribution and abundance of Culicoides imicola in Italy. Vet. Ital., 39 (47), Nannini D., Calistri P., Giovannini A., Di Ventura M., Cafiero M.A., Ferrari G., Santucci U. & Caporale V. (2004). Health management of large transhumant animal populations and risk of bluetongue spread to disease-free areas. In Bluetongue, Part II (N.J. MacLachlan & J.E. Pearson, eds). Proc. Third International Symposium, Taormina, October Vet. Ital., 40 (4), Office International des Épizooties (OIE) (2003). Terrestrial animal health code, 12th Ed. OIE, Paris (oie.int/eng/normes/mcode/a_summry.htm accessed on 3 May 2004). 384 Veterinaria Italiana, 40 (3), 2004

339 Vet. Ital., 40 (3), Epidemiology and vectors Bluetongue surveillance in the Campania Region of Italy using a geographic information system to create risk maps V. Caligiuri (1), G.A. Giuliano (1), V. Vitale (1), L. Chiavacci (2), S. Travaglio (2), L. Manelli (3), S. Piscedda (3), M. Giardina (3) & R. Mainolfi (3) (1) Osservatorio Epidemiologico Veterinario Regione Campania, Istituto Zooprofilattico Sperimentale del Mezzogiorno, Via Salute 2, Portici, Italy (2) Istituto Zooprofilattico Sperimentale del Piemonte, Liguria e Valle D Aosta, Via Bologna, Torino, Italy (3) Formez, via Campi Flegrei 34, Arco Felice, Pozzuoli, Italy Summary The aim of the project was the implementation of a geographic information system (GIS) to study areas of the Campania Region that are potentially at risk for bluetongue (BT) infection. As a first step, environmental, meteorological and climatic features were surveyed to evaluate areas where Culicoides could be present. A risk map was developed and five zones with different risk levels were defined. Data from Culicoides trapping were surveyed to evaluate the capability of the risk map to predict presence/absence of vectors. Finally, data from 2002 BT surveillance (outbreaks and serologically positive sentinels [SPS]) were compared to the map. Captures of Culicoides, SPS and BT in areas the map indicated as a medium/high risk level, seem to confirm reliability of the produced map. However, very few C. imicola were captured in these medium/high risk areas. Keywords Bluetongue Campania Culicoides Geographic information system Italy Mapping Regression Remote sensing Surveillance Vector-borne disease. Introduction Bluetongue (BT) is an infectious disease of ruminants caused by bluetongue virus (BTV) transmitted by biting midges of the genus Culicoides of which C. imicola has been implicated as the major vector in the Mediterranean Basin. The first outbreak of BT in Italy was reported in Sardinia in 2000 and in the same year many other regions of southern Italy were affected by the disease (8). Areas in which C. imicola are found are considered at risk to BT but usually the vector is not identified until after outbreaks of the disease have occurred. Determining Culicoides distribution before the disease outbreaks occur could be useful to target control measures, such as vaccination, use of insecticide and housing of susceptible animals in periods of peak vector activity (13). The goal of this study was to implement a geographic information system (GIS) to define a risk map (12) for the presence/absence of Culicoides. Climatic and environmental factors primarily govern the distribution of the vector, but soil type may also influence the establishment of C. imicola in any given zone (9). C. imicola persists in areas where the average daily maximum temperature exceeds 12.5 C, requiring wet, organically enriched soil or mud and an optimal annual rainfall of mm. Other factors, such as irrigation and standing water near water troughs, may influence humidity and promote breeding with areas in the vicinity of animal holdings being the most suitable. The vector can fly for distances of a few hundred metres, but warm-humid winds can transport them for distances of greater than 7 km and to altitudes of m (6). Methods Two LandSat 7 (30 mt pixels) remote sensing images of the Campania Region were prepared in 2000, georeferenced in UTM ED50 with Erdas Imagine 8.4 software (Fig. 1). Using LandSat images with the minimum distance algorithm, the land use was determined and listed under the following Veterinaria Italiana, 40 (3),

340 classes: evergreen and deciduous trees, grazing ground, water, uncultivated areas, cultivated areas, urban areas, hazelnut areas, burnt areas, pits and sand. The land use was then further classified to identify the areas where livestock occur (3, 10). were taken into consideration. The information on capture covers the period from February 2001 to November Results Based on biological characteristics of C. imicola, the most suitable habitat was defined and a soil map was produced (Fig. 2); information by satellite imaging and possible animal presence (depending on territorial characteristics) were combined and five risk classes were identified, from low probability (score 0) to one of very high probability (score 4) for the presence of C. imicola. Figure 1 LandSat 7 image georeferenced in UTM ED50 From a digital elevation model (DEM) using ArcGis 8.1 software (especially with spatial analyst extension), the following maps were obtained: slope, elevation, aspect and stream network. The maps were reclassified giving the areas significance according to the predicted habitat of Culicoides species. Using meteorological data from 30 weather stations (points on the map with a meteorological value) and performing statistical interpolations, a temperature chart was obtained using the regression method between height and temperature (the correlation value was: ). The humidity and precipitation chart was drawn using the inverse distance weighted (IDW) interpolator to create a grid map with a value for any point on the map (11). All the maps were combined to obtain an initial result. Data on insect captures refer to the monthly average number of Culicoides collected in fixed traps according to the national entomological surveillance plan in Italy. The choice of area for the positioning of traps was made both on the basis of previous capture results and on the areas predicted to have a greater number of domestic ruminants. One fixed trap was positioned every km 2, which is equal to four cells of the grid into which the territory of the Campania Region was subdivided under the BT surveillance plan. Each cell presented a square of 20 km 2. The traps were located on farms chosen by the veterinary services according to the presence of cattle and dependent on the agreement of the owner. The capture of insects was performed once a week between sunset and sunrise and the trapped insects were sent to the reference centre in Teramo to be counted and identified (1). Both the quantity of Culicoides as well as the presence of Culicoides imicola Water, pits, sand Urban and burnt areas Woodlands Uncultivated and cultivated areas Grazing grounds Figure 2 Land use reclassification based on the presence of Culicoides probability The presence of clay soil may also influence the presence and breeding of C. imicola. Therefore, based on DEM, a map was produced emphasising moisture-retentive zones (Fig. 3) where breeding by C. imicola is most likely to occur. This map also takes into account slope, elevation and aspect. Marshland Figure 3 Zones in which swamps can develop 386 Veterinaria Italiana, 40 (3), 2004

341 Temperature, rainfall and humidity maps were also produced (Figs 4, 5 and 6). The mean maximum temperature in the Campania Region is over 12.5 C for many consecutive months, so only mean minimum temperatures for the winter months (November to March) were used. Climatic and territorial classifications were combined to map the probability of presence/absence of C. imicola (Figs 7 and 8) with probability of presence of Culicoides subdivided amongst five risk levels (low, medium-low, medium, medium-high and high). Results from fixed traps in medium and mediumhigh risk areas were mapped (Fig. 8 and Table I). Mean minimal temperatures High: Low: Reclassification of temperatures Figure 4 Mean minimal temperatures in the winter months Low Medium-low Medium Medium-high High Figure 7 Map of risk level >1 000 mm <300 mm mm Figure 5 Rain table Outbreaks Results of insect captures Seroconverted sentinel animals Cell = 20 km Low Medium-low Medium Medium-high High Figure 8 Map of risk level based on insect capture results, bluetongue outbreaks and bluetongue seroconverted sentinel animals <=60% 60-70% >70% Figure 6 Medium value of humidity in recent years Culicoides were captured in all of the traps, but C. imicola was identified at only three trap sites (Table II). Serologically positive sentinels (SPS) and BT outbreaks were also included in the risk map (Fig. 8). SPS were distributed throughout the region but more occurred in the south. Unfortunately, only 139 outbreaks of 251 were georeferenced and Veterinaria Italiana, 40 (3),

342 mapped; all emanated from the south/south-eastern part of the region. Outbreaks and positive sentinels were found mostly in the medium or high-risk level zones (Tables III and IV). About 10% of the outbreaks, and 12% of the positive sentinels, occurred in medium-low risk level areas, while 90% and 88%, respectively, occurred in medium-high or high-risk level areas (Fig. 8). Table I Number of Culicoides trapped according to predicted risk levels Risk level Count Percentage 1 Low Medium-low Medium Medium-high High 0 0 Total Table IV Bluetongue outbreaks segregated by risk level, November 2001-November 2002 Risk level Count Percentage 1 Low Medium-low Medium Medium-high High 0 0 Total Livestock distribution, positive sentinels and BT outbreaks are presented in Figure 9. Table II Insect collections, 2002 Trap ID No. of collections 2002 No. of collections (May- October) Mean no. of Culicoides (May- October) No. of collections positive for C. imicola (November) Table III Positive sentinels by risk area Risk level Count Percentage 1 Low Medium-low Medium Medium-high High 1 1 Total Seroconverted animal Outbreaks Point Point Cattle Sheep/goats Buffalo 1 point = 50 1 point = 50 1 point = 50 Figure 9 Livestock distribution and identification of seroconverted livestock and bluetongue outbreaks Discussion The risk map identified 90% of the Campania Region to be in the medium-high risk zone for presence of C. imicola. Data from traps seem not to confirm the reliability in predicting the presence of C. imicola. However, positive captures of Culicoides in all 11 fixed traps, and finding SPS and BT infection in predicted medium/high-risk areas, suggests reliability of the model (2). Other Culicoides species could be involved in transmission of the disease and this observation has been supported by other studies (5, 7). Moreover, it is possible that the capture methodology needs to be improved (more captures and more traps) to identify low levels of C. imicola (4). The distribution of susceptible animal populations, which was not included, could possibly 388 Veterinaria Italiana, 40 (3), 2004

343 improve this model. Unfortunately, georeferenced farms are not yet available. Analogous methodology should be applied in other regions with different territorial and climatic characteristics to verify the capability of the model to predict the presence or absence of Culicoides species. If the role of other Culicoides species in BTV transmission should be confirmed, the model could be tested in that context. References 1. Anon. (2001). Sistema nazionale di sorveglianza della febbre catarrale degli ovini (Bluetongue) piano di sorveglianza entomologica Allegato I, Parte II, Rev , Baylis M., Mellor P.S., Wittmann E.J. & Rogers D.J. (2001). Prediction of areas around the Mediterranean at risk of bluetongue by modelling the distribution of its vector using satellite imaging. Vet. Rec., 149, Burrough P.A. (1986). Principles of geographical information system for land resources assessment. Clarendon Press, Oxford, Monograph No Calistri P., Goffredo M., Caporale V. & Meiswinkel R. (2003). The distribution of Culicoides imicola in Italy: application and evaluation of current Mediterranean models based on climate. J. Vet. Med. B, 50, Caracappa S., Torina A., Guercio A., Vitale F., Calabro A., Purpari G., Ferrantelli V., Vitale M. & Mellors P.S. (2003). Identification of a novel bluetongue vector species of Culicoides in Sicily. Vet. Rec., 153 (3), De Liberato C., Purse B.V., Goffredo M., Scholl F. & Scaramozzino P. (2003). Geographical and seasonal distribution of the bluetongue virus vector, Culicoides imicola, in central Italy. Med. Vet. Entomol., 17 (4), Gerry A.C., Mullens B.A., MacLachlan N.J. & Mecham J. (2001). Culicoides sonorensis, vectorial capacity, bluetongue virus, infection rate, cattle. J. Med. Entomol., 38, Goffredo M., Conte A., Cocciolito R. & Meiswinkel R. (2003). Distribuzione e abbondanza di Culicoides imicola in Italia. Vet. Ital., 47, Koslowsky S., Staubach Ch., Kramer M. & Wieler L.H. (2003). Bluetongue disease in Deutschland? Risikoabschätzung mit Hilfe eines geographischen Informationssystems (GIS). In Bericht des 25 Kongresses der Deutschen Veterinärmedizinischen Gesellschaft e.v.: Schwerpunktthema Zoonosen, , Berlin, Piscedda S. (1999). Introduzione ai Sistemi Informativi Geografici, Quaderni del Progetto Panda, Ministero delle Politiche Agricole e Forestali, Tomlin D. (1990). Geographic information systems and cartographic modeling. Prentice-Hall, Englewood Cliffs, New Jersey, Ward M.P. & Carpenter T.E. (2000). Techniques for analysis of disease clustering in space and in time in veterinary epidemiology. Rev. Prev. Vet. Med., 45, Wittmann E.J., Mellor P.S. & Baylis M. (2001). Using climate data to map the potential distribution of Culicoides imicola (Diptera: Ceratopogonidae) in Europe. Rev. Sci. Tech. Off. Int. Epiz., 20 (3), Veterinaria Italiana, 40 (3),

344 Vet. Ital., 40 (3), Molecular investigations of orbivirus/vector interactions C.L. Campbell, M.J. McNulty, G.J. Letchworth & W.C. Wilson Arthropod-Borne Animal Diseases Research Laboratory, PO Box 3965, Laramie, WY 82071, United States of America Summary Defining predictors for insect-transmitted virus (arbovirus) disease cycles requires an understanding of the molecular interactions between the virus and vector insect. Studies of orbiviruses from numerous geographic regions have indicated that virus genes are affected by insect population differences. Therefore, the authors have initiated genetic studies of Culicoides sonorensis, isolating cdnas for characterisation of differential insect gene expression, as well as a gene discovery project. Previous work identified insect transcripts elevated in orbivirus-infected female midguts at one day post infection (pi). Here, we report cdnas that were more abundant in midguts two days following an epizootic haemorrhagic disease virus feeding, as well in head/salivary glands at three days pi. Of the cdnas identified in midguts at two days pi, three encode translational machinery components, and three encode components that affect cellular structural features. Of the differentially expressed salivary gland cdnas, only one was homologous to a previously identified gene, a putative odorant binding protein. Keywords Bluetongue Culicoides sonorensis Cytoskeleton Differential gene expression Ribosomal protein subunit Sensory appendage protein Subtractive library Translation Vector. Introduction Bluetongue (BT) virus (BTV) infection of livestock in the United States of America (USA) causes limited clinical losses. However, there is continued concern because of the potential danger of importing exotic virus strains of unknown virulence to livestock in the USA. Mounting virus phylogenetic analyses have resulted in the delineation of orbiviruses into geographic types (2, 6, 7, 19, 22) potentially caused by evolutionary pressure from regional vector populations. BTV gene segments appear to evolve independently in a host-specific fashion suggesting that both invertebrate and vertebrate hosts influence genetic selection (6). To better define possible interactions between orbiviruses, such as BTV and epizootic haemorrhagic disease virus (EHDV), and the insect vector host, the Arthropod-Borne Animal Diseases Research Laboratory has focused on the primary vector species encountered in the USA, Culicoides sonorensis. These studies will provide the basis for comparative genomic studies as well as investigations of specific interactions between insect vector proteins and exotic viruses. Eventually information gathered from these studies should provide insight into risk assessment for importation of exotic vector species or virus strains and provide targets for genetic manipulations to increase virus resistance and the development of new control strategies for interrupting insect-transmitted virus (arbovirus) disease cycles. The authors have chosen several approaches to characterise both genetic and environmental factors that may influence the ability of Culicoides spp. to amplify and transmit an arbovirus, as follows: 1) identification of differentially expressed transcripts in orbivirus-infected midge target tissues most relevant to virus infection 2) a gene discovery project of midguts and salivary glands to identify Culicoid genes for further study 3) characterisation of possible environmental factors that, combined with the Culicoid genetic characteristics, determine vector competence. This paper, however, focuses on the identification of differentially expressed insect cdnas in target tissues and outlines our developing efforts towards a C. sonorensis tissue-specific gene discovery project. The purpose of the differential expression studies was to identify cdnas that might comprise possible barriers to orbivirus infection in Culicoides spp. and to 390 Veterinaria Italiana, 40 (3), 2004

345 develop markers to assess differences between resistant and susceptible vector populations. The primary focus was on those genes expressed early in the infection process because the encoded proteins are likely to be important in affecting virus infection and may include insect gene products that assist in virus replication or during attachment. A previous report from this laboratory described seven of eight midge cdnas that had elevated transcript levels in EHDV-infected midge midguts compared to serumfed controls one day following a virus meal (9). The majority of the cdnas identified in the previous study comprise three major groups, those that were homologous to genes coding for translation machinery components (RPS6, eif3, eif5a), those potentially involved in cellular differentiation (LAR, FZ2) or those putatively associated with actin (LAR, SAC, actin) (9). Based on the general understanding of orbivirus replication in host cells (18), we proposed that host translation factors, such as those translation elongation factors and the ribosomal binding subunit listed above, are recruited for viral replication. We also postulated that the putative actin-binding cdnas and actin might be involved in virus movement such as has been shown for other viruses such as West Nile virus (12). Our previous study has been extended in this paper to describe elevated transcripts in midguts and heads/salivary gland tissues later during the infection cycle to better characterise the progression of orbivirus infection in the female midge midgut. Midguts were analysed for elevated transcript levels at two days post infection (pi). As orbiviruses are expected to escape from the midgut milieu by three days pi (21), we chose to identify elevated transcripts in the distal head/salivary gland tissues during this time. Reverse Northern blot analysis provided preliminary confirmation of cdnas from subtracted libraries designed to enrich for elevated transcripts in tissues of EHDV-infected midges. These analyses resulted in the preliminary identification of 14 transcripts from midguts at two days pi and 4 transcripts from heads/salivary glands at three days pi that are more abundant during EHDV infection. Methods Detection of differential transcript levels Midges were fed an EHDV-containing (6.0 log 10 TCID 50 /ml) meal in foetal bovine serum plus phagostimulants, whereas control midges were fed a similar meal without virus (9). Midges were held for two or three days at 27 C prior to tissue dissection and processing. At two days following the virus meal, midguts were dissected from both groups of midges and total RNA was extracted as previously described (9). At three days following the virus meal, heads with salivary glands from female midges were removed and RNA was extracted. Total RNA was used to prepare subtracted libraries (BD Biosciences- Clontech) and cdna inserts were cloned into PCR TOPO 4.0 sequencing plasmids. Reverse Northern blots (13) using Alk-Phos Direct labelling (Amersham) revealed cdnas in virus-infected midguts or heads/salivary glands, more abundant than those of serum-fed controls. cdnas were sequenced and subjected to tblastx analysis for preliminary identification based on sequence similarity (1). Expressed sequence tag gene discovery project RNA was prepared from either serum-fed colonised C. sonorensis female midguts or from salivary glands removed from 2- to 4-day-old naive females. All tissues were dissected and stored in RNAlater (Ambion). The midgut cdna library was prepared from poly A+ RNA and cloned into psport1 (Life Technologies). The salivary gland library was generated by PCR amplification of total RNA and cloned into pdnr-lib plasmids (Clontech). Results and discussion A total of 279 cdnas were analysed by reverse Northern blot for the two-day-midgut subtracted library. Only those cdnas that showed increased intensity of hybridisation over serum-fed controls, following two independent reverse Northern blots, were analysed further. Of these, 14 cdnas showed a stronger intensity of hybridisation when probed with alkaline phosphatase-labelled EHDV-infected total midgut cdnas versus serum-fed total midgut cdnas (data not shown). These cdnas are listed in Table I. The number of elevated transcripts comprised only 5% of the number of cdnas screened suggesting that orbivirus infection affects a relatively small number of insect transcripts (24, 25). These data indicated that increased transcript levels during EHDV infection of Culicoides comprise a small group of cdnas coding for three major groups of putative proteins: translation factors, proteins affecting cellular structural components and putative effectors of cellular differentiation. Translation factors constitute one group of cdnas more abundant in EHDV-infected midguts at two days pi. Similar to the findings previously reported (9), elevated transcripts at two days pi include two ribosomal protein subunits (RPL32, RPL1) (Table I). Many viruses, including orbiviruses, repress overall host cell metabolism, including transcription and Veterinaria Italiana, 40 (3),

346 Table I Midge midgut transcripts elevated at two days post infection cdna Accession number bp Homology GI number E-value Protein domain 402 AY Gelsolin (GEL) E AY Unknown E -007 Zinc finger, zf-c2h AY Drosophila melanogaster cue-related E -021 EGF-laminin, ferredoxin, LDL_receptor b, 3E Not submitted to 614 Novel GenBank 535 AY BTB domain protein E Not submitted to 237 TRY1-like E -005 GenBank Not submitted to 386 Ribosomal protein L 32 (RPL32) E -044 Ribosomal L32, 3.8E-21 GenBank Not submitted to 661 Novel GenBank 616 AY Clathrin heavy chain E AY eif 3 subunit E -053 Ferredoxin, eif3c N, 6.5E AY Ribosomal protein L 1 (RPL1) E -037 Ribosomal L4, 1.7E AY Tropomysin-like E Not submitted to 327 Novel GenBank 639 AY COX III E -040 cdna complementary DNA bp base pair ( bp indicates cdna fragment length) GI GenInfo identifier no match was identified. Homology indicates the proposed molecular function of the most closely related gene (GenBank accession number) of known function available in GenBank using a tblastx search E-value indicates the confidence level of the homology assignment Protein domains were determined by Prosite or Pfam search; Pfam searches are accompanied by an E-value translation of host mrnas (18). However, drawing a parallel in another virus system, host cells infected with poliovirus exhibit un-repressed translation of ribosomal protein subunits (10) such as RPL32 identified in this study (Table I), in addition to post translational modification of RPS6 (9), identified in our previous report. These parallels in another virus system corroborate the supposition that these proteins are recruited for virus replication; however, further studies are required to confirm this hypothesis. A second category of cdnas enriched in EHDVinfected midguts at two days pi were cdnas encoding proteins that affect cellular structural features. As stated previously, several putative actinbinding cdnas were isolated from midguts one day pi (9). Many types of viruses recruit actin for various steps of viral propagation (reviewed in 14). At two days pi, a cdna encoding the Culicoides homolog of gelsolin was isolated. Gelsolin is an actin filament severing protein (3). Perhaps gelsolin participates in viral particle release from the cytoskeletal matrix during the replicative process. Other cdnas encoding structural proteins include a tropomyosinlike cdna enriched in EHDV-infected midge midguts (Table I). The function of tropomyosin as a contractile fibre does not provide any clues into its possible function in assisting or preventing viral propagation. A Culicoides homolog to clathrin heavy chain was also enriched in EHDV-infected insect midguts (Table I). Although clathrin is known for virus endocytosis (15), research of a closely related orbivirus, BT, suggests, indirectly, that perhaps this molecule could be involved in non-lytic virus release from insect cells. The BT NS3 protein directly interacts with calpactin of the annexin II complex (4). Annexin II complexes associate with clathrin heavy chain in vivo (23), and clathrin heavy chain participates in exocytosis (16, 17). The enrichment of the clathrin heavy chain homolog in EHDV-infected midguts suggests that EHDV may stimulate and use a nonlytic viral egress in insect cells. The last group of cdnas enriched in midguts by EHDV infection is associated with cellular 392 Veterinaria Italiana, 40 (3), 2004

347 differentiation. This includes a Culicoides homolog of the Drosophila cue gene. Although not well studied, the Drosophila cue homolog has predicted epidermal growth factor (EGF)/laminin protein domains that participate in sperm differentiation and maturation (11). The nature of its participation during virus infection is unclear. Nevertheless, the Culicoides cue gene (cdna 518) is the third differentially expressed cdna potentially encoding an effector of cellular differentiation. This follows the earlier expression of cdnas for CsFZ2, coding for the putative Culicoides homolog of the WNT receptor, and CsLAR, encoding a putative temporally expressed cell adhesion molecule (9). Of the head/salivary gland library, 157 cdnas were analysed by reverse Northern analysis, and 4 transcripts from EHDV-infected insects had consistently stronger hybridisation intensities than serum-fed controls following several rounds of reverse Northern analysis (Table II). Only one of the four differentially expressed cdnas, number 818, was similar to any known genes. cdna 818 is most closely related to a putative odorant binding protein that may possess pheromone binding activity (5). This is the first report of increased levels of a cdna encoding odorant-binding protein during an orbivirus infection. This finding lends credence to the idea that behavioural studies of virus-infected insect vectors may provide insight into the evolutionary relationships between viruses and insect vectors. Perhaps EHDV infection increases the host-finding behaviour similar to the previous report of Plasmodium infection increasing the biting rate of mosquitoes (20). Taken together, these data suggest that increased transcript levels during EHDV infection of Culicoides comprise a small group of cdnas coding for three major groups of putative proteins: translation factors, putative modifiers of cellular differentiation and those affecting cellular structural components. When considered in the context of current knowledge of the corresponding homologs in other organisms, collectively, these cdnas are forming a picture of the nature of the gene expression response of the insect to an orbivirus infection. The apparent absence of immune response genes in these screens might suggest a synergistic relationship between Culicoids and orbiviruses. Alternatively, immune response genes may not have been detected for a variety of reasons; for example, they may act transiently or may be regulated post-transcriptionally or post-translationally. To prepare for future studies of the molecular interactions of the arbovirus and vector insect, we have undertaken a gene discovery expressed sequence tag (EST) project designed to increase our repertoire of Culicoides cdnas. The focus is on the primary tissue barriers to vector infection, replication and transmission. The growing dataset of EST sequences from C. sonorensis midgut and salivary glands are being analysed, categorised and annotated. In preliminary annotation, over 874 unique gene alleles have already been identified (data not shown). Table II Midge head/salivary gland transcripts elevated at three days post infection cdna Accession number bp Homology GI number E-value Protein domain 752 AY Novel Anaphylatoxin, VWFC 818 AY SAP1, sensory appendage protein (antennae) E -046 OS-D, 1E Not submitted to GenBank 829 Not submitted to GenBank 476 Novel 437 Novel cdna complementary DNA bp base pair ( bp indicates cdna fragment length) GI GenInfo identifier OS-D insect pheromone-binding family VWFC a type of C-terminal cystine knot no match was identified Homology indicates the proposed molecular function of the most closely related gene (GI number) of known function available in GenBank using a tblastx search E-value indicates the confidence level of the homology assignment Protein domains were determined by Prosite or Pfam search; Pfam searches are accompanied by an E-value Veterinaria Italiana, 40 (3),

348 The 25% proportion of digestive enzymes amongst the individual EST sequences is indicative of the presence of redundant transcripts, as would be expected from midges that had recently received a serum meal (Fig. 1). Digestive enzyme Novel Unknown Unknown Ubiquitin Novel Signal transduction Structural Secretory Enzyme, digestive Immune Enzyme, metabolic Redox Mitochondrial Endocytosis Transcrip./translation Apoptosis Transporter 626 clones to date Figure 1 Expressed sequence tag (EST) pilot gene discovery project showing the relative frequency of various gene categories identified from initial EST sequences cdnas were prepared from serum-fed colony Culicoides midgut mrnas Homology to known genes was determined by tblast search of NCBI public database Unknown genes represent those that are homologous to genes in the public databases with no assigned molecular function Novel cdnas represent those that do not have significant homology to any genes in GenBank Most genes identified in the differential expression studies were not isolated in the EST gene discovery project. Among the 21 confirmed or provisional differentially expressed cdnas identified from midguts at one and two days pi, only RPS6 and translation factor eif4a were also isolated from the serum-fed midgut cdna library. Therefore, the differentially expressed cdnas may represent rare transcripts that would not otherwise be isolated except through a subtractive method as described here. The development of a C. sonorensis tissuespecific EST database and the combined application of modern and classical arbovirology will move BTV and EHDV vector ecology/competence studies into a new era. These studies are presumably defining genetic markers associated with vectorial capacity. Additional markers are also being generated that allow assessment of the environmental factors involved in vector competence (8). When combined, this growing set of reagents should provide opportunities for assessing the risk of virus infection within a population, persistence in a geographical location via overwintering insects, and potential for geographical movement. In addition, novel targets for interrupting the orbiviral transmission cycle could be identified for potential development of therapeutic agents. Acknowledgements The authors would like to thank Shirley Luckhart, N. James McLachlan, Bruce Seal and Ann Powers for their critical review of this manuscript. References 1. Altschul S.F., Gish W., Miller W., Myers E.W. & Lipman D.J. (1990). Basic local alignment search tool. J. Molec. Biol., 215 (3), Balasuriya U.B., Nadler S.A., Wilson W.C., Smythe A.B., Savini G., de Santis P., Monaco F., Zhang N., Pritchard I., Tabachnick W.J. & MacLachlan N.J. (2005). Global isolates of bluetongue virus segregated into region-specific topotypes based on phylogenetic analyses of their NS3 genes. J. Gen. Virol. (submitted). 3. Bearer E.L. (1991). Direct observation of actin filament severing by gelsolin and binding by gcap39 and CapZ. J. Cell Biol., 115 (6), Beaton A.R., Rodriguez J., Reddy Y.K. & Roy P. (2002). The membrane trafficking protein calpactin forms a complex with bluetongue virus protein NS3 and mediates virus release. Proc. Natl Acad. Sci. USA., 99 (20), Biessmann H., Walter M.F., Dimitratos S. & Woods D. (2002). Isolation of cdna clones encoding putative odourant binding proteins from the antennae of the malaria-transmitting mosquito, Anopheles gambiae. Insect Molec. Biol., 11 (2), Bonneau K.R., Zhang N., Zhu J., Zhang F., Li Z., Zhang K., Xiao L., Xiang W. & MacLachlan N.J. (1999). Sequence comparison of the L2 and S10 genes of bluetongue viruses from the United States and the People s Republic of China. Virus Res., 61 (2), Bonneau K.R., Zhang N.Z., Wilson W.C., Zhu J.B., Zhang F.Q., Li Z.H., Zhang K.L., Xiao L., Xiang W.B. & MacLachlan N.J. (2000). Phylogenetic analysis of the S7 gene does not segregate Chinese 394 Veterinaria Italiana, 40 (3), 2004