Assessment and an updated list of the mosquitoes of Saudi Arabia

Background Mosquito-borne pathogens are important causes of diseases in the Kingdom of Saudi Arabia. Knowledge of the mosquito fauna is needed for the appropriate control of the vectors that transmit the pathogens and prevent the diseases they cause. An important first step is to have an up-to-date list of the species known to be present in the country. Original occurrence records were obtained from published literature and critically scrutinized to compile a list of the mosquito species that occur within the borders of the Kingdom. Results Fifty-one species have been recorded in the Kingdom; however, the occurrence of two of these species is unlikely. Thus, the mosquito fauna of the Kingdom comprises 49 species that include 18 anophelines and 31 culicines. Published records are provided for each species. Problematic records based on misidentifications and inappropriate sources are discussed and annotated for clarity. Conclusion Integrated morphological and molecular methods of identification are needed to refine the list of species and accurately document their distributions in the Kingdom.


Background
The Arabian Peninsula (c.3 million km 2 ) includes the Kingdom of Saudi Arabia (KSA), Oman, Qatar, United Arab Emirates and Yemen. The KSA occupies the major portion of the peninsula (c.1.97 m km 2 ), and includes 13 provinces. Mosquito-borne pathogens cause several diseases in the KSA. Malaria has been considered an endemic disease in the country from as early as 1900 [1].
Many studies have focused on the identification and ecology of the mosquito fauna of the peninsula, including the KSA [2][3][4][5][6][7]. The distribution and abundance of mosquito species are influenced by host availability and human activities, changes in land cover and climatic conditions, such as temperature and rainfall [8]. These conditions significantly affect the vectorial capacity of anopheline and other mosquitoes for pathogen transmission. Understanding these factors is essential for disease control.
Mosquito-borne pathogens, including Plasmodium species, dengue virus, Rift Valley fever virus and microfilariae, cause diseases in the KSA [9][10][11]. Correct identification of vector species and knowledge of their distributions and biology are important requirements for disease control.
Historically, mosquitoes have been collected in different areas of the KSA during ecological studies and for construction of morphological keys for their identification. Prior to 1956 [2], little was known about the mosquito fauna of the Arabian Peninsula, with virtually nothing known about mosquitoes in the central and northern regions of the peninsula. Records of mosquito species collected in the KSA are found principally in papers published before the end of the last century [2,4,5,[12][13][14]. Additional species have been recorded in more recent publications, but until now a list of all species of mosquitoes present in the country has not been published. Therefore, the purpose of this work was to produce an up-to-date list of the mosquito species that are definitely known to occur in the KSA, annotated with published collection records.

Discussion
The ecological diversity of the KSA allows multiple opportunities for the evolution of variation in both mosquitoes and their pathogens. Large-scale movements of people, including the annual pilgrimage to Mecca, largescale use of foreign workers and new developmental projects engender environmental heterogeneity and complex epidemiology. The wide variation in ecological settings may complicate the understanding of spatial/temporal characteristics and population dynamics of mosquito vectors of pathogens that cause diseases in humans.
Investigators may be reluctant to conduct ecological studies in the KSA because field work is labour-intensive and time-consuming. Moreover, due to habitat heterogeneity, which in the KSA is often extreme, there are species of only seven mosquito genera in the country. Here we present the first complete list (see above) of the mosquito species that definitely are known to occur in the KSA. In some cases, ecological and morphological studies have shown that some species are members of species complexes. There are members of four Anopheles species complexes in the country, including the An. gambiae, An [70,71]. Anopheles arabiensis is the only member of the complex with a range that extends outside of Africa into the southwestern and western areas of the Arabian Peninsula, occurring in the KSA and Yemen, where it is the major malaria vector [2,3,7,[72][73][74][75]. Most ecological studies conducted in the KSA recorded the presence of An. arabiensis (often as An. gambiae) in different life stages [27,29,30,51]. Moreover, molecular analysis based on the internal transcribed spacer (ITS2) region of rDNA has shown that An. arabiensis (GenBank: KM068071) is the only member of An. gambiae complex present in the KSA [59].
The An. fluviatilis complex is a group of mosquitoes that is widely distributed in southern and southwestern Asia and includes important malaria vectors [76]. The complex includes four species informally denoted as species S, T, U and V [77][78][79][80][81]. Anopheles fluviatilis S is mainly anthropophilic, endophilic and a very competent malaria vector [52]. Species T and U are primarily zoophilic, exophagic and exophilic, and are regarded as poor-or non-vectors of malaria even though they are responsible for malaria transmission in mountainous and hilly areas in Asia, including Iran [80,82]. Anopheles fluviatilis (s.l.) was recorded in the eastern and western regions of the KSA by Mattingly & Knight [2], Wills et al. [31] and Al-Ghamdi et al. [51]. It is not known whether species T or U or both are present in the KSA.
The An. culicifacies complex consists of five sibling species, informally named species A, B, C, D and E, based on cytotaxonomic and molecular evidence [53,76,83]. These species have distinct geographical distributions, blood-feeding behaviour and vectorial capacity for malaria transmission [54,84,85]. The complex is widely distributed from southern China to Iran, the Arabian Peninsula and Ethiopia [86]. Species E is the most important malaria vector in the Indian subcontinent, and is highly anthropophilic and endophilic [87]. The other sibling species have low anthropophilic index or are highly zoophilic, and are considered poor-or nonmalaria vectors in India and Sri Lanka [52,88]. Available data indicate there are at least two species of the complex in Iran [89], but only the presence of species A has been confirmed based on cytogenetic and molecular evidence; the second species may be a new member of the complex [90]. Further study is needed to determine which member or members of the complex are present in the KSA.
The An. subpictus complex consists of four species in India, informally denoted as species A, B, C and D, based on morphological, chromosomal and molecular differences [91,92]. As is the case in Iran [90], it is not known which species of the complex is present in the KSA.
Anopheles stephensi is the main malaria vector in the eastern region of the KSA, but members of the An. fluviatilis and An. culicifacies complexes, which are important malaria vectors in other parts of Asia [2,3,55,74], may contribute to malaria transmission. In Asia, An. stephensi, An. culicifacies (s.l.) and An. fluviatilis (s.l.) consist of different ecoforms that are restricted to rural or urban areas with different degrees of involvement in malaria transmission [54,76,85]. Anopheles stephensi is a major vector in Asia and the Middle East, including, in addition to the eastern KSA, areas of Iran, Pakistan, Afghanistan, India and southern China [87]. It includes three ecological variants (egg phenotypes): the mysorensis, typical and intermediate forms, which are based and distinguished on egg morphology [87,[93][94][95][96]. The typical (type) form is mainly found in urban areas, whereas the intermediate and mysorensis forms are mainly found in rural environments [87]. The blood-feeding behaviour and role of these forms in malaria transmission varies in different geographical regions. In Iran, the type form is highly anthropophilic and a major malaria vector, whereas the mysorensis form is mainly zoophilic and a poor-or non-vector [97]. In contrast, the mysorensis form is an important malaria vector in India [98]. These differences might be due to the type of host available and the presence of certain mosquito genotypes with different host preference and susceptibility to malarial parasites. Molecular study using two different markers (ITS2 and cox1) did not reveal any differences between An. stephensi (ITS2, GenBank: KM052589; and cox1, GenBank: KJ528887-KJ528894) in two distant cities of the KSA (Al-Ahsaa and Dammam) where different phenotypes might be present [59].
Anopheles azaniae was first encountered in Saudi Arabia during the collection of mosquitoes for Rift Valley virus isolation in the Jizan Region [17], and subsequently collected in the same region by Alahmed et al. [18] and Khater et al. [7]. The record of this species in the Riyadh Region [32], which is a very large eastern region that shares its southern border with Yemen, requires confirmation. The record of this species in the more northerly western Al-Madinah Al-Munawarah Region [33] is doubtful and a probable misidentification of either An. dthali or An. rhodesiensis rupicolus. Anopheles azaniae was previously known to occur in neighbouring Yemen [14], where two morphological forms were recently discovered, the typical form and a previously unrecognized form not immediately identifiable as An. azaniae [99]. A single specimen of the atypical form had a mixed human/ bovine blood meal, indicating that females are anthropophilic and should be considered a potential vector of malarial protozoa. It seems possible that further studies in the southwest of the KSA may detect the presence of the atypical form.
A species of the genus Coquillettidia collected in the Eastern Province of the KSA was identified as a species of Mansonia in the study of Wills et al. [31]. According to the online catalogue of the Walter Reed Biosystematics Unit [24], only two species of the tribe Mansoniini are known to occur in more or less arid countries of the Middle East, Cq. (Coquillettidia) buxtoni in Israel and Syria and Cq. (Coq.) richiardii in Iran, Syria and Turkey. With that in mind, it seems likely that the generic identification of Wills et al. [31] was based on keys available at the time, e.g. Mattingly [100], which treated Coquillettidia as a subgenus of Mansonia. In view of the proximity of eastern areas of the KSA to Iran, we are confident that the mosquitoes identified as a species of Mansonia by Wills et al. [31] were most likely specimens of Cq. richiardii.
Many of the occurrence records for Cx. pipiens and Cx. quinquefasciatus in the KSA are problematic. Culex pipiens generally occurs north of a line transecting the KSA between Jeddah in the west and Al-Hasa in the east. Culex quinquefasciatus occurs in the more southerly region of the Arabian Peninsula, with isolated populations of Cx. pipiens occurring in the southwestern mountains [5]. Introgression occurs between the two species where their distributions overlap in central KSA (see figure 3 in Harbach [101]). Both species exhibit considerable variation in the larval and adult stages, and they cannot always be distinguished with certainty. They are only reliably distinguished by characters of the male genitalia; however, their separation is complicated by the occurrence of hybrid forms in the central transitional zone and eastern coastal region of the country. Hybrids are morphologically more variable than the parental species, and are only reliably distinguished from the parental species by the DV/D ratio calculated from measurements based on the form and relationships of the dorsal and ventral arms of the male genitalia [102,103]. It is not possible to distinguish hybrid larvae and females from the parental species. For this reason, some hybrid specimens collected in and around the zone of introgression have undoubtedly been misidentified as either Cx. pipiens or Cx. quinquefasciatus. Larvae and adults collected north of the hybrid zone that have been identified as Cx. quinquefasciatus are misidentified specimens of Cx. pipiens, and those collected south of the zone that have been identified as Cx. pipiens, with the possible exception of those collected in the southwestern mountains, are misidentified specimens of Cx. quinquefasciatus. Future studies based on dissections of male genitalia or DNA sequence analysis of larvae and adults are needed to substantiate species identifications and collection records. Culex quinquefasciatus is an important vector of the microfilariae that cause lymphatic filariasis.
Culex torrentium is listed as a species of the KSA in the online catalogue of the Walter Reed Biosystematics Unit [24] based on a number of reports, including Alahmed et al. [18,19], Al Ahmed et al. [20,28], Kheir et al. [34], Alahmed [35] and Hassan et al. [41]. However, it is obvious that the identification of Cx. torrentium in the KSA by these authors is based on unreliable characters of larvae and adult habitus principally using the keys of Harbach [4,5]. This species can only be distinguished from members of the Cx. pipiens complex with certainty based on features of the male genitalia. Additionally, Cx. torrentium is a Palaearctic species whose distribution only extends south into Turkey and possibly into northern Iran [90,104,105]. As noted by Harbach [5], early records of Cx. torrentium in Iraq are doubtful, and the species is not now listed as present in that country [106,107]. For these reasons, Cx. torrentium is excluded from the list of mosquito species known to occur in the KSA.
Four species of the subgenus Culex of the genus Culex occur in the southwestern Afrotropical Region of the Arabian Peninsula, i.e. Cx. decens, Cx. duttoni, Cx. simpsoni and Cx. univittatus. These species were only known to occur in Yemen [5,108] until they were more recently found in the Asir Region [63], Jizan (Jazan) Region [18] and Najran Province [28], which border Yemen in the southwest corner of the KSA. The records of two of these species, Cx. simpsoni and Cx. univittatus, require confirmation because their identification was based on larvae and adults. They are only reliably distinguished from Cx. sinaiticus and Cx. perexiguus, respectively, based on the morphology of the male genitalia. Records of the two species in more northerly areas of the country are more doubtful.
Since Culex salisburiensis has been known to be present in Yemen since 1953 [108], there is no reason to doubt its occurrence in the adjacent Afrotropical area of the KSA [21,67]. On the contrary, the reported finding of Cx. wigglesworthi in the Asir Region by Al Ashry et al. [67] is certainly a misidentification. The authors collected 4307 Culex larvae, two of which were identified as Cx. nebulosus, three as Cx. perexiguus and two as Cx. wigglesworthi. Larvae were identified using the keys of Hopkins [109], Mattingly & Knight [2] and Harbach [4,5]. The identification of Cx. nebulosus and Cx. perexiguus may be correct as these two species are included in the keys of Mattingly & Knight (Cx. nebulosus) and Harbach (Cx. nebulosus and Cx. perexiguus). Use of the badly outdated keys of Hopkins [109] for African species published in 1952, leading to the identification of Cx. wigglesworthi, is problematic, nonetheless because it is a sub-Sharan species with closest records in southern Sudan, South Sudan and Kenya. Consequently, as it seems likely that the two larvae were misidentified specimens of either Cx. perexiguus or Cx. univittatus, Cx. wigglesworthi is excluded from the list of mosquito species known to occur in the KSA.
Alahmed [35], Alahmed et al. [36] and Khater et al. [7] recorded the presence of Uranotaenia (Pseudoficalbia) unguiculata in eastern and southwestern areas of the KSA. The authors, however, were apparently unaware of the earlier description of Ur. unguiculata pefflyi by Stone [69], and the recognized presence of only that form in eastern parts of the KSA [31]. Consequently, as it is unlikely that the typical form, Ur. unguiculata unguiculata, is present in the country, only the subspecies described and named by Stone [69] is included in the list.

Conclusions
It is obvious from this study that researchers working in the KSA have generally been unaware of the limitations of available keys, have failed to critically evaluate published records and, at least in the case of Culex, have not confirmed identifications with correlated diagnostic features of the male genitalia. Future studies in the KSA must overcome these shortcomings by individually rearing larvae to obtain adults of both sexes for integrated morphological and molecular study, and use the resultant unequivocal molecular markers for the identification of larvae and adults collected in subsequent studies. Based on currently available information, it is concluded that 49 species of Culicidae, 18 anophelines and 31 culicines, are present in various regions of the KSA.