Bionomic Characteristics of Major Vector Species and Malaria Transmission in Senegambian Border Areas in Senegal

1. Introduction

Malaria remains one of the leading causes of hospitalization and death, making it one of the deadliest parasitic diseases. In Senegal, malaria is therefore a key priority of the national malaria control programme [1]. Consequently, several efforts have been made in term of vector control throughout interventions such as long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS), the two main tools most use in the context of the country. In terms of disease prevention, these tools have made a significant contribution to disease control through their widespread deployment in targeted areas [2]. These current core malaria vector interventions, both of which primarily target Anopheles vectors that feed and rest indoors [3]. In addition, insecticide-based IRS and LLINs have selected for interventions-based evolutionary adaptations that enable the vectors to be refractory to the interventions hence insecticide resistance acquired or to escape from these insecticide-based interventions. The main adaptations observed relate to insecticide resistance, as well as shifts in biting and resting behaviours [4,5].

These adaptations or, more importantly, shifts in vector behaviours mean that we are increasingly seeing vectors biting and resting outdoors or in outdoor shelters [6,7]. This situation poses a real problem for health control programmes, as the tools at their disposal do not enable them to reach these targeted vectors. These vectors are therefore, responsible for the risk of malaria transmission to outside or, more importantly, for residual transmission [8]. In Senegal, malaria is transmitted largely by species such as An. gambiae, An. arabiensis, An. coluzzii and An. funestus [7,9]. Species such as An. pharoensis, An. nili and An. melas also play a minor role in transmission at local level [8,10].

Entomological surveillance and field observations instead suggest heterogeneous environmental conditions; therefore, distribution of these species varies considerably from the north to the south of the country. Species of the An. gambiae complex are found throughout the country. Both An. gambiae and An. coluzzii have been reported in the country, with a predominance An. coluzzii in the north and An. gambiae in the south and the center. In most cases, An. arabiensis coexists with An. coluzzii in the north and with An. gambiae in the south, these species are sympatric almost everywhere across the country, with relative abundances varying according to eco-climatic zones [11]. As for An. melas, a species dependent on brackish water, it is mainly found in littoral and coastal areas and in mangrove ecosystems. Its range extends from north to south and inland into all areas where salt water is present, particularly along certain watercourses [12]. Anopheles funestus has a more restricted distribution compared with the major species of An. gambiae complex. The presence of An. funestus in the Lower Senegal River Valley, north-west of the Lower Ferlo Valley, in the Lake Guiers area, appears to be linked to environmental changes resulting from hydro-agricultural development and crop irrigation. In areas where it coexists with An. gambiae s.l. species, the late emergence of An. funestus and the availability of its breeding sites during the dry season enable it to ensure the persistence of malaria transmission [10].

The proliferation of these species in breeding sites is influenced by rainfall. The peak of malaria transmission is clearly concentrated in the period from September to November, coinciding with the end of the rainy season, a very hot and humid period [13]. During this period, transmission is primarily carried out by An. gambiae, An. coluzzii and An. arabiensis, whilst during the dry season, An. funestus takes over. In areas where market gardening and irrigated farming are common, it is possible that, outside the rainy seasons, vector population densities may be high, particularly for An. gambiae at the start of the flooding period and An. funestus as vegetation establishes itself, which may contribute to prolonged malaria transmission during the dry season [14,15].

The specific composition of these vectors in the southern region, particularly in the Senegambian region, varies, especially depending on the season. Furthermore, the transmission drivers are not well understood, and are associated with high malaria incidence rates in the Senegambian region. In view of this situation, and to tackle these vectors in the Senegambian region, the National Malaria Control Programmes (NMCPs) of Senegal and The Gambia have drawn up a memorandum of understanding for mass distribution of LLINs at the border [2].

In this context, the study aimed to determine bionomic characteristics of major vector species and malaria transmission in Senegambian cross border, with goal of gaining a better understanding of the entomological factors involved in malaria transmission alongside the Senegambian border. This study formed part of a broader operational entomological assessment conducted at four sentinel sites along the Senegambian border, combining vector bionomic characterization with human behavioural observations to better understand determinants of residual malaria transmission.

2. Materials and Methods

2.1. Study Areas

The study was conducted in Nemataba and Medina Dianguette in Velingara district, and in Ndrame Ndimb and Paoskoto in Nioro district from July 2023 to March 2024 (Figure 1). The study sites were well described in Diop et al. [16].

2.2. Mosquito Sampling and Laboratory Processing

Host-seeking and indoor resting mosquitoes were sampled bimonthly from July 2023 to March 2024 using human landing catches (HLCs) and pyrethrum spray catches (PSCs), respectively. For HLC activities, three households were randomly selected as capture points where one room was used at each point. Capture sessions took place in two consecutive nights, between 8 p.m. and 6 a.m., simultaneously involving two capture operators one placed inside the room and the other outside. At each capture point, two capture operators worked in pairs between 8 p.m. and 1 a.m., before replaced by another pair working during the second half of the night, i.e. from 1 a.m. to 6 a.m. In order to minimize biases linked to the differential attractiveness of the operators, the latter were rotated by alternating pairs between the time slots (8 p.m./1 a.m. and 1 a.m./6 a.m.) and the capture points. A supervisor was appointed to ensure that collection activities were correctly conducted, by monitoring work instructions and offering coffee to the capture operators to keep them from dozing off. Collected mosquitoes were counted and identified using conventional keys [17]. A subset of host seeking females (486 An. gambiae s.l. and 148 An. funestus) was dissected to determine the physiological age [18].

Indoor resting females were collected using PSC in ten randomly selected rooms in each study site, from 7 a.m. to 9 a.m. The technique consisted of laying white sheets on the floor of selected rooms, then spraying inside with overactivated solution of pyrethroids (by adding piperonyl butoxide), after first taking out food and blocking all exits that could serve as an escape for mosquitoes. Ten minutes after spraying, the sheets were carefully removed and Anopheles mosquitoes were collected using flexible forceps and then placed into numbered Petri dishes. Collected Anopheles mosquitoes were identified and classified according to the repletion state (unfed, fed, half-gravid and gravid). All the sampled anopheline specimens were individually stored in numbered Eppendorf tubes containing desiccant and kept for subsequent laboratory processing.

At the Laboratory of Vector and Parasite Ecology (LEVP), University of Cheikh Anta Diop (UCAD), Dakar, Senegal, a subsample of 52.46% (n = 724) An. gambiae s.l. and 46.75% (n = 381) An. funestus specimens were separately dissected. Head-thoraces, fed female abdomens and legs or wings, were used to determine Plasmodium infection rates, feeding host and species identification, respectively. Blood meal origin of indoor resting females was determined by Enzyme-Linked Immunosorbent Assay (ELISA) method [19]. The infectivity of host-seeking females was detected using another ELISA technique that targeted circumsporozoïtique antigen (ELISA-CSP) as described by Wirtz et al. [20]. Genomic deoxyribonucleic acid (DNA) was extracted from individual mosquitoes using the protocol of Collins et al. [21]. The sibling species of An. gambiae complex were identified using polymerase chain reaction (PCR) as described by Wilkins et al. [22]. Also, An. funestus group was identified using PCR described by Koekemoer et al. [23].

2.3. Data Analysis

Human biting rates were calculated by dividing the number of Anopheles females caught on human via HLC by the number of mosquito collectors who operated per night. Indoor resting densities were obtained by dividing the number of females collected inside dwellings by the number of rooms sprayed. Infection rates were obtained by dividing the number of infected female (ELISA-CSP positive) by the number of tested females. Parity rates were calculated by dividing the number of parous females by the total dissected. Entomological inoculation rates (EIR) were expressed as the product of infection rates and human biting rates. Human blood index was obtained as the ratio between the number of blood meals taken from humans and the total number of blood meals identified. Data were entered into an Excel database. The Wilcoxon rank-sum test was used to compare mean human biting rates between sites and indoors vs. outdoors, and mean indoor resting densities between sites. The average number of parous females were also compared between sites using Fisher’s exact test. All statistical analysis and graphs were performed using ggplot2, dplyr, tidyr and ggrepel packages under R software (version 4.5.1). The significance level was set at 5%.

3. Results

3.1. Diversity and Distribution of the Anopheline fauna

A total of 2,301 Anopheles female mosquitoes were collected during the entire study period, all collection methods combined, including 38.46% (n = 885) in Medina Dianguette, 30.86% (n = 710) in Ndrame Ndimb, 26.16% (n = 602) in Paoskoto and 4.52% (n = 104) in Nemataba. The collected samples consisted of 860 (97.18%) An. gambiae s.l. and 25 (2.82%) An. rufipes in Medina Dianguette against 63 (60.58%) An. gambiae s.l., 34 (32.69%) An. rufipes and 7 (6.73%) An. nili in Nemataba. While, in Ndrame Ndimb, the collected samples consisted of 369 (51.97%) An. gambiae s.l., 319 (44.93%) An. funestus, 18 (2.54%) An. rufipes, 2 (0.28%) An. pharoensis, 1 (0.14%) An. wellcomei and 1 (0.14%) An. ziemanni against 496 (82.39%) An. funestus, 88 (14.62%) An. gambiae s.l., 16 (2.66%) An. wellcomei and 2 (0.33%) An. rufipes in Paoskoto. The most more important specific diversity was observed in Ndrame Ndimb and the lowest specific diversity was recorded Medina Dianguette (Figure 2). As An. gambiae s.l. was predominant in Medina Dianguette, Nemataba and Ndrame Ndimb (with a good proportion of An. funestus), and An. funestus was predominant in Paoskoto, sub-sequent comparisons were made among these species.

3.2. Relative Abundance of An. gambiae s.l. and An. funestus

Out of 1,380 An. gambiae s.l. collected, 62.32% (n = 860) were collected from Medina Dianguette, with 88.14% (n = 758) by HLCs and 11.86% (n = 102) by PSCs, against 4.56% (n = 63) collected from Nemataba, with 80.95% (n = 51) by HLCs and 19.05% (n = 12) by PSCs. While, the 26.74% (n = 369) An. gambiae s.l. were collected from Ndrame Ndimb, with 8.13% (n = 30) by HLCs and 91.87% (n = 339) by PSCs, against 6.38% (n = 88) An. gambiae s.l. collected from Paoskoto, with 14.77% (n = 13) by HLCs and 85.23% (n = 75) by PSCs. This analysis shows that, in Medina Dianguette and Nemataba, more An. gambiae s.l. females were captured in the HLCs than in the PSCs collections, conversely, in Ndrame Ndimb and Paoskoto, more An. gambiae s.l. females were found in the PSCs than in the HLCs collections.

Out of 815 An. funestus collected, 60.86% (n = 496) were collected from Paoskoto, with 34.88% (n = 173) by HLCs and 65.12% (n = 323) by PSCs, against 39.14% (n = 319) collected from Ndrame Ndimb, with 6.90% (n = 22) by HLCs and 93.10% (n = 297) by PSCs. Both in Ndrame Ndimb and Paoskoto, more An. funestus females were found in the PSCs than in the HLCs collections.

In Nemataba, An. gambiae s.l. populations collected by HLCs, were mostly abundant in September with 73.33% indoors and 91.67% outdoors, however, they were most abundant in July with 80.54% indoors and 72.16% outdoors in Medina Dianguette. While, An. gambiae s.l. populations collected by HLCs were most abundant in July with 80% indoors and 75% outdoors in Ndrame Ndimb, and in Paoskoto in outdoors (50%) and in September for indoors population (60%). In Paoskoto, An. funestus collected by HLCs were most abundant in January with 30.43% indoors and 38.27% outdoors while, in Ndrame Ndimb, the most abundance were recorded indoors (40%) in January and outdoors in September and January with 42.86% (Table 1).

For indoor resting females, the relative abundance of An. gambiae s.l. was higher in September with 41.67% and 39.22% in Nemataba and Medina Dianguette, respectively. Conversely, the most abundance of An. gambiae s.l. were recorded in July with 73.75% and 64% in Ndrame Ndimb and Paoskoto, respectively. While, An. funestus was most abundant in September (32.20%) and January (61.28%) in Paoskoto and Ndrame Ndimb, respectively (Table 1).

3.3. Molecular Identification of An. gambiae Sibling Species and An. funestus Group

A sub-sample of 479 An. gambiae s.l. specimens were successfully identified by PCR in Medina Dianguette (n = 224), Ndrame Ndimb (n = 182), Nemataba (n = 27) and Paoskoto (n = 46). Molecular analysis revealed the presence of An. gambiae (50.90%, n = 114), An. coluzzii (41.96%, n = 94), An. arabiensis (4.91%, n = 11) and An. gambiae/An. coluzzii hybrids (2.23%, n = 5) in Medina Dianguette, whereas An. arabiensis (40.74%, n = 11), An. coluzzii (33.33%, n = 9) and An. gambiae (25.92%, n = 7) in Nemataba. While in Ndrame Ndimb, An. arabiensis (79.12%, n = 144), An. gambiae (17.58%, n = 32), An. coluzzii (2.20%, n = 4), and An. gambiae/An. coluzzii hybrids (1.10%, n = 2) were observed against An. arabiensis (95.65%, n = 44) and An. gambiae (4.35%, n = 2) in Paoskoto. Anopheles arabiensis was the most common species in both Ndrame Ndimb and Paoskoto, whereas An. gambiae and An. coluzzii were predominant in Medina Dianguette and a slight predominance of An. arabiensis was noted over An. gambiae and An. coluzzii in Nemataba (Figure 3). Molecular analysis performed on 248 An. funestus specimens in Ndrame Ndimb (n = 110) and Paoskoto (n = 138), revealed the exclusive presence of An. funestus s.s. 

3.4. Human Biting Rates of An. gambiae s.l. and An. funestus

A total of 60 person-nights were used, during 10 capture sessions, in each site. The mean human biting rates (HBRs) of An. gambiae s.l. were higher, in Medina Dianguette (14.73 bites per person per night (b/p/n)) than in Nemataba (1.05 b/p/n), and in Ndrame Ndimb (6.55 b/p/n) than in Paoskoto (1.47 b/p/n). However, Wilcoxon tests did not reveal any significant differences between the sites (p-value = 0.059 and p-value = 0.249, respectively). There was no significant difference between indoors (13.53, 0.5, 0.33 and 0.17 b/p/n) and outdoors (11.73, 1.2, 0.67 and 0.27 b/p/n) HBRs in Medina Dianguette (w = 13, p-value = 1), Nemataba (w = 13.5, p-value = 0.91), Ndrame Ndimb (w = 6, p-value = 0.18) and Paoskoto (w = 10, p-value = 0.65), respectively. However, mostly peak HBRs were recorded in July with (54.5, 8 b/p/n) indoors and (42.33, 15 b/p/n) outdoors in Medina Dianguette and Ndrame Ndimb, respectively. While, the HBRs peaked in September with 2 b/p/n indoors and 6.5 b/p/n outdoors in Nemataba, and in indoors (3 b/p/n) in Paoskoto, whereas, HBRs peaked in July with 4 b/p/n outdoors in Paoskoto (Figure 4).

The mean human biting rates of An. funestus were higher in Paoskoto (2.88 b/p/n) than in Ndrame Ndimb (0.37 b/p/n), and the difference was statistically significant (w = 0, p-value = 0.012). There was no significant difference between indoors (0.5 and 3.07 b/p/n) and outdoors (0.23 and 2.7 b/p/n) HBRs in Ndrame Ndimb (w = 17, p-value = 0.39) and Paoskoto (w = 15.5, p-value = 0.6), respectively. However, mostly peak HBRs were recorded in January with 4.67 b/p/n indoors and 5.17 b/p/n outdoors in Paoskoto, and in Ndrame Ndimb, HBRs peaked indoors (1 b/p/n) in September and outdoors (0.5 b/p/n) in January (Figure 4).

3.5. Nocturnal Activity of An. gambiae s.l. and An. funestus

Hourly biting profiles of these species showed heterogeneous variations between sites. In Medina Dianguette, HBRs showed that An. gambiae s.l. was more aggressive during the second half of the night and the highest peaks appeared outdoors between 2 a.m. to 3 a.m. and indoors 3 a.m. to 4 a.m. While in Nemataba, An. gambiae s.l. showed earlier peak indoors and outdoors between 9 p.m. to 10 p.m., an activity more intense outdoors with two peaks between 11 p.m. to 12 a.m. and 2 a.m. to 3 a.m., and peaks indoors from 1 a.m. to 3 a.m. In Ndrame Ndimb, An. gambiae s.l. biting pressure was higher outdoors where the highest peaks appeared between 10 p.m. to 11 p.m. and 12 a.m. to 1 a.m. and four similar peaks were noted indoors (8 p.m./9 p.m., 10 p.m./11 p.m., 2 a.m./3 a.m. and 5 a.m./6 a.m.). Anopheles funestus showed a first peak indoors between 11p.m. to 12 a.m., and the highest peaks appeared indoors and outdoors between 3 a.m. to 4 a.m. Conversely, in Paoskoto, An. gambiae s.l. biting pressure was lower and the peaks were appeared indoors between 1 a.m. to 2 a.m. and outdoors between 2 a.m. to 3 a.m. However, An. funestus biting pressure showed earlier peaks indoors and outdoors between 8 p.m. to 9 p.m. The highest peaks appeared in the second half of the night indoors between 2 a.m. 3 a.m. and 4 a.m. to 5 a.m., and outdoors between 1 a.m. 2 a.m. and 3 a.m. to 4 a.m. (Figure 5).

3.6. Physiological State of An. gambiae s.l. and An. funestus

The abdominal statuses of 620 An. funestus and 528 An. gambiae s.l. resting females were determined. Among the 620 An. funestus females, 53.53% and 48.61% of them collected in Ndrame Ndimb and Paoskoto, respectively, had recently fed. Of the 528 An. gambiae s.l., 53.92%, 48.08%, 41.67% and 41.33% of them collected in Medina Dianguette, Ndrame Ndimb, Nemataba and Paoskoto, respectively, had recently fed (Table 2).

3.7. Indoor Resting Densities of An. gambiae s.l. and An. funestus

The mean indoor resting density (IRD) of An. gambiae s.l. was significantly (w = 25, p-value = 7.9e-3) lower in Nemataba [0.24 females per room (f/r)] than in Medina Dianguette (2.04 f/r). Conversely, there was no significant difference (w = 19, p-value = 0.209) in IRDs between Paoskoto (1.5 f/r) and Ndrame Ndimb (6.78 f/r). While, there was no significant difference (w = 8.5, p-value = 0.463) in IDRs of An. funestus between Ndrame Ndimb (5.94 f/r) and Paoskoto (6.46 f/r).

The highest IRDs of An. gambiae s.l. were recorded in July in Ndrame Ndimb (25 f/r) and Paoskoto (4.8 f/r), and in September in Medina Dianguette and Nemataba, with 4 f/r and 0.5 f/r, respectively. The highest IRDs of An. funestus were observed in September in Paoskoto and in January in Ndrame Ndimb, with 10.4 f/r and 18.2 f/r, respectively (Figure 6).

3.8. Trophic Preferences and Human Blood Index

The blood meals of 238 An. gambiae s.l. and 233 An. funestus females were tested, including 159 An. gambiae s.l. and 113 An. funestus from Ndrame Ndimb, 120 An. funestus and 27 An. gambiae s.l. from Paoskoto, 49 An. gambiae s.l. from Medina Dianguette and 3 An. gambiae s.l. from Nemataba. Out of the 159 An. gambiae s.l. tested in Ndrame Ndimb, 165 blood meals were successfully identified, with a human blood index (HBI) of 2.42%. Similarly, in Paoskoto, of the 27 An. gambiae s.l. tested, 38 blood meals were successfully identified, and none of them was taken on human giving a null HBI. Blood meals identified alone on equine hosts were 84.85% and 50% in Ndrame Ndimb and Paoskoto, respectively. Conversely, out of the 49 An. gambiae s.l. tested in Medina Dianguette, 48 blood meals were successfully identified, with a HBI of 77.08% while in Nemataba, the 3 blood meals of An. gambiae s.l. tested, were successfully identified, with a HBI of 66.67% (Table 3).

Among the 113 An. funestus tested in Ndrame Ndimb, 118 blood meals were successfully identified, with a HBI of 5.93%. Meanwhile, of the 120 An. funestus tested in Paoskoto, 137 blood meals were successfully identified, with a HBI of 3.65%. Blood meals identified alone on equine hosts were 49.15% and 47.44% in Ndrame Ndimb and Paoskoto, respectively. Globally, equines were the first host of choice for An. gambiae s.l. and An. funestus populations in Ndrame Ndimb and Paoskoto whereas, in Medina Dianguette and Nemataba, humans were the first host of choice for An. gambiae s.l. 

3.9. Pariry Rates of An. gambiae s.l. and An. funestus

The mean parity rates of An. gambiae s.l. were 44.96% and 60% in Medina Dianguette and Nemataba, respectively. Although, the parity rate was higher in Nemataba than Medina Dianguette, the difference was not statistically significant (OR = 1.63, CI_95% [0.80 – 3.46], p-value = 0.19). There was no significant difference (OR = 1.49, IC_95% [0.30 – 9.89], p-value = 0.73) in mean parity rates of An. gambiae s.l. between Ndrame Ndimb (33.33%) and Paoskoto (25%).

Similarly, no significant difference (OR = 0.93, IC_95% [0.36 – 2.47], p-value = 1) in mean parity rates of An. funestus was noted between Ndrame Ndimb (56%) and Paoskoto (57.72%) (Table 4).

3.10. Infection Rates (IRs) and Malaria Transmission

A total of 226 An. gambiae s.l. females were tested by ELISA-CSP, including 188, 16, 15 and 7 from Medina Dianguette, Nemataba, Ndrame Ndimb and Paoskoto, respectively. Amongst the An. gambiae s.l. tested, one was found to be infected in Medina Dianguette and in Paoskoto within 0.53% and 14.28% IRs, respectively. No female was found infected in Ndrame Ndimb and Nemataba. The entomological inoculation rates were 0.078 infected bites per person per night (ib/p/n) and 0.209 ib/p/n in Medina Dianguette and Paoskoto, respectively. The transmission was not detected in Nemataba and Ndrame Ndimb during the study period.

Amongst the 98 An. funestus tested by ELISA-CSP, including 15 in Ndrame Ndimb and 83 in Paoskoto, none of them was found to be infected, indicated that An. funestus was not involved in malaria transmission during the study period.

Table 5. Infection rates of An. gambiae s.l. and An. funestus in Ndrame Ndimb, Paoskoto, Nemataba and Medina Dianguette.

Sites	An. funestus			An. gambiae s.l.
	Tested	Positive	IR (%)	Tested	Positive	IR (%)
Medina Dianguette	-	-	-	188	1	0.53
Ndrame Ndimb	15	0	0	15	0	0
Nemataba	-	-	-	16	0	0
Paoskoto	83	0	0	7	1	14.28
Total	98	0	0	226	2	0.88

IR = infection rate.

Table 5. Infection rates of An. gambiae s.l. and An. funestus in Ndrame Ndimb, Paoskoto, Nemataba and Medina Dianguette.

Sites	An. funestus			An. gambiae s.l.
	Tested	Positive	IR (%)	Tested	Positive	IR (%)
Medina Dianguette	-	-	-	188	1	0.53
Ndrame Ndimb	15	0	0	15	0	0
Nemataba	-	-	-	16	0	0
Paoskoto	83	0	0	7	1	14.28
Total	98	0	0	226	2	0.88

IR = infection rate.

4. Discussion

The investigation carried out alongside the Senegambian border into bionomic characteristics of major vector species and malaria transmission has shown that An. arabiensis was the most common species in both Ndrame Ndimb and Paoskoto. Whereas, An. gambiae and An. coluzzii were predominant in Medina Dianguette, and a slight predominance of An. arabiensis was noted over An. gambiae and An. coluzzii in Nemataba. The predominance of An. arabiensis in Ndrame Ndimb and Paoskoto is consistent with the ecological conditions of this area, as these two sites are located in central Senegal, a slightly drier area. As An. arabiensis is highly tolerant of dry and arid environments, several studies have demonstrated its predominance in the center of the country [15,24,25,26]. Its predominance in Nemataba might be explained by the timing of the study collection, the majority of which was carried out during the dry season, which might reflect seasonal shifts in species composition [10,16]. Anopheles gambiae and An. coluzzii are typically more abundant in humid area, Medina Dianguette is located in south part of the country characterize by bioclimatic conditions favorable to the development of larval and the survival of adult populations in the humid savannah and forest ecosystems [17,27]. In An. funestus group, An. funestus s.s. was exclusive present in Ndrame Ndimb and Paoskoto. This finding corroborates several studies showing the exclusive present of An. funestus s.s. in Senegal [15,28,29,30].

Relative abundance of An. gambiae s.l. species was the highest during the rainy season (July and September), this high abundance might be explained by the presence of rainwater, which provide suitable breeding sites leading to the proliferation of mosquitoes [7,17]. While for An. funestus, the most abundance were recorded in September and January. The persistent high abundance of An. funestus during the dry season might be due to the permanent nature of these breeding sites [14,17]. Changes to the environment brought about by hydro-agricultural developments and crop irrigation mean that Ndrame Ndimb and Paoskoto are increasingly characterized by a specific ecological profile typical of An. funestus. Irrigation is also known as a factor that increasing Anopheles breeding sites [31]. This discrepancy between species highlights the variable capacity of Anopheles adaptations across ecological contexts.

The high peaks of HBRs of An. gambiae s.l. were recorded in July in Medina Dianguette and Ndrame Ndimb, both indoors and outdoors. While, HBRs peaked in September indoors and outdoors in Nemataba, and in Paoskoto, HBRs peaked indoors in September and outdoors in July. The higher density of An. gambiae s.l. following the rainy season is consistent with findings from other studies that link rainfall with the proliferation of breeding sites for these species [7,32,33]. However, mostly peak HBRs of An. funestus were recorded in January indoors and outdoors in Paoskoto, and in Ndrame Ndimb, HBRs peaked indoors in September and outdoors in January, this finding was consistent with previous study in central Senegal [28]. Both in An. gambiae s.l. and An. funestus, there was no significance difference in HBRs between indoors and outdoors, which showed that these species bite, as well as indoors and outdoors. Anopheles gambiae s.l. was more aggressive during the second half of the night in Medina Dianguette and in Paoskoto while, in Nemataba and Ndrame Ndimb, An. gambiae s.l. showed earlier peaks in the first time of the night and others peaks in the second half of the night. Despite a few peaks in the early evening in Nemataba and Ndrame Ndimb, An. gambiae s.l. remains active during the second half of the night, consistent with his known nocturnal biting behaviour [7,34,35]. These peaks in the early evening, can be explained by the human host’s availability during this period given that they are still active [10], supported by Vajda et al. [36]. Anopheles funestus was more aggressive during the second half of the night both in Paoskoto and Ndrame Ndimb, consistent with previous study in the area [28]. Conversely, others studies showed that An. funestus bites during the day [15] and even more aggressive in broad daylight [30]. Evidence of nocturnal activity is particularly relevant for the evaluation and implementation of malaria control strategies based on human exposure patterns.

Anopheles gambiae s.l. were highly anthropophilic in Medina Dianguette and Nemataba, consistent with previous studies conducted in the south part of the country [7,10,27,37]. This high HBI demonstrates the affinity that the vector has with humans, despite the use of mosquito nets. While, the lower HBIs were noted in Ndrame Ndimb and Paoskoto, can be explained by the fact that in this area, An. arabiensis was the most common species within the An. gambiae complex, exhibits behavioural flexibility in relation to animals. Also, An. funestus exhibits a low level of anthropophilic, similar findings were reported by Gueye et al. [26], Faye et al. [28] in the same area. Globally in Ndrame Ndimb and Paoskoto, the widespread use of LLINs may explain these low HBIs in An. gambiae s.l. and An. funestus, meaning that, following the availability and accessibility of human host, they may resort to feed on alternative hosts [38,39]. It is also worth noting that in Ndrame Ndimb and Paoskoto, there are many horses and other domestic animals. Previous studies carried out in the central region of the country have reported HBIs for An. gambiae s.l. and An. funestus of over 70%, notably by Konaté et al. [39] in Dielmo. Otherwise, the low HBIs observed in Ndrame Ndimb and Paoskoto coincided with predominance of An. arabiensis and substantial equine feeding, suggesting that host availability and vector host preference may contribute to reduced human feeding. The concurrent human behavioural observations reported in the companion analysis [36] provide additional context for interpreting the extent to which reduced human-vector contact may also reflect LLIN protection.

In Medina Dianguette, Ndrame Ndimb and Paoskoto, An. gambiae s.l. consisted mainly of young females. The predominance of nulliparous females in several sites may be consistent with recent adult emergence following seasonal increases in vector abundance, although other factors, including differential mortality associated with vector-control interventions, may also contribute. While in Nemataba, An. gambiae s.l. consisted mainly of older females, this finding corroborates previous studies conducted in Senegal where An. gambiae s.l. showed high parity rates [8,14,25,35]. Therefore, An. funestus consisted mainly of older females both in Ndrame Ndimb and Paoskoto, previous study in Dielmo reported high parity rate [30]. Conversely, these observations contrast with those of Dia et al. [14], of Faye et al. [28] in the north and center of the country, which show a relatively young population of An. funestus.

Only one An. gambiae s.l. specimen in Medina Dianguette and in Paoskoto were found infected by Plasmodium. Similarly, low infection rates were recorded by Sy et al. [8] in central Senegal. These infection rates observed, were low compared to those obtained from these areas within An. gambiae s.l. by Konaté et al. [35]. A study carried out in the southern part of the country, in contrast to the low infection rate in Medina Dianguette, showed that the species An. arabiensis, An. gambiae and An. coluzzii were infected by Plasmodium, with high infection rates in different villages [40]. No female An. gambiae s.l. were found to be infected by Plasmodium in Ndrame Ndimb and Nemataba. This absence of malaria transmission may be explained by the fact that the An. gambiae s.l. population in Ndrame Ndimb consists mainly of young females, and although the An. gambiae s.l. population in Nemataba has reached the critical epidemiological age, no malaria transmission was detected. Anopheles funestus population was not involved in malaria transmission during the study period. In contrast, previous studies conducted in central Senegal have showed that An. funestus was involved in malaria transmission [15,28,30]. Overall, the use of vector control strategies may explain this lack of transmission in Ndrame Ndimb and Paoskoto. Additionally, the opportunistic feeding behaviour of An. gambiae s.l. and An. funestus, with a lower HBIs in Ndrame Ndimb and Paoskoto, may further influence the infection rate and malaria transmission [41].

The study sites were located solely in Senegal, having sites on both the Senegalese and Gambian sides, and along the entire border, would enable a better characterization of the bionomics of major malaria vectors. Given the short duration of the study, monthly longitudinal monitoring throughout the year would help to better determine variations in abundance, as well as the biting and resting behaviours of the vectors. Laboratory analyses, with greater financial resources, would enable all samples to be processed in order to maximize the chance to detect infected females. The concurrent human behavioural analysis [36] provides an important complementarity perspective by integrating vector biting patterns with human location, activity, sleeping patterns and LLIN use to identify when and where human exposure occurs despite existing vector-control interventions. The use of additional collection methods, such as window traps, Prokopack aspirator in outdoor shelters would have been beneficial to better characterize the resting behaviour of vector populations. Importantly, the vector bionomic patterns described here should be interpreted alongside the human behaviour observations collected concurrently at the same sites. While, the present analysis characterizes species composition, biting and resting behaviours, host feeding and physiological characteristics, the companion analysis integrates these entomological data with hourly observations of human location, activity, sleeping and LLIN use to quantify the extent to which vector activity overlaps with human exposure and to identify gaps in protection. This complementary approach demonstrates how species and behaviour-specific vector characteristics translate into heterogeneous human exposure across sites.

5. Conclusions

This study provides preliminary information about vector bionomic characteristics in the Senegambian cross border. Specific composition showed that An. arabiensis were more prevalent in central of the country and in the south, An. gambiae and An. coluzzii were predominant in Medina Dianguette while in Nemataba, there was a slight predominance of An. arabiensis over An. gambiae and An. coluzzii. Anopheles funestus was only caught in Ndrame Ndimb and Paoskoto. Both An. gambiae s.l. and An. funestus bite, as well as indoors and outdoors, An. gambiae s.l. was more aggressive from July to September while, An. funestues was more aggressive from September to January. Anopheles gambiae s.l. was consisted mainly of young females in Medina Dianguette, Ndrame Ndimb and Paoskoto, While in Nemataba, it consisted mainly of older females. Also, An. funestus was consisted mainly of older females both in Ndrame Ndimb and Paoskoto, Equines were the first host of choice for An. gambiae s.l. and An. funestus populations in Ndrame Ndimb and Paoskoto whereas, in Medina Dianguette and Nemataba, humans were the first host of choice for An. gambiae s.l. Plasmodium infection was detected in An. gambiae s.l. from Medina Dianguette and Paoskoto, although sample sizes were limited. These findings provide basic data on vector bionomic in the Senegambian cross border, which can allow malaria control programmes to plan interventions for better decision-making.