No ethical permit was required for this study. However, there was a focus group discussion with the community and household heads where verbal consent was obtained for mosquito collections in the community after the study aims and objectives were explained. During this research study, we did not perform insecticide spraying, night collections, or human bait for mosquito collection. All mosquitoes were sampled during daytime using electrical aspirators activated with batteries.

Study site description. Benin lies between the Equator and the Tropic of Cancer at latitudes ranging from 6°30′ N to 12°30′ N and longitude from 1° E to 3°40′ E. This country shares boundaries with Togo in the West, Burkina Faso and Niger in the North, and Nigeria in the East. Four main climatic zones are found in the country. The North Sudanese climatic region, which is characterized by one long dry season and a short rainy season, with low relative humidity and rainfall that is the lowest in the country (800 to 1000 mm per year). Large water bodies are found in this region and temperatures are the highest, and could reach 45°C during dry seasons. The second region is the wet Sudanese climatic zone (Atacorian). This climatic region is dominated by hills of up to 800 m of altitude and several small water bodies, which makes the region colder. Annual rainfall ranges from 1200 to 1300 mm per year, the vegetation is partially of wet savanna type, the temperature in this part of the country is the lowest. The third region is the sub-Sudanese climatic region that covers the center of the country and part of the South. This climatic region has one long rainy season and one short dry season. Rainfall is between 900 and 1200 mm, the region is less hilly and the vegetation is of wet savanna type. The fourth region is the southern sub-equatorial climatic region that spans the southern part of the country and extends up to coastal areas of Benin. This region is made up of two rainy seasons and two dry seasons. The relative humidity is high, temperatures are relatively low and the vegetation is a mosaic of coastal, wetlands, forest, and wet savanna type. Several water bodies join together in this part of the country before being channeled into the sea ( Figure 1).

Mosquito sampling. From December 2014 to January 2016, indoor collections of adult female mosquito were made between 06 to 10am in several localities along South-North transect of Benin using four electric aspirators. Mosquito collections were carried out in different localities and the GPS was used to determine the latitude and longitude for each sampled locality. Maps of surveyed sites and the distribution of An. funestus in Benin were developed using recorded latitudes and longitudes. For each surveyed locality, a minimum of 30 rooms were randomly selected for mosquito aspirations. These rooms were selected in a way to cover the various ecologies found in each locality. At least three days were spent in each surveyed site but for localities where An. funestus were found, the number of days was extended to five days to obtain a good number of mosquitoes to be used for various analyses. Aspirated mosquitoes were identified morphologically ⁹ , counted and the total number for each species was recorded. All blood-fed and gravid An. funestus (F ₀) collected inside houses were taken to the IITA insectary in Cotonou (Benin), where they were kept in small cups until fully gravid. The forced egg laying technique described by Morgan et al. ²⁶ was then used to induce female An. funestus to lay eggs. Egg batches and emerging larvae from the same female mosquito were reared together and later pooled with larvae from other females, if these females were found belonging to the same molecular species. Larvae were fed daily with Tetramin™ baby fish food and the water of each larvae bowl was changed every two days to reduce the mortality. The F ₁ adults generated were randomly mixed in cages for subsequent experiments.

Mosquito densities per room (m/r) were estimated during four annual climatic seasons: rainy season, transition from rainy to dry season, dry season and transition from dry to rainy season. This estimation was based on the total number of An. funestus s.l. collected during each season divided by the number of rooms surveyed for mosquito collections in that season. Seasonal variations of An. funestus densities were determined per room in two localities in Benin: the locality of Tanongou in the North (wet Sudanese/Atacorian climatic region) and the locality of Kpome in the South (subequatorial climatic region). Kpome and Tanongou were selected to represent the southern and northern regions respectively, due to the high density of An. funestus recorded in these localities.

For each locality, female mosquito specimens that were morphologically identified as belonging to An. funestus group ⁹ were subjected to DNA extractions using Qiagen DNeasy Kit followed by PCR for species identification, as described by Koekemoer et al. ³¹ .

The Plasmodium infection rate was determined using the TaqMan assay ³² . The reaction was performed in a 10µl final volume reaction containing 1×SensiMix (Bioline), 800 nM of each primer and 200 nM of probes labeled with fluorophores: FAM for detecting P. falciparum, and HEX for P. ovale, P. vivax and P. malariae (P. ovm). P. falciparum sample and a mixture of P. ovale, P. vivax and P. malariae were used as positive controls. The real-time PCR Agilent MX 3005 system was used for amplification with the following cycling conditions: 95°C for 10 minutes for denaturation, followed by 40 cycles of 15 seconds at 92°C and 1 minute at 60°C.

Protocols and standard insecticide treated papers supplied by WHO ³³ were used to test for insecticide susceptibility of An. funestus from selected localities in the northern and southern where there was a consistent number of ovipositing females. These selected localities were Tanongou, northern Benin in the wet Sudanese climatic region (Atacorian region), and Doukonta, southern Benin in the sub-equatorial climatic region. We assessed the susceptibility pattern of An. funestus s.s. from both localities to two insecticides of public health interest: pyrethroids type I permethrin (0.75%) used for insecticide treated nets (ITNs), and organochlorines DDT (4%) used in insecticide residual spraying (IRS). Exposed mosquitoes were fed with 10% sugar solution after 1hr of insecticide exposure after which mortalities were recorded 24hrs post exposure to insecticide treated papers ³³ . The wild population of An. funestus was exposed to non-treated insecticide papers as a control ³³ due to lack of susceptible strains of An. funestus, ( An. funestus FANG). Prior to the experiment, the effectiveness of insecticide treated papers was confirmed by exposing the susceptible strain An. gambiae kisumu to insecticide impregnated papers. WHO criteria were used to determine resistance status with mortality between 98–100% indicating susceptibility, 90–97% potential resistance, and less than 90% resistance ³³ .

To assess the role of L119F mutation in DDT resistance, wild female An. funestus s.s. collected from each selected location were genotyped using the Taqman assay, as previously demonstrated ³⁰ . The reaction was performed in a 10μl final volume containing 1×SensiMix (Bioline, London, UK), 800 nM of each primer and 200 nM of each probe using an Agilent MX3005P machine. The following cycling conditions were used: 10 min at 95°C, 40 cycles of 15s at 92°C and 1 min at 60°C. Two probes labelled with fluorochromes FAM and HEX were used. The FAM was used to detect the mutant allele, while the HEX detected the wild type allele.

MedCalc easy-to-use online statistical software ³⁴ was used to test for significant difference of Plasmodium infection rate and L119F-GSTe2 genotyping data in the South compared to the North of Benin.

Out of the 46 surveyed localities ( Figure 1 and Supplementary Table 1) in this study, An. funestus species were found in eight localities, generally in sympatry with An. gambiae, and spread in two geo-climatic regions of Benin. In addition, most of the sites where An. funestus species were collected were found in the western part of the country (six out of eight localities with An. funestus; Figure 1).

A total of 3179 mosquitoes belonging to different species were caught during this survey. These mosquito populations from indoor collections were dominated by Anopheles spp. 82.89% (2635), followed by Culex spp. 14.90% (474), Mansonia spp. 1.38% (44) and Aedes spp. 0.81% (26) ( Supplementary Table 1). Out of the morphologically identified Anopheles spp., An. gambiae s.l. constituted 79% (2083), followed by An. funestus s.l. with 21% (552). No other Anopheles species was collected during the sampling period.

An. funestus was not found in either the dry Sudanese climatic region (no An. funestus collected in the 11 surveyed localities), nor in the transition region between the Sudanese and the sub-equatorial climatic regions (the sub-Sudanese climatic region), where no An. funestus was found in the nine surveyed localities ( Figure 1 and Supplementary Table 1). All An. funestus samples collected were either from the southern sub-equatorial region ( An. funestus found in five out of the 17 surveyed localities) or the northern wet Sudanese climatic region of the Atacora (three localities with An. funestus out of nine surveyed). It is worth indicating that most sampled specimens of An. funestus were found in the western part of Benin (six localities out of eight with An. funestus). Out of the 552 morphologically identified An. funestus sampled during this survey, 319 samples were from localities situated in the sub-equatorial climatic region and 233 from the wet Sudanese climatic region (Atacorian region). High densities of An. funestus were recorded in Kpome (243 An. funestus) and Tanongou (229 An. funestus), localities from the sub-equatorial climatic region and the wet Sudanese climatic region, respectively ( Figure 1).

Generated data from Kpome during the four monitored seasons revealed a higher An. funestus density per room (m/r) during the transition period from dry to rainy season (3 m/r). The lowest number of An. funestus (0.2 m/r) was recorded during rainy season. Densities of 1 and 2.3 m/r were documented during the transition from rainy to dry season and the dry season, respectively. A similar trend was observed in Tanongou with a higher density of An. funestus recorded during the transition from dry to rainy season (1.3 m/r), followed by the dry season with a density of 0.4 m/r, the transition from rainy to dry season and the rainy season had densities of 0.2 and 0.1 m/r, respectively. Comparative analysis of An. funestus densities at Kpome and Tanongou revealed a relatively higher rate of An. funestus mosquitoes per room at Kpome throughout all the four identified seasons compared to Tanongou ( Figure 2).

PCR species detection of the 552 morphologically identified An. funestus individuals revealed a predominance of An. funestus s.s. in the two climatic regions where An. funestus was found in Benin. In the wet Sudanese climatic region, and more specifically in Tanongou, An. funestus s.s. was found in sympatry with its sister species An. leesoni. Out of the 229 An. funestus s.l. aspirated indoors at Tanongou, 178 were An. funestus s.s. and 51 were An. leesoni. In contrast, in the southern locality of Kpome where the highest density of An. funestus was recorded (243 An. funestus s.l.), as well as Doukonta and Pahou, no other member of the group apart from An. funestus s.s. was found ( Table 1).

Taqman results (n=552) showed that An. funestus mosquitoes from the sub-equatorial climatic localities of the southern Benin were significantly infected with Plasmodium compared with those from the wet Sudanese localities of the northwestern Benin (Atacorian region) (P=0.0001). An. funestus from Kpome, Pahou and Doukonta in southern Benin had Plasmodium infection rates of 18.51, 15.78 and 13.33%, respectively. However, in northwestern Benin, only An. funestus s.s. from Tanongou was infected with Plasmodium with an infection rate of 5.62% ( Table 2). Plasmodium infection was absent in all the 51 An. leesoni specimens analysed during this course of research ( Table 2).

Insecticide susceptibility tests of An. funestus s.s. from Doukonta, Pahou ⁶ and Kpome ⁷ in the South, and Tanongou in northern Benin were assessed. In total, 100 females each (F ₁ generated from F ₀ oviposition) of An. funestus s.s. from Doukonta were exposed to DDT and permethrin. Similarly, 100 An. funestus s.s. from Tanongou were exposed to permethrin and DDT. Results revealed low mortalities to DDT (8±0.5%) and permethrin (11±0.5%) for An. funestus s.s. from Doukonta, whereas the Tanongou population had higher mortality rates to DDT (90±3.18%) and permethrin (100%). This shows that there is a higher resistance in Doukonta compared to Tanongou ( Figure 3). Similarly, high resistance levels have been previously documented in southern localities of Pahou and Kpome ^{6, 7} .

Genotyping of the L119F-Gste2 mutation in wild An. funestus population from each of the selected locations revealed the presence of the resistant 119F allele at a high frequency: 96% in Kpome ⁷ , 83.2% in Doukonta (southern Benin), while in Tanongou (North Benin), 35% mutant allelic frequency was recorded. No susceptible allele (SS) was observed either in Kpome or Doukonta mosquitoes, showing that the 119F gene is close to fixation in the An. funestus populations of these two locations in the southern Benin. A significant difference (P≤0.0001) was observed between the 119F allelic frequency recorded in Kpome and Doukonta, where a high resistance to DDT was observed compared to Tanongou ( Figure 4).

An. funestus was mainly found in the southern and the northwestern localities of Benin in this study. In these two geo-climatic regions, there seem to be a high tendency of this species to colonize the western areas of the country (north and southwestern). The relatively high presence of this vector in the western part of Benin could be explained by the humidity, relatively low temperatures associated with the hilly landscape, and the presence of rivers and streams covered with vegetation ³⁵ . This shows that this species prefers more permanent water bodies with vegetation usually found along rivers, streams and lakes ³⁶ , whereas An. gambiae tends to oviposit in temporary breeding sites, such as puddles and animal foot prints ³⁷ . Very little or no population of An. funestus was found in the dry Sudanese climatic region of northeastern Benin. The low presence of this mosquito species in this dry hot region (low rain falls and temperature reaching 45°C during dry seasons) is either due to the period of sampling or the low presence of permanent fresh water bodies covered with vegetation coupled with dryness of the region ³⁸ .

The density of An. funestus species collected indoor in this research further confirms their endophilic behavior ³⁹ . Two species of the An. funestus group were identified during this study: An. funestus s.s. and An. leesoni. Contrary to An. funestus s.s., there was no trace of Plasmodium DNA in the 51 samples of An. leesoni analyzed. This result confirms the low/no implication of An. leesoni in the transmission of malaria, as previously documented ^{14, 39} , which is notable in West Africa as this species is known to be highly zoophilic. While placing a low epidemiological interest on An. leesoni, this study further highlights the need for a high focus on An. funestus s.s. for improved control of malaria in Benin ⁷ . Recorded infection rates of An. funestus were more than three times higher in screened localities of southern localities (Kpome, Pahou, Doukonta) compared to the North (Tanongou), suggesting a higher implication of An. funestus in malaria transmission in the southern part of the country where its density is also high. The high Plasmodium infection rates observed in southern Benin are similar to some infection rates documented in several African countries in this species; Plasmodium falciparum infection rates of 22 ¹¹ and 27% ¹² have been found in An. funestus populations of South Africa. In countries from the western part of Africa, a mean rate of infectivity between 3 and 15% has been observed, including in Burkina Faso ^{14, 40} and recently in Ghana ⁴¹ . In Burkina Faso, Dabire et al. ⁴⁰ documented the presence of Plasmodium in An. funestus (20% infection rate) from Lena during the month of August 2000. In Benin, two studies recently conducted in southern localities revealed Plasmodium infection rates of 13.6 and 18.27% in An. funestus ^{7, 42} . This study has shown a similar trend in the densities of An. funestus in both screened ecological zones throughout the year. High densities of An. funestus mosquitoes were recorded during the transition from dry to rainy season followed by the dry season, then the transition from the rainy to dry season and finally the rainy season, where the least density of An. funestus were recorded. The involvement of An. funestus in the transmission of malaria during dry seasons was also documented in Ghana ⁴³ , Nigeria ¹⁵ , Burkina Faso ^{14, 40} , and more recently in southern Benin ⁴⁴ .

Comparative analysis of insecticide resistance profiles in An. funestus populations from Doukonta (southern Benin) and Tanongou (northern Benin) reveals that An. funestus s.s. from Doukonta are relatively more resistant to DDT and permethrin (mortality rates of 8±0.5 and 11±0.5%, respectively) than those from Tanongou, where only a moderate resistance was observed to DDT (mortality rate of 90±3.18%) and a full susceptibility to permethrin (100%). High resistance to DDT and permethrin had previously been reported in populations of An. funestus from two other localities of southern Benin, Pahou and Kpome ^{6, 7} . In addition to the use of agricultural insecticides in both the northern and southern surveyed sites, the high insecticide resistance observed in the South could be associated with environmental factors, such as urbanization, which increases the level of xenobiotics (pollution) in Anopheles breeding sites and could favor the selection of cross resistance to permethrin and DDT in southern Benin compared to northwestern Benin with less urbanization and pollution ⁴⁵ . Recorded resistance profiles could also be associated with a relatively high flow of genes among An. funestus populations in southern Benin compared to the North, particularly if there are some barriers to gene flow, which needs to be investigated further. Other factors of resistance selection, such as the relatively high use of ITNs/IRS (use of public health insecticides) in the southern Benin compared to the North, might have also contributed to observed high resistance profile of mosquitoes ^{6, 46– 50} . Similar observations have been documented on An. gambiae s.l. in the North and South of Benin where increased pyrethroid resistance is also prevalent in An. gambiae s.l. species in South Benin ^{51– 53} than in the North, mirroring the pattern that was observed here for An. funestus. Resistance to DDT and permethrin is also widely distributed in An. gambiae in Benin ^{4, 54} .

The high frequency of the 119F-GSTe2 resistant allele in Kpome and Doukonta where high phenotypic resistance to DDT was also observed; both results suggest that this mutation plays an important role in DDT resistance in West Africa, as previously documented ³⁰ . Indeed, consistent frequencies of this resistance allele were also recorded in other DDT resistant populations in Central and West Africa notably in Cameroon (52%), Ghana (44%) and Burkina Faso (25%) in accordance with the previously reported prevalence of DDT resistance in these countries ^{27– 29} . The resistant 119F allele was detected in An. funestus populations from Tanongou, but with a relatively low frequency (35%), reflecting the moderate level of DDT resistance recorded. This result is in line with the detection of low frequencies of this resistant allele in the eastern African An. funestus of Uganda (20.4%) and Kenya (7.8%), which is associated with a moderate level of DDT phenotypic resistance observed in this region ^{19, 26} . However, this observation is different in southern Africa where this mutation is completely absent despite recent reports of DDT resistance ²⁵ , suggesting that DDT resistance in southern Africa is driven by a different mechanism to that observed in West and Central Africa. These heterogeneities in L119F frequencies suggest that there are different mechanisms responsible for the DDT resistance in An. funestus populations across Africa.