Diversity and Molecular Characterization of Mosquitoes (Diptera: Culicidae) in Selected Ecological Regions in Kenya.

Mosquitoes play a predominant role as leading agents in the spread of vector-borne diseases and the consequent mortality in humans. Despite reports on increase of new and recurrent mosquito borne-disease outbreaks such as chikungunya, dengue fever and Rift Valley fever in Kenya, little is known about the genetic characteristics and diversity of the vector species that have been incriminated in transmission of disease pathogens. In this study, mosquito species were collected from Kisumu city, Kilifi town and Nairobi city and we determined their genetic diversity and phylogenetic relationships. PCR was used to amplify the partial cytochrome oxidase subunit 1 (CO1) gene of mosquito samples. Molecular-genetic and phylogenetic analysis of the partial cytochrome oxidase subunit 1 (CO1) gene were employed to identify their relationship with known mosquito species. Fourteen (14) haplotypes belonging to genus Aedes, nine (9) haplotypes belonging to genus Anopheles and twelve (12) haplotypes belonging to genus Culex were identified in this study. Findings from this study revealed a potentially new haplotype belonging to Anopheles genus and reported the first molecular characterization of Aedes cumminsii in Kenya. Sequence results revealed variation in mosquito species from Kilifi, Kisumu and Nairobi. Since vector competence varies greatly across species as well as species-complexes and is strongly associated with specific behavioural adaptations, proper species identification is important for vector control programs.


Introduction
Mosquitoes are vectors responsible for transmission of numerous pathogens causing diseases such as, malaria, lymphatic filariasis, avian malaria and arboviruses including dengue virus, chikungunya virus, yellow fever virus, West Nile virus, and Zika virus (Charrel et al., 2007;Semenza, 2014). Africa is one of the major hosts of mosquitoes responsible for mosquito-borne viruses (Braack et al., 2018) that are of great medical importance and contribute to the current global public health threat (Enserink, 2007;Gubler, 2002;Higgs, 2014). Seasonal and environmental changes play a role in the global distribution of mosquito species and the arboviruses they transmit (Anyamba et al., 2001;Hasnan et al., 2016). The global spread of vectorborne diseases has resulted in multiple calls on nations to enhance surveillance of emerging arboviruses that requires understanding of the species composition and distribution of potential mosquito vectors (Grout et al., 2017;Kollars et al., 2016).
In the recent past, there has been an increasing spread of mosquito-borne viruses such as chikungunya virus, dengue virus and Rift Valley fever virus in Kenya, thus prompting a need for further research (Johnson et al., 1982;Konongoi et al., 2018). The available literature on mosquitoes in Kenya mainly addresses aspects of morphological identification of mosquito vectors and limited molecular characterization (Lutomiah et al., 2013;Mwangangi et al., 2013). Despite mosquitoes being a key public health challenge in Kenya, little is known about their species diversity and distribution along different ecological zones such as the Kenyan coast and Kenya's capital city. Subsequently, population genetic studies on mosquito vectors in Kenya have focused on the Anopheles genus because of its significance in endemic malaria transmission (Chen et al., 2006;Chen et al., 2004;Lukindu et al., 2018). In addition, most of the studies on mosquito vector composition and diversity are based on mosquitoes confined to a single habitat or with a limited habitat range (Ajamma et al., 2016a;Muturi et al., 2006). The species composition and distribution of Anopheline mosquitoes in Kenya, particularly along the Kenyan coast, have broadly been reported over that of Culicine mosquitoes (Mbogo et al., 2003;Midega et al., 2007;Midega et al., 2010). Moreover, little has been documented on the species composition and diversity of all mosquito groups by use of molecular markers. As such, understanding the species composition and diversity patterns of the suggested vectors is pivotal to the judicious deployment of existing vector control strategies and the development of new effective vector control interventions (Kraemer et al., 2016).
In this study, we employed molecular genetic techniques, involving PCR and sequencing of cytochrome oxidase subunit 1 (CO1) gene to identify and characterize mosquito species in Nairobi, Kisumu and Kilifi Counties in Kenya.

Study sites
This study was carried out at Nairobi, Kilifi and Kisumu Counties in Kenya. Kilifi and Kisumu regions were chosen purposively due to their high abundance of mosquito vectors (WHO, 2017) and vector-borne disease burden, while Nairobi region was selected because it's a major international and domestic destination for both humans and parasites (Wesolowski et al., 2012). Two sampling sites were randomly selected from each of the three regions as follows: Kisumu; Ahero and Kisumu town, Kilifi; Kilifi town and Mazingira Park and Nairobi; Nairobi city centre, and Northern Bypass (Figure 1).

Sampling strategy
The trapping of mosquitoes was carried out in the respective counties during the dry season (January-February 2018) and wet season (March-April 2018). The captures were conducted day and night using the Pyrethrum Spray Catch (PSC) method as used by (Ndiath et al., 2011). The specimens were adult mosquitoes, which were morphologically sorted in the field into their respective genera, and transported in liquid nitrogen to the laboratory for further molecular analysis. A total of 2,438 adult mosquitoes were collected. Of these, 894, 824 and 720 adult mosquito samples were collected in Nairobi, Kisumu and Kilifi respectively. From the overall collection, 300 hundred mosquitoes per county were randomly selected for PCR. A total of 25 sequences per study region were used for phylogenetic and genetic diversity analysis as described in an earlier study (Hale et al., 2012).

PCR analysis
Total genomic DNA was extracted from whole body of individual mosquitoes using the Collins' protocol (Collins et al., 1987) with minor modifications. A DNA homogenizing buffer (containing 0.1 M NaCl, 0.2 M sucrose, 0.01 M EDTA and 0.03 M Tris pH 8) was mixed with a lysis buffer (containing 0.25 M EDTA, 2.5% w/v SDS and 0.5 M Tris pH 9.2) in the ratio of 4:1 to make up the grinding buffer (GB). Each mosquito was homogenized in 100 цL of the GB, using a hand-held pestle homogenizer and incubated for 30 min at 55°C. Into each sample, 14 цL of 8 M potassium acetate (KAc), a deproteinating reagent was added and then incubated for 30 min at room temperature before centrifuging at 13,000 rpm for 15 min to get the supernatant that contained the nucleic acid component. 95% ethanol was used to precipitate the genomic DNA. Centrifugation at 13,000 rpm for 10 minutes was done to obtain the nucleic acid pellet. This was followed by a washing step using 70% ethanol. The DNA pellet was suspended in 100 µL of T.E buffer pH 7.2 and stored at -20°C awaiting subsequent experimental procedures. ProFlex SN 297802057 thermocycler under the following cycling parameters; initial denaturation for 15 min at 95°C, followed by 35 cycles of denaturation at 95°C for 30 sec, annealing at 50°C (Anopheles, Aedes, Culex) for 30 sec, and extension at 72°C for 30 sec, and a final extension at 72°C for 7 min. The PCR products from the amplification of the mitochondrial cytochrome c oxidase 1 (CO1) region of the mosquito after purification using QIAquick® gel extraction kit catalogue number 28706, were shipped for sequencing at Macrogen Inc., South Korea.

Sequence analysis
Resultant mitochondrial cytochrome c oxidase 1 (CO1) sequence chromatograms were edited and visualized using Chromas Lite version 2.6.5. The sequences were deposited in GenBank and accession numbers assigned accordingly. Consensus sequences were aligned using ClustalX version 2 ( Thompson et al., 1997), and visualized using Seaview version 4.7 (Gouy et al., 2010). Unique sequences (haplotypes) were identified using DnaSP version 6 (Librado & Rozas, 2009). Sequence polymorphisms were identified using DnaSP and visualized using Jalview version 2. 10.5 (Waterhouse et al., 2009). DNA sequence divergence was analysed using DnaSP. These unique sequences were compared with reference sequences from other parts of the world, selected to represent the Aedes, Anopheles and Culex genera previously reported and available from GenBank (Benson et al., 2011). Other sequences similar to the study sequences in GenBank obtained using the Blastn algorithm were also included in the analysis. Multiple alignment and comparison of the study sequences and GenBank references were performed using ClustalX. Phylogenetic and molecular evolutionary analyses were conducted using Software for Molecular Evolutionary Genetics (MEGA7) (Kumar et al., 2016). Phylogenetic trees were constructed using the maximum likelihood (ML) method rooted using Lutzomyia longipalpis. The phylogenetic trees were estimated using the best-fit general time-reversible (GTR) model of nucleotide substitution with gamma-distributed rate variation among sites. Bootstrap resampling process (1000 replications) was employed to assess the robustness of individual nodes of phylogeny (only >50% were indicated). The resultant tree was visualized using Dendroscope version 3 (Huson & Scornavacca, 2012).

Phylogenetic analysis
From each study site, 25 CO1 gene amplicons were sequenced for phylogenetic analysis. In total, 14 haplotypes belonging to genera Aedes, 9 haplotypes belonging to genera Anopheles and 12 haplotypes belonging to genera Culex were identified through CO1 sequence analysis. These sequences were deposited in GenBank and assigned accession numbers (Table 1). Sequence analysis revealed a unique Anopheles haplotype (GenBank accession number, MK300230) ( Figure 3). Subsequently, haplotypes of Anopheles gambiae, Anopheles funestus, Aedes cumminsii, Aedes aegypti, Culex pipiens and Culex sitiens were found to be distributed across Kilifi, Kisumu and Nairobi mosquito populations (Table 1).
Diversity indices for the three populations, based on sequenced results were calculated as shown in (Table 2). Average number nucleotide differences (k), nucleotide diversity Pi (π) and haplotype diversity (Hd) varied among the species (Table 2).
Phylogenetic analysis of fourteen (14) Aedes haplotypes from Kilifi and Nairobi with similar sequences based on Blastn (NCBI)  Nyanza-Kisumu, Kenya ( Figure 2). Genetic divergence between study Aedes haplotypes identified in Kilifi and Nairobi and Aedes species they clustered with (sequences of known species obtained from GenBank) was variable (Table 3). There was limited divergence between Aedes aegypti (Accession number KX420485) that has previously been identified in Nyanza-Kisumu, Kenya and study haplotypes MK300216, MK300222, MK300218 and MK300221. Aedes aegypti (Accession number KU380400.1) that has been reported in Nyanza-Kisumu, Kenya before showed limited divergence with study haplotype MK300217. Limited divergence was also identified between haplotype MK300224 and Aedes aegypti (Accession number HQ688297.1) that has been characterized in France. Greater divergence and heterogeneity was observed between Aedes aegypti and study haplotypes MK300225, MK300219 and MK300229. Study haplotypes MK300216, MK300220, MK300223, MK300227 and MK300228 formed a distinct clade with other Aedes aegypti of known identity ( Figure 2).
Phylogenetic analysis of haplotypes with similar sequences to those of known identity showed a clustering of study Anopheles haplotype MK300231 and MK300232 with Anopheles funestus. Notably, they also clustered with Anopheles funestus (Accession number MH299888.1 and KU380404.1) that has been reported in Kilifi and Baringo counties in Kenya respectively ( Figure 3). Study haplotype MK300233, MK300234, MK300235, MK300236, MK300237 and MK300238 clustered with Anopheles gambiae previously isolated in Uganda (Accession number MG753695.1, MG753730.1 and MG753745.1) ( Figure 3). Anopheles haplotype MK300230 formed its own distinct clade. This study haplotype MK300230 may be a new species or novel haplotype not yet described ( Figure 3). Genetic divergence between Anopheles haplotypes identified in Kisumu, Kilifi, Nairobi and Anopheles species from GenBank they clustered with was variable in some haplotypes while others were not variable (Table 4). There was very limited divergence and heterogeneity between Anopheles funestus and study haplotype MK300231 and MK300232. There was no divergence between Anopheles gambiae (Accession number DQ792577.1 and MG753695.1) and study haplotype MK300234. Anopheles gambiae (Accession number MG753695.1) has been identified in Uganda before. Study haplotypes MK300235, MK300233, MK300238, MK300236 and MK300237 showed limited divergence with Anopheles gambiae.
From the phylogenetic analysis, we further established that 12 Culex haplotypes from Kilifi, Kisumu and Nairobi, and similar sequences of known identity based on Blastn (NCBI) showed a clustering of study haplotype MK300240, MK300242, MK300246, MK300247, MK300248, MK300249 and MK300250 with Culex pipiens that have been identified in different regions of the world. Importantly, they clustered with Culex pipiens that has previously been identified in Nyanza-Kisumu, Kenya (Accession number KU380381.1, KU380372.1) (Figure 4). Study haplotypes MK300239, MK300241, MK300243, MK300244, MK300245 clustered with Culex sitiens that was earlier identified in Australia (Accession number MG712559.1) ( Figure 4). Genetic divergence between Culex haplotypes identified in Kisumu, Kilifi, Nairobi and reference Culex species was slightly variable in some species, while other species showed no divergence (Table 5).  Lukindu et al., 2018). The Kilifi mosquito population had the greatest diversity and abundance of mosquito species, possibly due to its geographical position, human activities, and natural climatic conditions.

This study identified
Aedes cumminsii has been morphologically identified in Kenya before (Mwangangi et al., 2006), however, this study reports the first molecular characterization of Aedes cumminsii in Kenya.
Aedes haplotypes between Kilifi and Nairobi populations were    et al., 2017). There is therefore increased diversity in Aedes aegypti species from Nairobi; diversity and spread of Aedes aegypti has been attributed to the increase in arboviral infections (Woolhouse et al., 1997). The diversity of Aedes aegypti in Nairobi could be the result of high population density (Gubler & Clark, 1995), poor sanitation and waste disposal as well as water management (Monath, 1994). The Kilifi population had genetically diverse forms of Aedes aegypti (Table 2). Aedes aegypti is widespread on the Kenyan coast (McDonald, 1977;Teesdale, 1955 (Powell & Tabachnick, 2013) and contribute to diversity of the species. In addition, invasion risk related to human travel has become far more severe (Egizi et al., 2016;Wilder-Smith & Gubler, 2008). Phylogenetic relationship between Aedes species from this study and other Aedes species of known identity from GenBank showed clustering with Aedes cumminsii and Aedes aegypti at a high bootstrap value (>90%) at the defining node on the phylogenetic tree ( Figure 2). However, genetic diversity between Aedes species from this study and those of known identity from GenBank was variable (Table 3).
Anopheles species were distributed across the three study populations Kisumu, Nairobi and Kilifi (Table 1). Anopheles species between Kilifi, Kisumu and Nairobi populations were highly divergent as analyzed using molecular markers. Nairobi had only one haplotype of Anopheles gambiae (Table 1).
Anopheles mosquitoes have also been reported in places where malaria has been eradicated and also in malaria non endemic regions thus increasing the risk of reintroduction of malaria as well as spreading of malaria to new areas (Martens & Hall, 2000   Culex pipiens was distributed across Kilifi, Kisumu and Nairobi population while Culex sitiens was only identified in Kilifi population (Table 1). Culicidae is a large and abundant group that occurs throughout temperate and tropical regions of the world, as well as the peri Arctic Circle (Schäfer & Lundström, 2001). Culex mosquitos are an important vector of the zoonotic infection filariasis. Human filariasis infection is a major public health concern. Approximately 57% of those at risk of infection is in the South-East Asia Region and 37% in the African Region ("WHO | Epidemiology," 2018). Although Culex pipiens is ornithophilic it can also feed on humans and mammals (Reisen et al., 1990) and thus capable to transmit West Nile virus to humans. Culex pipiens (Linnaeus) has been identified as the primary vector of West Nile virus (Turell et al., 2000). Kenyan strain of Culex pipiens has been confirmed to be capable of transmitting West Nile virus and its circulation among humans in Kenya has been detected (Lutomiah et al., 2011;Morrill et al., 1991).  (Table 5). This study identified Culex sitiens in the Kilifi population only, Culex sitiens has been found to tolerate saline waters, in Oman it has been successfully isolated from brackish water (Roberts, 1996). Consequently, parasites such as Microsporidium, Amblyospora have been isolated from Culex sitiens mosquito in Coastal Kenya (Sabwa et al., 1984).

Conclusion
Results from this study demonstrate that mosquito vectors that have been associated to arboviral pathogens are distributed across Kilifi, Nairobi and Kisumu counties. 35 haplotypes belonging to genus Anopheles, Culex and Aedes have been identified, genetic diversity of this haplotypes varies with some genus recording high diversity where's others had low diversity. A potentially new haplotype belonging to Anopheles genus has been identified. This implies further research on genetic characterization of mosquitoes in Kenya for an appropriate vector control and management program across the whole country.

Kentaro Itokawa
Pathogen Genomics Center, National Institute of Infectious Diseases, Tokyo, Japan The study described in the manuscript authored by Makanda et al. describes the genetic diversity of mosquitoes captured in three regions in Kenya using the mitochondrial CO1 gene. They amplified a part of CO1 by PCR and sequenced the amplicon. Their analysis includes phylogenetic clustering of each CO1 haplotype with already registered sequence entries in Genbank and calculation of some basic population genetic parameters such as nucleotide diversity (Pi) and haplotype gene diversity (Hd). The main discovery and conclusions they have drawn out from the results were "A potentially novel haplotype MK300230 was identified in Kilifi population." Anopheles "The first molecular characterization of in Kenya" Aedes cumminsii "higher diversity of haplotypes in the Kisumu population, having detected Anopheles Anopheles and " gambiae Anopheles funestus However, there are doubtful points at each finding they are claiming in this manuscript. Many of those problems seem to stem from species identification merely by sequence without leaving insect specimens for retrospective morphological assessment (they used the whole body of insects for DNA extraction, according to the manuscript). Because of this, it would be extremely difficult to address the concerns I will describe below.

About the "A potentially novel haplotype MK300230" Anopheles
In the phylogenetic tree in Fig 3, the haplotype MK300230 locates outside of the known Anopheles group clade. This is very strange and interesting if that insect individual actually possessed morphological characters that look like . Then I queried the sequence of MK300230 to GenBank database by Anopheles blastn search and I found two notable hits, KY831299.1 (99% identical for 588 bp) and JN298693.1 (97% identical for 658 bp) both of which were registered as sequences obtained from some species of Tipuloidea. Unfortunately, MK300230 would not be a sequence for or even not for a mosquito Anopheles species. Probably, what the authors have actually captured was a kind of crane flies, though no clue to confirm it remains because the whole specimens have gone for DNA extraction.

"The first molecular characterization of Aedes cumminsii in Kenya"
The claim that the insect represented by MK300225 belongs is difficult to accept from Aedes cumminsii 1. 2.

3.
The claim that the insect represented by MK300225 belongs is difficult to accept from Aedes cumminsii the phylogenetic tree in Fig 2; there is only one reference sequence for (MG242484).

Aedes cumminsii
Even if MG242484 was the closest sequence entry to MK30025 in current database, it is a dangerous leap to conclude those two sequences are "haplotypes of same species". With a similar reason, assigning MK300227 and MK300228 to in the Fig 2 and MK300247 to in Fig 4 are also Aedes aegypti Culex pipiens problematic because they are not clustered with reference sequences of each corresponding species at least in those figures.
Because of those issues with ambiguous species assignment of each haplotype sequences, I can not consider the calculated population genetic parameters as valid.
Utilizing DNA sequence for taxonomical identification requires extreme care for the selection of reference sequences to be used. Sadly, the description of species associated with DNA sequences exiting in GenBank database is not always correct, but they are based on various levels of evidence. If one classifies his/her sequence merely based on similarity to a sequence that has a wrong species description and registered this new sequence with a wrong species assignment, this entry can be another source of another false finding by another researcher. At least, the authors should correct the ORGANISM section of MK300230 entry in GenBank which is now saying "Anopheles splendidus".
Preserving specimens is important to avoid many of those pitfalls. For mosquito, there are some useful method like NaOH crude extraction (e.g. Lars Rudbeck & Jørgen Dissing, 1998, BioTequniques 25 (4)), which enable preparing DNA template for PCR from a single leg with only cheap cost.The study described in the manuscript authored by Makanda et al described genetic diversity of mosquitoes captured in three regions in Kenya using mitochondrial CO1 gene. They amplified a part of CO1 by PCR and sequenced the amplicon. Their analysis includes phylogenetic clustering of each CO1 haplotype with already registered sequence entries in genbank and calculation of some basic population genetic parameters such as nucleotide diversity (Pi) and haplotype gene diversity (Hd). The main discovery and conclusions they have drawn out from the results were… "A potentially novel haplotype MK300230 was identified in Kilifi population." Anopheles "The first molecular characterization of in Kenya" Aedes cumminsii "higher diversity of haplotypes in the Kisumu population, having detected Anopheles Anopheles and " gambiae Anopheles funestus However, there are doubtful points at each finding they are claiming in this manuscript. Many of those problems seem to stem on identification of species merely by sequence without leaving insect specimens for retrospective morphological assessment (they used whole body for DNA extraction according to the manuscript). Because of this, it would be extremely difficult to address the concerns I will describe below.
About the "A potentially novel haplotype MK300230" Anopheles In the phylogenetic tree in Fig 3, the haplotype MK300230 locates outside of the known group Anopheles clade. This is very strange and interesting if that insect individual actually possessed morphological characters which looks like . Then I queried the sequence of MK300230 to genbank database Anopheles by blastn search, I found two notable hits, KY831299.1 (99% identical for 588 bp) and JN298693.1 (97% identical for 658 bp) both of which were registered as sequences obtained from some species of Tipuloidea. Unfortunately, MK300230 would not be a sequence for or even not for any Anopheles mosquito species. Probably, what the authors have actually captured was a kind of crane flies, though no clue to confirm it remains if the whole specimens had been used for DNA extraction.
"The first molecular characterization of in Kenya"

Aedes cumminsii
The claim that the insect represented by MK300225 belong is difficult to accept from the Aedes cumminsii The claim that the insect represented by MK300225 belong is difficult to accept from the Aedes cumminsii phylogenetic tree in Fig 2; there is only one reference sequence for (MG242484). Even Aedes cumminsii of MG242484 was the closest sequence entry to MK30025 in the current database, it is a dangerous leap to conclude those two sequences are "haplotypes of same species". For a similar reason, assigning MK300227 and MK300228 to in Fig 2 and MK300247 to in Fig 4 is also Aedes aegypti Culex pipiens problematic because they are not clustered with reference sequences of each corresponding species at least in those figures. Because of those problems, Utilizing DNA sequence for taxonomical identification requires extreme care for selection of reference sequences to be used. Sadly, the description of species associated with DNA sequences exiting in GenBank database is not always correct, but they are based on various levels of evidence. If one classifies his/her sequence merely based on similarity to a sequence that has a wrong species description and registered this new sequence with a wrong species assignment, this entry can be another source of another false finding by another researcher. At least, the authors should correct the ORGANISM section of MK300230 entry in GenBank which is now saying " " which is obviously incorrect.

Anopheles splendidus
Preserving specimens is important to avoid many of those pitfalls. For mosquitoes, there are some useful methods like NaOH crude extraction (e.g. Lars Rudbeck & Jørgen Dissing, 1998, BioTequniques 25 (4)), which enable preparing DNA template for PCR from only a single leg with just a cheap cost.

Suggestion
The author should revise the species assignment for each sequence more carefully and consider using ambiguous descriptions for less confident sequences (e.g. Aedes sp.).

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Yes , which permits unrestricted use, distribution, and reproduction in any medium, provided the original Attribution License work is properly cited.

Laban Njoroge
Invertebrate Zoology Section, The National Museums of Kenya, Nairobi, Kenya Abstract and Introduction: There are a number of grammatical errors in the manuscript. The following are examples in the abstract: 'In this study, we identified mosquito species across Kisumu, Kilifi and Nairobi Counties' -I believe mosquitoes were collected from only few places in the three counties and therefore should read were collected from and not across. There should be a word 'the' before consequent mortality in line 1 of the abstract. Commas are missing in some areas where they should be e.g. in the abstract. '.........Despite reports on increase of new and recurrent mosquito borne-disease outbreaks such as chikungunya, dengue fever and Rift valley fever in Kenya little is known about the genetic characteristics.' A comma after Kenya.
'PCR was used to amplify and sequence the partial cytochrome oxidase subunit 1 (CO1) gene.' This sentence creates an impression that PCR was used for both amplificationa nd sequencing. I believe PCR was used to amplify the genes but sequencing was done using another method. This should be clarified. It is needless to repeat the word haplotypes three types. In the introduction: '.............pathogens causing diseases such as; malaria, lymphatic filariases, avian malaria' -That semi colon after such as should not be there. There is lack of standardization in writing names. e.g. disease names such as Chikungunya, Dengue are started with a capital letter in the main body but with small letters in the Keywords Some scientific names of some species are wrongly written e.g. Aedes cummnisii should be Aedes cumminsii. Culicine mosquitoes not Culcine mosquitoes in some areas and others mosquitos.
Methods: There aren't sufficient details of the methods to allow replication by others. For instance, there is no mention of where the Pyrethrum Spray Catches were done. Was in inside houses, Bus waiting lounges, garages? The reference for PSC collection method should clearly indicate that it is as used by Ndiath et al. 2011. I have rated the study design study as partly appropriate as being a study that targeted diversity, one sampling method that targets only indoor resting mosquitoes was not the best. There is probably a need to justify why only PSC was used and point one method as the reason for the low diversity collected. Again, it is not explained why bus stops in the three counties were preferred as sampling sites. If the idea was to see the contribution of transportation to the mixing of populations, then that didn't come out clearly.
'Only a few in Nairobi ( 'This may contribute to the high susceptibility to dengue-outbreak reported in the region (Baba et al' -Should read this may contribute to their high susceptibility 'The similarities in the genetic composition between the An. gambiae in Kenya and Uganda is most likely due to the proximity of the two countries to one another and the exchanges is more likely over land as opposed to across lake Victoria as claimed in the discussion. This study has indicated high diversity of haplotypes in the Kisumu population' -I do not think Anopheles two species only can be regarded as high diversity. Probably you should use the word higher in comparison with Nairobi and Kilifi. 'The low diversity of species in Kilifi and Nairobi may be attributed to the Great Rift Valley,' -Anopheles there is an abundance of Anopheles especially in Kilifi (see Mwangangi et al 2012 which you have in the references). The problem is the choice of sampling method employed. PSC targets indoor resting mosquitoes only while the highest diversity are found outdoors

Conclusion:
The conclusion is sounding a bit weak and it is more of a discussion than a conclusion. There isn't a strong conclusion about the findings on diversity and molecular characterization of mosquitoes encountered. 1.

3.
Reviewer Expertise: I am a mosquito taxonomist and ecologist. I am not an expert in sequencing and phylogenetic analyses and therefore that bit may require another expert.
I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.