Identification and characterisation of mosquitoes from different locations in Qatar in 2017–2019

Mosquito-borne infections have considerable consequences for public health. The mere presence of a single case of vector-borne disease (VBD) introduces a risk to the local community particularly when associated with the compatible vector, host, and suitable environmental factors. Presently, there is no well-established vector control and surveillance programme in Qatar; therefore, the likelihood of VBDs spreading is undetermined. As a result, there is a pressing need to address this gap and enable successful management of VBDs. This study presents the results of three consecutive field surveys conducted between 2017 and 2019 with the aim of defining the types and distribution of mosquitoes that are of public health importance in Qatar. The results of the adult mosquito trappings show that the southern house mosquito Culex quinquefasciatus is the most widespread and abundant mosquito species, followed by Cx. perexiguus, both species representing a risk of West Nile virus transmission. All sampling methods show that the malaria vector Anopheles stephensi is widespread including in urbanised areas, suggesting a risk of local malaria transmission. The wetland mosquito Aedes caspius is also widespread, representing a risk of Rift Valley fever virus transmission. The dengue vector Ae. aegypti was not detected and can be considered neither widespread nor abundant, suggesting a minimal risk for local transmission of dengue, chikungunya and Zika viruses. Interestingly, the study detected Culiseta longiareolata for the first time in Qatar. Regular field studies are needed to further address the knowledge gaps in terms of distribution, ecology, and biting habits of different mosquito species currently present in Qatar to accurately assess the risk of mosquito-borne diseases.


Introduction
In recent years, the importance of vector-borne diseases (VBDs) has increased at the global and regional levels [29]. Several factors including the rapid growth of the human population, unprecedented urbanisation, increases in movement of humans and animals (travel and trade), and environmental challenges including climate change significantly impact the life cycle, the transmission and the geographical distribution of pathogens [17]. In non-endemic countries such as Qatar, the very first and crucial step in the prevention and control of VBDs requires the identification and appraisal of potential vector populations followed by mapping of the human and animal populations at-risk of acquiring (and transmitting) the pathogen. Currently, being a non-endemic country, the vector control and surveillance programmes were never well established in Qatar. Therefore, the Ministry of Public Health, Qatar, with technical assistance from Eastern Mediterranean Regional Office (EMRO) of the World Health Organization (WHO), have recently assessed the situation of vectors and their respective VBDs in Qatar [24,25]. Analysis of the situation revealed a significant knowledge gap regarding the presence and distribution of mosquito species in different parts of the country, including rural-urban distribution. To address this issue, it was recommended to further strengthen Qatar's technical capacity in the field of entomology, and in particular with emphasis on developing competencies toward vectors identification and surveillance. Subsequently, several field surveys were organised to assess the presence of key species of mosquitoes in different regions of Qatar, together with capacity-building activities.
At the time of the above-mentioned situation analysis, we conducted a literature review which included a total of nine studies, and together reported the occurrence of 20 mosquito taxa (Culicidae) in Qatar (Table 5) [25]. However, the majority of these 20 mosquito taxa were reported by a single publication. Moreover, in these cases, the authors often did not provide findings specifications, including for species described beyond their established distribution range, and thus their presence in Qatar requires further confirmation. Also, two studies reported taxa, i.e. Culiseta sp. and Coquillettidia sp., that remain yet to be identified at the species level. The literature review guided us in identifying the existing gap(s) regarding the distribution of different species of mosquito across various regions of Qatar. Furthermore, entomological reports from many neighbouring countries informed us about the presence of several mosquito species in the Middle East region (e.g. 49 species in Saudi Arabia [4]), which increases the probability of discovering other mosquito species (and sub-species) in Qatar. Therefore, we conducted field surveys to gather accurate and updated data about the presence and distribution of various mosquito species and carried out the risk assessment for mosquito-borne diseases in different regions of Qatar. Here, we report the main findings from the 3 sessions of field survey: (i) a longitudinal monitoring performed between August 2017 and August 2018; (ii) a series of samplings collected during the situation analysis mission, in September 2017; and (iii) a cross-sectional study undertaken in January 2019.

Study area
Qatar (24-26°N, 50-51°E) is a small peninsular country, located on the north-eastern coast of the Arabian Peninsula, Middle East (Fig. 1). The total area of Qatar is approximately 11,600 km 2 and the total population is around 2,750,000 consisting of a large number of immigrants that varies from year to year (75.5% in the year 2015) [5]. Topographically, most of Qatar consists of a flat rocky plain (the highest point is 103 m), with a small range of limestone hills in the North-West and massive sand dunes in the South. The land is comprised of urban areas at 13%, rural areas at 84.5%, and has around 5.7% (670 sq. km in 2016) of agricultural land [5]. The country is divided into eight municipalities. Qatar's climate is classified as a hot desert (Köppen-Geiger category BWh), with an annual mean temperature of 27.1°C and mean rainfall of 72 mm (most rainfall is between October and May) [6]. A series of repeated sampling (longitudinal) sessions were carried out to collect adult mosquito samples from across the country over a period of one year, to account for seasonal data. A total of nine locations were selected across the country to account for different environment sub-types that would influence mosquito breeding such as farms, gardening centres, and zoos (Table 1, Fig. 1).
Adult mosquitoes were collected through MozzTech Mosquito Traps (Ridpest, Malaysia) baited with Octenol and CO 2 that is produced by photocatalytic reaction of titanium dioxide exposed to black light. The traps were set for two consecutive nights each week between August 2017 and August 2018. The mosquitoes caught by this process were collected daily in the morning, and then frozen once transported to the laboratory for sorting and identification under a stereo microscope. To obtain an overview and insight about the mosquito breeding habits in Qatar, five sites previously known to local municipality's pest control workers as common sites for mosquito breeding were inspected for three days (September 18-20, 2017) ( Table 1, Fig. 1). Two strategies were used to collect larval samples: (i) using a net with a fine mesh and then transferring the samples to a 1-L white plastic tray for observation; and (ii) filling the tray by directly dipping it in water. Larvae and pupae collected using these techniques were transferred with water to a vial for transport to the laboratory. There, 4th instar larvae were transferred to a 70% ethanol solution and young larvae and pupae were kept until they grew to 4th instar or emergence of adults. In addition, resting catches were performed by using sweep nets around vegetation, and human landing catches were performed by netting around a person. In both cases, adults were collected from the net via a mouth aspirator and brought to the laboratory.

Session 3: Cross-sectional field study, January 2019
A cross-sectional study was conducted with the aim of updating the pre-existing database of the mosquito fauna of Qatar, for species presence at as many sites as possible. A total of 18 sites were selected across the country for collecting the mosquito samples. These sites were selected to ensure rapid collection and transport of the samples to the laboratory within a one-day trip. These sites covered all possible ranges of environments, e.g. urban building areas, farms, garden centres, industrial areas, sewage lakes, wetlands, worker houses, and zoos (Table 1, Figs. 1 and 2). All the samples were collected between January 15-23, 2019. The choices of sites were guided by municipalities' pest control workers, satellite images and/or visually along roads in the course of journeys. Larval samplings, resting catches and human landing catches were performed at

Mosquito identification Morphology
Mosquito larvae and adults (females and males) were classified as belonging to a species or, if not possible, to a group of morphologically closely related species based on standard identification keys using stereomicroscope [3,7,11,12,23]. Several subsamples of mosquito larvae and adults were preserved in ethanol (larvae and immature exuviae, male genitalia) or pinned in an insect box (adults). Molecular identification by DNA isolation and amplification of the mitochondrial cytochrome oxidase subunit I gene (COI) for Culex sp. or of the ribosomal internal transcribed spacer 2 (ITS2) for Anopheles sp. was performed on only a small fraction of total specimens, as described elsewhere [16,26]. New sequences were deposited in GenBank with accession numbers OL653979, OL654412, OL672837, OL672843, and OL672844. In addition, a rapid polymerase chain reaction (PCR) assay that uses polymorphisms in the second intron of the acetylcholinesterase-2 (ACE2) locus was run for the identification of specimens of the Cx. pipiens complex and possible hybrids [28].

Longitudinal data, 2017-2018
Thousands of mosquitoes were collected in session 1, but the presence of considerable by-catches (attracted by the black light) and the poor quality of preservation did not allow all specimens to be properly sorted and identified. However, to obtain an estimate of sampling outcomes under our time constraints, we performed subsampling and analysed one randomly chosen sample per month and per site.
We analysed 99 samples, yielding detection of seven mosquito species or groups, the most abundant being Culex quinquefasciatus species group (Cx. (Culex) pipiens (Linnaeus, 1758), Cx. (Cux.) quinquefasciatus Say, 1823, and Cx. (Cux.) perexiguus Theobald, 1903, which are almost impossible to distinguish as driedand often damagedadults) detected at all sites, followed by Anopheles (Cellia) stephensi Liston, 1901 collected at four sites (Table 2). No other Anopheles species was detected here. Culex quinquefasciatus gr. was highly abundant almost all over the year, whereas An. stephensi showed medium abundance in Oct-Nov and Jun-Jul (Table 3). A third species, Aedes (Ochlerotatus) caspius (Pallas, 1771), was detected at three sites only and at several periods over the year, but in small numbers. In addition, the species Culiseta (Allotheobaldia) longiareolata (Macquart, 1838) was found at three sites.

Field studies, September 2017 and January 2019
In sessions 2 and 3, a total of 20 sites were surveyed with 6 samples collected in 2017, and 27 in 2019 (Tables 1 and 4). This comprises 20 larval samplings, 2 adult human landing catches, 3 adult resting catches, and 8 adult trappings. Larval samplings yielded 933 larvae and 97 pupae, and entrapped adult mosquitoes accounted for 20 males and 101 females.
Seven mosquito species from four genera were observed: one Aedes, one Anopheles, four Culex, and one Culiseta (Table 5). All seven species were observed at both larval and adult (trapped or reared from immatures) stages, allowing accurate morphological identification. One specimen of An. stephensi (sample Q14b, adult female), two of Cx. perexiguus (samples Q04, adult male, and Q20, larvae) and one of Cx. (Cux.) tritaeniorhynchus Giles, 1901 (sample Q10, adult male) were submitted to molecular identification and obtained COI or ITS2 sequences were compared with vouchers deposited in GenBank. Our An. stephensi sequence showed 100% similarity with specimens from Iran and Iraq; Cx. perexiguus sequences showed 100% identity with specimens from the United Arab Emirates, while the Cx. tritaeniorhynchus sequence showed >99% similarity with specimens from India, all confirming our morphological identification. Specimens of Cx. quinquefasciatus were also submitted to molecular identification. A total of 45 specimens (adults and larvae, 1-6 specimens per sample, from all samples harbouring Cx. quinquefasciatus except Q04 and Q19) were submitted to a PCR targeting the ACE2 locus and all obtained band traces on the gel showed characteristic Cx. quinquefasciatus bands (274 bp). Preliminary genomic analysis also suggested that there is no notable trace of hybridisation with Cx. pipiens in the analysed genomes (Yuki Haba, pers. comm.). Culex quinquefasciatus was clearly the more abundant  of the species, collected at 13 sites among 20 in total (Fig. 3), distributed in all land use categories (Fig. 4), and representing 48% of the collected individuals in total (Fig. 5). The lesser encountered species, Cs. longiareolata, was only found at two sites while all five remaining species were collected from five to eight different sites (Fig. 3). In terms of numbers of individuals, Cx. perexiguus and Cx. (Barraudius) pusillus Macquart, 1850 represented 20% and 18%, respectively, while the four remaining species represented less than 5%. Human landing catches revealed the occurrence of Ae. caspius only, while adult trappings also caught Cx. quinquefasciatus (88% of the caught individuals), Cx. tritaeniorhynchus (10%) and An. stephensi (1%), besides Ae. caspius (1%) (Fig. 5). Comparing the species composition according to land use categories showed that all categories have significant mosquito diversity with at least four species among the seven found here. All species but Cs. longiareolata were found to occur in wetlands, and all but Cx. pusillus in rural habitats. Similarly to Cx. quinquefasciatus, Cx. tritaeniorhynchus and An. stephensi were found in all land use categories (Fig. 4).

Discussion
Highly accurate and up-to-date data about the presence and distribution of various vector species are needed by public health authorities to assess the potential threat and devise effective counter strategies for VBDs. In the present study, three field survey sessions were conducted between 2017 and 2019 with the primary aim of collecting data on geographical, topographical, and seasonal distribution of various species of mosquitoes, in different regions of Qatar.

Field data outcomes
The samples from our entomological survey were collected from various sites to account for different factors that may influence the breeding capabilities and distribution of mosquitoes, including farms, garden centres, industrial areas, sewage lakes and sewage treatment plants, urban building areas, wetlands, worker houses, and zoos. In our survey, one or more species of mosquitoes were found at every inspected location, with the southern house mosquito species Cx. quinquefasciatus showing the widest geographical distribution. This is not surprising as this species is well adapted to breed in a wide range of habitats, from artificial collection of water in man-made containers to natural water bodies [7,12]. Our overall findings were in accordance with the known preferences of the species [7,12]. For example, the immature samples of Cx. tritaeniorhynchus and Cx. perexiguus were collected more frequently from flooded land than artificial containers, while specimens of Cx. pusillus and Ae. caspius were frequently found in wetlands with brackish water. However, we were surprised to find Cx. pusillus in a metallic cistern filled with fresh water. The Cs. longiareolata samples, both adults and immatures, were collected from four different sites. This is the first time Cs. longiareolata specimens were detected in Qatar. Wetlands and rural habitats showed the highest mosquito fauna diversity (six species among seven) in comparison to other habitats such as agricultural land, suburban and urban habitats, which harboured at least four species. All these findings are of public health significance in terms of risk for nuisance or potential for pathogen transmission.

Critical review of the species list
No invasive species were found during our surveys. Despite large scale inspection of many man-made containers located in both urban and suburban habitat, our surveys did not find even a single sample of Aedes (Stegomyia) aegypti (Linnaeus 1762), suggesting that this species is potentially uncommon in Qatar. The occurrence of the yellow fever mosquito, Ae. aegypti, was reported in Qatar in a single reference without providing any sampling details [2]. Nevertheless, the presence of Ae. aegypti in Qatar is hardly surprising, as it is reported to breed in several neighbouring countries. We need to be watchful about its possible import into Qatar by being vigilant at places of entry for goods (port, airport, road crossings). Table 3. Relative abundance and seasonality of mosquito species collected in the longitudinal adult monitoring, monthly, August 2017-September 2018, according to one sample per month per site. One black dot = 1-10 individuals; Two black dots = 11-50 individuals; Three black dots = >50 individuals.

Culex tritaeniorhynchus
Culex pusillus Similarly, an investigation for the possible introduction and presence of another invasive species, the Asian tiger mosquito Ae. (Stg.) albopictus (Skuse, 1894), which also inhabits artificial collection of water (e.g. containers) should be performed. Additionally, authorities need to be especially vigilant since this species is spreading worldwide and is even found in the Middle East (e.g. in Iran, Gulf of Oman coast; [9]). Intense international trade makes its introduction possible, and the local climate looks suitable for its establishment [10]. Two brackish-water wetland mosquitoes are reported to occur in Qatar. The first, Ae. caspius, looks to be widespread in the country based on our findings (Table 5). Previous studies have also reported the presence of these species in Qatar for long time. It is possible that the population of this particular however, it has never been reported from any other country in the Middle East except Iraq and Turkey [2,22]. In addition, Ae. caspius adults show morphological variabilities, which could cause its misidentification as Ae. dorsalis [7]. Therefore, the present study recommends that the presence of Ae. dorsalis should be further studied in Qatar with sample collections, morphological observations and molecular identification. Four Anopheles species are reported to inhabit Qatar (Table 5). The most frequently reported species, An. stephensi, was also observed in our study. While the presence of An. (Cel.) multicolor Cambouliu, 1902 is suggested by two field studies [15,18], the two other species An. (Cel.) culicifacies s.l. Giles, 1901 and An. (Cel.) sergentii (Theobald, 1907) are listed without any field observation data [2,11] and therefore their presence has to be substantiated.
The mosquito species belonging to the genus Culex are the most widespread mosquitoes in Qatar. In the Middle East, the Culex pipiens complex comprises the two forms pipiens and molestus, and Cx. quinquefasciatus [13,22]. However, distinguishing these species by morphology is a difficult task that requires meticulous specimen examination [7]. In our study, all specimens were identified as Cx. quinquefasciatus, including by molecular examination. Several articles on the Qatari fauna refer to the Cx. pipiens complex [1,18,19], while others mention both Cx pipiens form molestus and Cx. quinquefasciatus to occur [14,15]. Therefore, further sampling and molecular examination is recommended to confirm the identity of the Culex pipiens complex members in Qatar.
Culex (Cux.) univittatus Theobald, 1901 and Cx. perexiguus are two other closely related species that exhibit very similar external morphology at all life stages [7]. Both species have been reported in the Arabian Peninsula [12] as well as in Qatar [14,18]. In our study, we identified only Cx. perexiguus, confirmed by molecular identification. As for the pipiens complex, there is unclear morphological differentiation and thus further molecular examination is recommended for specimens attributed by morphology to Cx. univittatus [20]. The presence of Cx. pusillus and Cx. tritaeniorhynchus in Qatar was confirmed by our field studies, whereas five other Culex species reported in the literature were not found viz. Cx. ( [2,12,22], but to date, there has been only a single record in the literature and thus the occurrence of these five species in Qatar remains to be confirmed. Lastly, there is only one official record of detection of Culiseta sp. (under its synonym Theobaldia) [1] and for Coquillettidia sp. in Qatar [21]. The mention of Culiseta may refer to Cs. longiareolata that we report here for the first time, and the presence of Coquillettidia sp. has to be further investigated.

Recommendations to further explore local mosquito fauna
Additional and extended field surveys should be performed at regular interval to provide the most comprehensive   knowledge about the mosquito fauna in Qatar. The most comprehensive strategy would be to undertake a field survey at as many sites as possible throughout the country, covering all kinds of environments and applying various sampling and trapping methods, more intensely during the rainy season but also the rest of year.
While city parks may not provide relevant mosquito fauna data because of their regular treatment by insecticides, wildlife conservation centres and animal holdings are important to investigate. In addition, surveys should focus on points of entry (ports, airports) as well as labour camps and industrial zones for possible alien species introductions. There are chances of discovering previously undetected mosquito species in Qatar given the existence of many other species in neighbouring countries (e.g. 36 species in Saudi Arabia [2]). However, the most pressing priority must be to design field surveys to confirm the existence of the mosquito species reported to occur in Qatar only by a single study/sample (Table 5). A quick way of achieving this could be re-analysis of the already collected specimens preferentially by a third party (providing the samples are preserved by the institutes after completion of field surveys) [13,14,19]. Another way of achieving this would be to sample at the same locations as mentioned by authors in those studies, possibly at the same time of the year.
Besides mapping the mosquito population in Qatar, entomological surveys should also aim to evaluate the risk of mosquito-borne pathogen transmission by collecting data on distribution, abundance, seasonality and biting behaviour of species. Such surveys may focus on (1) Anopheles species as potential vectors of malaria parasites, (2) Ae. aegypti and Ae. albopictus as potential vectors of chikungunya, dengue, and Zika viruses, (3) Ae. caspius as a potential vector of Rift Valley fever virus, and (4) Cx. pipiens complex, Cx. perexiguus, Cx. tritaeniorhynchus, and Cx. univittatus as potential vectors of West Nile virus. Finally, cross-sectional and longitudinal data collections are needed to support the building of mid-and long-term surveillance and control strategies.

Summary outcome and prospects
Our field studies have immensely extended the length, breadth, and depth of Qatar's existing mosquito fauna database. Our field surveys were neither able to confirm nor refute the existence of Ae. aegypti in Qatar; however, given the extensive geographical coverage and length of sample collection, we can confidently say that Ae. aegypti is neither widespread nor abundant in Qatar. This suggests that there is a minimal risk for local transmission of dengue, chikungunya or Zika viruses. The malaria vector An. stephensi is widespread and common, including in urbanised areas, suggesting a risk of local transmission of malaria parasites. The wetland mosquito Ae. caspius is likewise widespread and is probably responsible for biting Table 5. Mosquito taxa reported to occur in Qatar in the literature, with date of first report, our findings, and assessed occurrence status. Black dots = confirmed presence.

Taxon
First report