C. D. Ramsdale¹, B. Alten², S. S. Çaglar² and N. Özer²

1 Varndean Lodge, London Road, Brighton BN1 6YA, England; ² Department of Biology, Faculty of Science, Hacettepe University, Beytepe 06532, Ankara, Turkey

An updated checklist of the mosquitoes of Turkey is presented together with notes and annotations.

The first list of the mosquitoes of Turkey is to be found in Martini’s (1929-31) Die Fliegen der palaearktischen Region. A succession of workers, notably Akalin (1936), Irdem (1940; 1941; 1943), Süyev (1953), Parrish (1959), Gökberk (1961; 1970), Erel (1967), Postiglione et al. (1970; 1972; 1973), Coluzzi et al. (1974), Lane (1982; 1985; 1992) and Alten et al. (2000), have since added to or amended this list as new data became available.

Recent major taxonomic revision affecting genera, subgenera and species of the tribe Aedini (Reinert, 1999a, 1999b, 1999c, 2000a, 2000b, 2000c, 2000d) places western Palaearctic members of this tribe in the genus Aedes (containing the subgenera Aedes, Aedimorphus, Fredwardsius and Stegomyia) or the genus Ochlerotatus (containing the subgenera Finlaya, Ochlerotatus and Rusticoidus). Thus, the 50 currently recognised endemic species of mosquito in Turkey are now placed in the genera Anopheles (10), Aedes (3), Ochlerotatus (15), Culex (13), Culiseta (6), Coquillettidia (1), Orthopodomyia (1) and Uranotaenia (1).

The revised list presented here will undoubtedly lengthen as morphological studies are reinforced by cyto-genetic, iso-enzyme and/or DNA sequencing analysis, and as more field data become available.

Genus Anopheles Meigen, 1818

recorded from Northern Iraq, but not Turkey (Note 7)

Genus Aedes Meigen, 1818

subgenus Stegomyia Theobald, 1901

Genus Ochlerotatus Lynch Arribálzaga, 1891

Genus Culex Linnaeus, 1758

Genus Culiseta Felt, 1904

Genus Coquillettidia Dyar, 1905

Genus Orthopodomyia Theobald, 1904

Genus Uranotaenia Lynch Arribálzaga, 1891

1. First recorded as An. bifurcatus or An. amaurus, An. claviger is widely distributed in Turkey. Examination of larvae from several localities in the Marmara, Aegean, Mediterranean and south-eastern Anatolia regions (i.e. in the hottest parts of the country), and of museum specimens in the collections of the Malaria Institute in Adana, failed to detect An. petragnani, the western Mediterranean member of the Claviger Complex (Postiglione et al., 1972).

2. The Hyrcanus Group comprises a large number of species with a combined distribution covering the Oriental and Palaearctic Regions. It includes several malaria vectors, including some in the western Palaearctic. Though well studied in the Oriental Region and in the Far Eastern Palaearctic (e.g. Reid, 1968; Harrison, 1972, 1973; Harrison et al., 1973, 1990; Xu & Feng, 1975; Takai & Kanda, 1986; Chow, 1991; Baimai et al., 1993), the few investigations made in the western Palaearctic have been fragmentary and inconclusive.

Several western Palaearctic forms differing in certain aspects of adult morphology and behaviour (Livadas & Sphangos, 1941; Ward, 1972; Postiglione et al., 1973; Critescu et al., 1975; Gutsevich, 1976; Ramsdale & Haas, 1978; Ramsdale & Snow, 2000) include the named forms flerowi, mahmouti, marzinovski, mesopotamiae, pictus and popovi currently treated as synonyms (Knight & Stone, 1977). Another, pseudopictus Grassi, 1899 reported from widely separated localities in the Mediterranean Basin and formerly treated as a synonym of hyrcanus, was elevated to species status (Glick, 1992). Because of the dearth of information, Reid (personal communication, 1968) suggested the embracing term ‘An hyrcanus’ be used for all western Palaearctic forms pending the accumulation of data permitting adoption of meaningful specific and/or subspecific names for each taxon.

Laboratory crosses showed that some western Palaearctic forms are reproductively isolated, but that two populations respectively from southern France and southern Anatolia are conspecific (Ross Institute, 1976; 1977). At least two Anatolian populations and two Afghanistan populations are morphologically separable (Postiglione et al., 1973; Ross Institute, 1976; 1977). Furthermore, in these and other countries the taxon is a confirmed or suspected vector of malaria (Zahar, 1974, 1990a, 1990b; Onori et al., 1975; Anufrieva et al., 1977; Ramsdale & Haas, 1978).

In the absence of knowledge of the true status of the different Western Palaearctic forms, behavioural, phenological, ecological and vectorial data continue to be applied to an agglomeration of species, Reid’s ‘An. hyrcanus’. In view of the large numbers of people vulnerable to attack from An. hyrcanus when sleeping outdoors during the hot months, and the current expansion of irrigation for intensive crop production taking place in south-eastern Anatolia, accumulation of knowledge on the vectorial status of each of the hyrcanus species present and of means of distinguishing each is of particular concern to the eastern Mediterranean area.

Old records from Turkey are of the mahmuti (Martini, 1929-31), sinensis (Parrish, 1959), and hyrcanus type forms (Akalin, 1936; Süyev, 1953; Parrish, 1959).

3. As currently understood the Maculipennis Complex includes nine Palaearctic species, atroparvus, beklemishevi, labranchiae, martinius, melanoon, messeae, maculipennis, sacharovi and subalpinus (White, 1978; as modified by Cianchi et al., 1981; Zulueta et al., 1983), the last three of which occur in Turkey.

A consequence of former conflicting taxonomic opinion is that distributional records of individual species from many parts of Eurasia contain anomalous records from outside the normal distributions of the various members of the Complex (Ramsdale & Snow, 2000). With regard to Turkey, early records of messeae were misidentifications of subalpinus (Postiglione et al., 1972, 1973); some records of melanoon, formerly regarded as conspecific with subalpinus, may have been cryptic examples of the latter (females of subalpinus may occasionally lay a batch of dark eggs). In all cases where eggs attributed to melanoon were re-examined, identification proved to have been made on wetted eggs, which obscured the deck patterns. These patterns were found to re-appear as the liquid coating evaporated (C.D. Ramsdale, personal observation).

4. Two morphological variants of this taxon were described as different species, Anopheles marteri from Algeria, and An. sogdianus Keshishian from Tadzikhstan. A third form from Spain, An. marteri var. conquensis Torres Cañamares was assigned to synonymy with the type form by Shahgudian (1956), who additionally reduced sogdianus to subspecies ranking. Reappraisal of the taxon led to sogdianus being assigned to synonymy (Ribeiro et al., 1985). In the absence of chromosomal or isoenzyme analysis, this may be the most appropriate interim conclusion. At the same time, as noted by Shahgudian (1956) and Ribeiro et al. (1985), existing data suggest that the two forms may be allopatric, with the sogdianus form having a more northerly distribution. All specimens examined in Turkey were morphologically similar to Shahgudian’s description of sogdianus (Postiglione et al., 1973).

5. Anopheles pulcherrimus is a marsh breeding species with a distribution centred on the lowlands north and south of the Hindu Kush (Boyd, 1949). It is intermittently recorded from the vicinity of the Halbur Çay, a tributary of the Dicle (Tigris), where these rivers form the borders between the Silopi sector of Turkey, northern Iraq and north-east Syria. This area and the south flowing Khabur River of eastern Syria (which flows south from the Midyat Sector of Turkey to join the Euphrates), represent the current northeastern limits of the distribution of An. pulcherrimus (Christophers, 1920; Leeson, 1950; Abdel-Malek, 1958). The adults are capable of extremely long flights and, at the northern end of the Persian Gulf, have been found on board a ship anchored more than 15 km from the nearest land (Wright, 1918). Until now it is only reported from the Turkish – Iraqi frontier during years when autumn rains are sufficiently heavy to create long lasting marshes suitable for autumn oviposition and larval over-wintering. The growing use of flood irrigation in both Syria and Turkey, and consequent proliferation of potential over-wintering sites, extends the areas receptive to this species, particularly in south-east Anatolia. Anopheles pulcherrimus is a confirmed vector of malaria in parts of Afghanistan, Turkestan and Iraq (Badawy, 1970; Onori et al., 1975; Anufrieva et al., 1977; Sergiev et al., 1993).

6. Anopheles multicolor and An. sergentii were each recorded once only during surveys in the Çükürova (Irdem, 1940, 1941). Their inclusion in Süyev’s and Parrish’s lists do not represent separate records and the presence of neither species has been confirmed in almost 60 years, even though this is the best studied part of the country. As stated by Postiglione et al. (1973) it must be concluded that neither species is part of the endemic fauna of Turkey.

Anopheles multicolor is a halophilic North African desertic species with a distribution extending eastwards through Arabia to western Pakistan. The only records from north of the Mediterranean are from the most arid part of Spain and date from the middle of the last century (Encinas Grandes, 1982). In the countries east of the Mediterranean, there are records from Israel, southern Lebanon, southern Syria, Jordan, southern Iraq and southern Iran (Zahar, 1974).

The distribution of the desertic An. sergentii extends from Pakistan, through Arabia and North Africa to the Canary Islands. It is absent from mainland Europe (Ramsdale & Snow, 2000) but was recorded in the Mediterranean island of Pantellaria (d’Alesandro & Sacca, 1967); the most northerly records in the Eastern Mediterranean are from Israel, Jordan and Lebanon (Zahar, 1974).

7. Although not recorded from Turkey, An stephensi, an urban malaria vector with a distribution extending from India to Iraq and the Gulf States, should be mentioned here. This species extended its range during the past century, and probably also much earlier (Zulueta, 1987). During the early 1900s its distribution extended from the Indian subcontinent along the coastal belt of the Persian Gulf into the southern-most part of Iraq. In 1918 it was not found north of Al Amarah (31°55’N) (Christophers & Short, 1921). In the 1940s and 1950s it was found in several places in the alluvial plain around and north of Baghdad (at ca. 33° 30’N and at altitudes of up to 183m) (Macan, 1950; Pringle, 1954). Observations during the malaria eradication campaign showed that it was established in northern Iraq in the vicinity of Kirkük (35°30’N, altitude 400m) and at a similar latitude at Zannar on the Syrian border (Zahar, 1990). Political unrest has precluded entomological work in this area for a long time and there are no recent data. The arrival of An. stephensi in southern Turkey may not be imminent, but it is worthy of note that Kirkük is within 250 km of the town of Cizre in Turkey, and at the same altitude.

Most Oriental species capable of overwintering do so in the larval stage, which may be an important factor limiting their northern distribution. However, according to de Zulueta & Muir (1969), there is some indication that An. stephensi females may undergo gonotrophic dissociation in the northern extremities of the distribution. It might also be added that, due in great part to human activities, ecological change over the past century has been greater than during the several preceding millennia.

8. Aedes cretinus is recorded from Greece (Crete, Attica, Macedonia) (Edwards, 1921; Samanidou-Voyadjoglou & Darsie, 1993; Samanidou, 1998), Cyprus (Nicosia) (Lane, 1982), Turkey (Antalya, open and wooded coast) (Sahin, 1981; Lane, 1982, 1985; Alten et al., 2000) and Georgia (Black Sea coast) (Gutsevich et al., 1971). Its presence in southern Ukraine (Crimea) was thought probable by Gutsevich et al. (1971), but Gornostaeva (2000) considers its presence in European Russia to be doubtful. It was found to be plentiful in the vicinity of Athens (Samanidou, 1998) and Antalya (Alten et al., 2000). Elsewhere, despite its rather wide distribution it seems to be a rarely encountered species, a factor probably accounting for the paucity of biological information. Larvae have been found in tree holes together with those of An. plumbeus, Oc. geniculatus and Or. pulcripalpis. (Gutsevich et al., 1971).

9. There are no recent records of Ae. aegypti in Turkey. However, the possibility of its re-appearance should not be forgotten. Furthermore, establishment of Ae. albopictus (Skuse) in Italy, Albania and France has obvious implications for other Mediterranean countries, including Turkey. An efficient vector of many arboviruses, Ae. albopictus is already involved in disease transmission in North America (Mitchell, 1995; Mitchell et al., 1998), and its arrival in the Mediterranean inevitably increases receptivity in this region. Though the harsh winters of the Turkish uplands may exclude it from much of the interior of the country, all the rich, populous lowlands are suitable for colonisation.

Aedes albopictus is not Europe’s only recently imported species. Ochlerotatus atropalpus (Coquillett) arrived in Italy by the same route (Romi et al., 1997) and it is possible that other species may take advantage of tyre-borne or other means of transport. Turkey is just as receptive to immigrant species as other countries engaged in international commerce.

10. The Caspius Complex in Europe contains two sibling species (Cianchi et al., 1980), currently designated Species A and Species B, but their separate distributions have not yet been determined. Autogenous and anautogenous forms separable by esterase assay occur in Egypt (Gad et al., 1992), but it is not known if these are behavioural characteristics of different species.

11. The Detritus Complex (Pasteur et al., 1977) comprises two sibling species, Oc. detritus (Haliday), type locality Holywood, County Down, Ireland, and Oc. coluzzii (Rioux et al.), type locality Salins-de-Giraud, Basse-Camargue, France, each with slightly different ecological requirements (Rioux et al., 1998). There is no information on the species present in Turkey.

12. Discussing what was then regarded as the Caspius Group of species, Coluzzi & Sabatini (1968) remarked that the salt marsh Caspius/Dorsalis subgrouping possesses narrow postpronotal scales, whilst in the more specialized sea shore rock pool Mariae and tree hole Pulchritarsis subgroupings these are broad. However, they were of the opinion that scale formation may be a secondary manifestation of ecological adaptations because, overall, the Mariae and Caspius subgroupings exhibit greater morphological affinity. Later, Zavortink (1972) moved related New World tree hole species into a Pulchritarsis Section within the subgenus Ochlerotatus of the genus Aedes. Later, Danilov (1982) moved the Russian species of the Pulchritarsis Complex from Ochlerotatus to the subgenus Finlaya. Comprehensive morphological studies (Reinert, 1999a, 1999b, 1999c, 2000a, 2000b, 2000c, 2000d) have resulted in a major revision of the tribe Aedini in which a revised subgenus Finlaya is placed in the genus Ochlerotatus. Morphological characteristics of these subgenera indicate that both pulchritarsis and nigrocanus should be included in the subgenus Ochlerotatus.

13. Oclerotatus nigrocanus is known only from a holotype male collected in Turkey by Dr Vögel in 1926 and described by Martini (1927). Originally lodged in Hamburg, the holotype is currently in the Natural History Museum, London (Mattingly, 1954). Because identification was made on a single damaged specimen, Minar (1991) considered it a doubtful species, stating that it is probably a junior synonym of pulchritarsis, but did not list it as such. However, after re-examination and comparison with Oc. pulchritarsis, Lane (1992) stated that nigrocanus is patently a separate taxon and accordingly restored its species status.

There is a curious story regarding the data labels attached by Vögel to this unique specimen. One of these gave the site of collection as "zwischen Afyon und Eskischehir im Zug, 1926" (Lane, 1994). These are two western Anatolian stations roughly 110 linear, or 160 rail kms apart on the then important but meandering 1300 km long, Haydarpasha–Aleppo railway). In 1926, before the modernisation of communications, this was the most important of the few railway lines in Turkey. A branch straggling across the plateau from Eskishehir to the then still quite small town of Ankara, and a short spur from Alayunt to Kütahya, were the only connecting lines at that time and passengers coming from these places would have had to change trains (Lewis, 1965). The type locality given by Martini (1929-31) is "Anatolisch Bahn, Türkei". There is no indication of the direction of travel of the train on which this mosquito was a passenger. It may have boarded at any station between the Bosphorus and Syria, giving a type locality about 350-500 km long if travelling from Haydarpasha on the Bosphorus, or about 800-950 km long if coming from the direction of Aleppo.

14. Recorded as diversus by Süyev (1953) and Erel (1967), as diversus var. subtrichurus by Martini (1929-31) and Süyev (1953), as rusticus and as rusticus var. subtrichurus by Parrish (1959).

15. The most comprehensive information about the Culex Univittatus Complex in south-west Asia and southern Europe is that of Harbach (1988, 1999) who, on distributional evidence plus the examination of specimens from Turkey, Greece, and Italy, regards Cx. perexiguus as being the only member of this complex present in the Palaearctic Mediterranean Subregion.

16. A unique record of Cx. quinquefasciatus (Parrish, 1959) is included in the list of Cx. pipiens records. Turkey is outside the normal range of Cx. quinquefasciatus and this old record probably refers to Cx. pipiens. Urban Cx. pipiens (feeding on humans and breeding in hypogeal or other more or less enclosed situations) is the ubiquitous Turkish urban pest mosquito and was the most plausible vector in foci of Bancroftian filariasis transmission in various parts of the country.

17. Culex tipuliformis Theobald is a synonym of Culex vagans Wiedemann, an Oriental species with a distribution extending west to Pakistan and Iran. However, Kirkpatrick (1924-1925) in Egypt erroneously applied this name to Culex theileri Theobald, a mosquito with a wide Afrotropical, southern Palaearctic and northern Oriental distribution. In recording Cx. tipuliformis, Irdem (1939), Süyev (1953) and Erel (1967) were following Kirkpatrick in applying this name to Cx. theileri.

18. Records of the Nearctic Culex apicalis Adams in Europe (mostly pre-1950) refer to the Holarctic Cx. territans Walker (Mattingly, 1953). Culex sergentii Theobald was synonymised with Cx. impudicus Ficalbi by Mattingly (1955). The record of Cx. sergentii (Cx. apicalis) by Erel (1967) obviously refers to Cx. territans.

19. The subgenus Lasiosiphon contains a single species, Cx. adairi (synonym pluvialis), with a distribution extending from Egypt to Equatorial Africa. Süyev (1953) included ‘pluvialis (?)’ in his list of Turkish mosquitoes without any explanation. He was obviously doubtful about the record and did not give any background information. There has not been a confirmatory record in the 48 years since then. Turkey is well outside the normal distribution of this species, and the record may safely be regarded as erroneous.

20. Culiseta subochrea has been variously regarded as a variant or as a subspecies of Cs. annulata, or as a separate species (e.g. Wesenberg-Lund, 1920-21; Edwards, 1921; Stone et al., 1959; Knight & Stone, 1977; Maslov, 1989). The taxon was last elevated to species status by Ribeiro et al. (1977), an action noted by Ward (1984) and which has general acceptance. The biology of the two species is similar and both may be found throughout Europe and south-west Asia. Culiseta subochrea is relatively rare in Britain, where Cs. annulata is a common mosquito (Cranston et al., 1987), but is the more prevalent of the two in Iran (Zaim & Cranston, 1986). It would be surprising, therefore, if Cs. subochrea were not additionally present in Turkey, despite the absence of specific records.

21. Recorded as Mansonia richiardii var. martinii by Martini (1929-31), and as Mansonia richiardii by Süyev (1953) and Parrish (1959). A further Mediterranean species, Cq. buxtoni, is present in Romania (Nicolescu, 1995) and Ukraine (Gutsevich et al., 1971) to the north of Turkey, and in Syria and Israel (Parr, 1943) to the south. Though not yet recorded, this species may be expected to occur in Turkey.