Distribution of Culicoides in Greece

Culicoides biting midges (Diptera: Ceratopogonidae) were trapped between 1999 and 2004 at 122 locations in mainland Greece and on most of the larger Aegean and Ionian islands, using OVI light traps, in order to determine the distribution and seasonal activity of bluetongue virus vectors and other Culicoides species. Thirty-nine Culicoides species were identified, six of which (C. furcillatus, C. impunctatus, C. paolae, C. pictipennis, C. riethi, and C. scoticus) were identified for the first time in Greece. Two of these (C. impunctatus and C. scoticus) may be of veterinary importance due to their role as vectors of bluetongue virus and related orbiviruses. In addition, C. imicola was detected for the first time in mainland Greece. Biting midges of the genus Culicoides present a considerable nuisance, but they are primarily of interest as vectors of a number of viruses, including bluetongue virus (BTV). In Africa and parts of southern Europe the principal vector of BTV is Culicoides imicola, although up to 30 species of Culicoides worldwide are thought to be capable of transmission (Erasmus 1990, Meiswinkel et al. 2004). Prior to 1998, BTV rarely occurred in Europe outside of Cyprus and the Iberian peninsula (Mellor et al. 2008), although a brief outbreak of BTV-4 occurred on the Greek island of Lesbos between October and December 1979 (Mastroyanni et al. 1981) and antibodies to BTV were detected shortly afterwards in animals on Rhodes (Dragonas 1981). In October 1998, BTV-9 was detected on Rhodes, Kos, Leros, and Samos, in the eastern Aegean Sea near the coast of Anatolian Turkey (Nomikou et al. 2004). This marked the beginning of a prolonged period of BTV incursions into southern Europe, even into locations where C. imicola was absent, which has continued until the present day. More recently in 2006, BTV-8 spread directly into northern Europe, a region where C. imicola does not occur and where the virus continues to circulate. These outbreaks confirmed that Culicoides species other than C. imicola are capable of sustaining outbreaks of BTV in the field (Caracappa et al. 2003, Savini et al. 2005), and several of these are known to occur in Greece (Mellor and Pitzolis 1979). During the incursions into Greece, it became apparent that the risk posed by BTV and related viruses could not be accurately assessed without updating information on the distribution and seasonality of Culicoides species in Greece. Following the detection of BTV on the four Greek islands in 1998, trapping for Culicoides on the islands began immediately using light traps operated twice a week through the winter into 1999. In 1999, the outbreak spread to mainland Greece, and a national entomological surveillance project was initiated to study the distribution and abundance of Culicoides in Greece, which ran until 2005. The principal aim of this project was to determine the Culicoides species present on the Greek mainland and the larger Aegean and Ionian islands, while the secondary aim was to elucidate the seasonality of the confirmed BTV vector (C. imicola) and potential vectors (C. obsoletus and C. pulicaris group). In this paper, we present the entomological findings of this project. Mainland Greece was divided into 59 squares measuring 50 × 50 km (Figure 1) which were sampled for Culicoides between 1999 and 2004. Wherever possible, two farms at least 10 km apart were sampled in each quadrant. Each farm was sampled overnight on two occasions between the months of April and November. Farms where C. imicola was found during this period were sampled for a further two or three nights during the winter months (December-March). Trapping was also performed on Greek islands with substantial cattle, sheep and/or goat populations, namely Crete, Rhodes, Lesbos, Corfu, Kefallonia, Zakynthos, Kos, Limnos, Chios, Samos, Kalimnos, Paros, and Naxos, using the same temporal frequency described above and a number of trapping sites representative of the size of the island. Map of Greece showing trapping locations for Culicoides, indicated with white circles and the 50 × 50 km grid system used for site selection. To sample the farms, Onderstepoort blacklight (UV) suction traps (Venter and Meiswinkel 1994) were used. Blacklight has been demonstrated to be eight to ten times more attractive to vector insects than white light and increases monitoring sensitivity in areas where vector abundances are low (Wieser-Schimpf et al. 1990). The traps are of a robust design and can remain in the field for several years. If an electrical system was not available, modified traps using a 12-volt car battery (with battery replacement every two or three days) were used. A screen with a 2 mm mesh size was placed around the light trap to exclude larger insects. The trap was placed outdoors close to the animals of the farm at 1.8-2 m from the ground, and was operated from dusk until approximately two hours after dawn. The collected insects were placed into a weak solution of detergent, decanted into 75% alcohol, and labelled with the date and location of the collecting site before being transported to the Department of Parasitology at the Institute for Infectious and Parasitic Diseases (Ministry for Rural Development and Food) in Athens for analysis and identification of the Culicoides. Light-traps are less effective at collecting diurnally-active Culicoides species but have fewer weaknesses than alternatives such as chemically-baited traps. Identification was based on training provided during the Culicoides identification course offered at IAH Pirbright, and sample identification was confirmed by consultation with experts in the field of Culicoides taxonomy, Dr. John Boorman, Mr. Rudy Meiswinkel, and Dr. Jean-Claude Délecolle. In regions where bluetongue outbreaks had occurred, a small number of trapping sites were selected for year-round trapping. Samples were collected twice a month and identified using the protocol described above. In the course of this work, 39 Culicoides species were identified (Table 1), six of which (C. furcillatus, C. impunctatus, C. paolae, C. pictipennis, C. riethi, and C. scoticus) were identified for the first time in Greece. Two of these (C. impunctatus and C. scoticus) are believed to be potentially capable of BTV transmission and are discussed further below. C. furcillatus and C. paolae were each collected at one site only, at Skaloti (41.39°N 24.27°E) and Felekas (39.68°N 19,79°E), respectively. The distribution of sites at which C. pictipennis and C. riethi were found is shown in Figure 9. Non-vector Culicoides species discovered for the first time in Greece during the present study at multiple sites. (a) C. pictipennis; (b) C. riethi. Specimens of C. impunctatus, C. pictipennis, and C. paolae are available for consultation at the Institute for Infectious and Parasitic Diseases (Ministry for Rural Development and Food) in Athens (sample identifiers OMOLIO-18-May-2000, PYRGOS-19-June-2003, and FELEKAS/CORFU-October-2004, respectively), while specimens of C. furcillatus, C. riethi, and C. scoticus remain in the University of Strasbourg collection (reference LXXXIII). Due to trap failures, there were a number of gaps in the seasonal data collection, as shown in Figure 2. Seasonal presence/absence of C. imicola, C. obsoletus and pulicaris-group Culicoides at Nea Santa (northern mainland), Omolio (central mainland) and Dimila (Rhodes). See Figure 3 for precise locations. Locations of seasonal trapping sites; Nea Sanda (41.1°N, 25.8°E), Omolio (39.9°N, 22.7°E) and Dimilia (36.3°N, 28.0°E). See Figure 2 for seasonal presence/absence. Culicoides imicola is the main vector of BTV in the Old World and in the Mediterranean Basin. It was recorded for the first time in mainland Greece in autumn 1999, during the bluetongue (BT) outbreak (Patakakis 2004). In all, this species was recorded at 34 of 122 trapping locations across Greece (Figure 4) but was generally absent from northern and western sites. In the central mainland, it was caught between April and December (Figure 2). The largest catch (1,278 individuals) occurred at Dimilia on Rhodes in early September. Crete is now the only large island in the Mediterranean Sea on which C. imicola has not yet been recorded. Distribution of C. imicola in Greece. Culicoides obsoletus was recorded at 110 out of 122 trap locations. This is the most abundant species in the country and exists in a wide and diverse range of environmental conditions (Figure 5). It was particularly abundant in or near forests, and a catch of 4,472 individuals at Skaloti (close to the Bulgarian border) in August 2002 represented the largest number of any single species caught in a single trapping night. Adults are active by March and are present until at least December (Figure 2). Distribution of C. obsoletus in Greece. Culicoides scoticus was recorded only at Arzos, in the northwest region of the country (41.64°N, 26.37°E) in August 2002. This is the first time that this species has been recorded in Greece. Culicoides pulicaris was recorded at 71 out of 122 trap locations (Figure 6). Its absence from the Ionian islands may be due to the fact that trapping on the islands could be performed only during October, when population levels of several Culicoides species are decreasing. This species was generally caught in low numbers, except for a single catch of 1,102 C. pulicaris (out of a total of 3,105 Culicoides) in NW Greece during late July 2003. The species was recorded in Greece from March/April to November (Figure 2). Distribution of C. pulicaris in Greece. Culicoides punctatus was recorded in 86 out of 122 trapping locations scattered across much of Greece, though it was not found on Crete (Figure 7). It was the only species captured in large numbers in March in the northwestern part of the country (Ioannina) and is present from March to December (Figure 2), with peaks of abundance in May and July. Distribution of C. punctatus in Greece. Culicoides newsteadi was recorded in 88 out of 122 trapping locations (Figure 8). This species was captured in large numbers during August in northwestern Greece and during June in the Peloponnese. It is present from March to November. Distribution of C. newsteadi in Greece. Culicoides impunctatus was recorded only at Omolio, in central Greece (39.89°N, 22.66°E), in May 2000. This is the first recording of this species in Greece. The surveillance described in this paper resulted in a number of important findings. Although C. imicola was first detected in Greece in the 1980s, previous Culicoides surveys had indicated that its presence was limited to a few islands in the eastern Aegean (Boorman and Wilkinson 1983, Mellor et al. 1984, Boorman 1986). Our work significantly extends these earlier findings by recording the species at 32 locations across Greece, including many for the first time on the mainland (Figure 4). The survey in 1983 by Mellor et al. (1984), although conducted in only three districts of mainland Greece (Thessalonika, Larissa, and Attica), used three light-traps of two different designs, trapped at multiple sites within each district (including both cattle and sheep farms) and trapped on eight nights in each district, making the sampling effort within these three districts comparable to that in the present study. Despite this, no C. imicola were caught in the 1983 study and C. imicola were caught in all three of these districts during the more recent study, suggesting that our findings reflect a genuine expansion of this important vector species rather than an artefact of the methodology or selection of trapping locations. Molecular analysis of Mediterranean populations of C. imicola have also indicated that the populations in Anatolian Turkey, the Aegean Islands, and mainland Greece are closely related to one another (Nolan et al. 2008). These findings suggest that the distribution of C. imicola in Greece has expanded since the mid-1980s, and the species now seems to occur in much of the eastern part of the country. However, it has not yet been found to the west of the Pindos mountain range, which crosses mainland Greece from the north to the south. Similarly, further south in the Peloponnese, C. imicola has not been found west of the Taygetus mountains. It may be that these two ranges form a barrier to the windborne movement of adult C. imicola into western mainland Greece, but how durable this barrier will be remains to be seen. Culicoides obsoletus has been suspected as a potential vector of BTV for nearly thirty years (Mellor and Pitzolis 1979). More recently, these suspicions have been supported by field and laboratory findings across Europe. In mainland Italy, BTV-2 has been isolated on several occasions from parous, non-blood engorged obsoletus-group midges (De Liberato et al. 2005, Savini et al. 2005), while in Sicily BTV-2 has also been isolated from parous non-blood engorged C. pulicaris (Caracappa et al. 2003). Laboratory competence studies carried out on obsoletus-group and pulicaris-group midges in the UK have confirmed that both are able to replicate BTV to transmissible levels following oral infection (Carpenter et al. 2006a). During the current BTV-8 outbreak in northern Europe, a number of researchers have reported the detection of BTV RNA from pools of the obsoletus group of Culicoides (Mehlhorn et al. 2007, Meiswinkel et al. 2007, Dijkstra et al. 2008). However, results obtained by PCR alone indicate only the presence of viral RNA rather than live virus, and further investigation by virus isolation is required to confirm the epidemiological significance of these results (Carpenter et al. 2009). Nevertheless, these findings are potentially significant, and the abundance of C. obsoletus and detection of C. scoticus for the first time in Greece are clearly a cause for concern. C. imicola and Obsoletus-group Culicoides species have different ecological requirements. A recent analysis (Conte et al. 2007) observed that in Italy the largest catches of C. imicola were not often made in the same areas as the largest catches of Obsoletus-group Culicoides species, but their analysis excluded sites where smaller catches were made. In this study, Obsoletus-group Culicoides species were present at 27 of the 34 sites where C. imicola was present, suggesting that, while sites capable of supporting high densities of both may be rare, sites where the two overlap at lower densities are considerably more common. Finally, it is also possible that C. newsteadi may play a role in BTV transmission as it is closely related to C. pulicaris, but this remains to be confirmed, and although only a small fraction of C. impunctatus adults appear to be capable of developing a fully disseminated BTV infection (Jennings and Mellor 1988, Carpenter et al. 2006b), it is capable of reaching very high densities under suitable conditions such as those existing in western Scotland. These high densities may offset its poor competence levels. The widespread presence of C. newsteadi in Greece, and the detection of C. impunctatus, may therefore also have implications for the transmission of BTV in the country. Information on the species composition, distribution, and seasonality of Culicoides vectors is a necessary prerequisite for understanding the epidemiology of Culicoides-borne pathogens. The results of the survey described here therefore have the potential to significantly improve our understanding of the epidemiology of BT in Greece in recent years. M.J.P. wishes to thank Dr. Despoina Stugiou who helped to identify the insect collections, Dr. John Boorman, Mr. Rudy Meiswinkel, and Dr. Jean-Claude Délecolle for advice on insect identification and Ms. Vaïa Karagianni for technical assistance. This work was funded by EU project QLK2-2000200611.