Blood-Feeding Patterns and Resting Behavior of Aedes aegypti from Three Health Districts of Ouagadougou City, Burkina Faso

ABSTRACT.

Recent dengue outbreaks have occurred in Ouagadougou and Bobo-Dioulasso, the two major cities of Burkina Faso. Dengue is a viral disease transmitted primarily by Aedes aegypti, a highly anthropophilic mosquito that thrives in human-transformed environments and breeds predominantly in artificial containers. In 2018, we investigated the resting and blood-feeding habits of Ae. aegypti in urban settings of Ouagadougou. In a 3-month cross-sectional study starting in August 2018, indoors and outdoors resting adult mosquitoes were collected using Prokopack aspirators in three health districts (HD). All mosquitoes were morphologically identified, and DNA was extracted from blood-fed Ae. aegypti females. A multiplex polymerase chain reaction with specific primers was used to identify the origin of the blood meal. A total of 4,256 adult Ae. aegypti mosquitoes, including 1,908 females, were collected. A preference for exophily was recorded in Bogodogo and Nongremassom, although an unexpectedly higher proportion of blood-fed females were found indoors than outdoors. Respectively, 96.09%, 91.03%, and 95.54% of the blood meals successfully analyzed in Baskuy, Bogodogo, and Nongremassom were from a single human host, with the remainder from domestic mammals as single or multiple hosts. Modeling total Ae. aegypti and blood-fed female counts showed that among other predictors, human density, outdoor environment, and house type affect their total densities. Our study revealed an exophilic tendency as well as a pronounced anthropophilic preference of Ae. aegypti adults, critical findings to consider when planning accurate entomological surveillance and effective interventions against Ae. aegypti in urban settings.

INTRODUCTION

Aedes aegypti, a vector that continues to spread globally, is responsible for transmitting dengue, chikungunya, Zika, and yellow fever, leading to increased incidence and distribution of these arboviral diseases. In sub-Saharan Africa, dengue in particular has increased in recent decades.¹ Aedes aegypti is considered globally to be typically anthropophilic with preferential indoor resting habits²; however, recent data related to Ae. aegypti bionomics in Burkina Faso highlighted more exophilic behavior.³ Improving knowledge of Ae. aegypti resting behavior and identifying the preferred sites within households and surrounding environments can help optimize surveillance efforts and guide the implementation of vector control strategies such as targeted indoor residual spraying or the use of insecticide-treated materials.²

Resting behavior is likely to vary, especially in Africa, but few investigations have been conducted lately on Aedes mosquitoes in West Africa.⁴ Recently, some studies in Senegal have reported contrasting behaviors in two subpopulations of Ae. aegypti that are identified based on morphological differences.⁵ The two morphotypes, enduringly known as Ae. aegypti aegypti and Ae. aegypti formosus, are no longer recognized as distinct subspecies by some mosquito taxonomists.⁶ Notwithstanding their unsettled taxonomic status, the subpopulation more closely associated with domestic environments showed a greater tendency to rest indoors and feed on humans than the subpopulation that is generally found far from human settlements in Senegal.^5,7,8 In Burkina Faso and Ghana, this contrast is reduced, with studies revealing that outdoor collections of Ae. aegypti are 2–4 times higher than indoor collections.^3,9,10 Given these heterogeneities, more detailed knowledge of Ae. aegypti resting behavior is a prerequisite to identify priority locations for vector control interventions.

The relative distribution of blood meals among alternative vertebrate hosts can affect the transmission of vector-borne diseases.^11,12 When mosquitoes, such as Ae. aegypti, feed primarily on humans, the risk of disease transmission to humans increases because of the greater contact between humans and the vector.^13,14 However, when Ae. aegypti also feeds on alternative hosts, including both humans and animals, zoonotic disease transmission increases and possibly even dengue¹⁵ is increased. The highly anthropophilic domestic populations of Ae. aegypti live in close association with humans, breeding both inside and outside houses in various artificial containers that are typically filled with water from rains or residences.^3,16,17 A recent study in Senegal revealed a preference for feeding on humans (78.6%), with other blood meals sourced from domestic animals such as dogs, cats, horses, cattle, sheep, and rats.⁵ During the 2016 and 2017 dengue outbreaks in Burkina Faso, 92% and 76% of blood-fed mosquitoes, respectively, had human blood meals, with the remainder from dogs, cows, and pigs.³ In neighboring Ghana, 90% of bloodmeals were from humans and 5% were from dogs or cows.⁹

To assess the risk of dengue outbreaks and fill the knowledge gap of Ae. aegypti bionomics in urban areas in Africa, this study aimed to estimate behavioral traits of adult populations of Ae. aegypti, with specific emphasis on their blood-feeding and resting behaviors in the health districts (HDs) of Baskuy, Bogodogo, and Nongremassom in the city of Ouagadougou. These districts accounted for 61% of the country’s 13,135 dengue cases reported during the 2017 outbreak.¹⁸ Understanding the blood-feeding behavior patterns of the vector is crucial in evaluating the risk of dengue transmission, while identifying their resting behaviors can help in implementing effective strategies for vector control.¹⁹

MATERIALS AND METHODS

Study area.

During the 2018 rainy season, mosquito collections were conducted from August to October in Ouagadougou (12°22′07.64″N, 001°31′35.63″W), the capital city of Burkina Faso (Figure 1). The investigation focused on three HDs within the central health region, i.e., of Baskuy, Bogodogo, and Nongremassom, which have been extensively described by Ouédraogo et al.¹⁶ The HD of Baskuy is entirely urban and located in the central part of the city, whereas Bogodogo, in the northern part of the city, and Nongremassom, in the southeastern part of the city, have both urban and semiurban areas. These districts reported significant numbers of cases during the 2016–2017 dengue outbreak.

Figure 1.

Map showing visited areas in the three health districts of the study in Ouagadougou. The spots represent the neighborhoods where mosquitoes were collected. The sizes of spots are proportional to the densities of Aedes aegypti mosquitoes in these areas. Spots of the same color belong to the same health district.

Meteorological data.

The National Meteorological Agency provided the meteorological data for the year 2018 pertaining to the Ouagadougou station. This dataset includes the minimum, maximum, and average temperature, relative humidity, and daily rainfall. The 10-day cumulative rainfall was computed using these data.

Adult mosquito collections.

A house-to-house cross-sectional entomological survey was conducted to collect indoor and outdoor resting mosquitoes. We defined at first some neighborhoods distributed as regularly as feasible across the study area and then selected compounds at random within each neighborhood. Access to the property was occasionally not possible, in which case we selected the nearest compound.

In each visited compound, adult mosquitoes were aspirated with Prokopack devices from various spots indoors, such as living rooms, bedrooms, showers, and kitchens, as well as from the courtyard outdoors.^20,21 In addition to sampling of adults, immature stages were collected from larval containers and were counted. These data have already been published¹⁶; we kept only the immature stage count per house for modeling purposes. For indoor collections, we focused on corners, wall surfaces, hanging clothes, underneath beds, or other furniture. For outdoor collections, we focused on potential mosquito shelters in the courtyard, such as under trees, on water troughs, on walls, and near open containers or other places where mosquitoes may rest. Prokopack collection boxes were labeled with the date of collection, location, house number, position, and duration.

Collections lasted for 10 minutes indoors (living and sleeping rooms) and another 10 minutes outdoors²²; this was usually sufficient to cover the whole property. To visit properties when the owners were present, collections took place from 06:00 to 09:00 and from 16:00 to 19:00.

Upon our return to the laboratory, the mosquitoes were killed by chilling and transferred to Petri dishes for separation from other insects and debris before morphological identification. Data such as the number of people living in each visited compound, the number of houses, the type of house (modern, traditional, or semimodern, i.e., built with both traditional and modern materials), the presence and number of domestic animals, and the occurrence of vegetation were recorded.

Mosquito processing and morphological identification.

All collected mosquitoes were morphologically identified using a binocular dissecting scope (magnifications, ×50 and ×100 ) in accordance with the keys of Huang²³ and Rueda²⁴ for Aedes mosquitoes, while other culicine species were identified using the key of Edwards²⁵ and the Anopheles species were identified in accordance with the keys of Gillies and De Meillon.²⁶

After identification, the mosquitoes were counted and separated according to physiological status: unfed, blood-fed females (semifed or fully fed), and gravid (semigravid or fully gravid) based on observation of the mosquito abdomen.²⁷ They were then stored in Eppendorf tubes over silica gel and kept at −20°C. Blood-fed female Ae. aegypti mosquitoes were placed in labeled Eppendorf tubes for molecular identification of the origin of the blood meal by polymerase chain reaction (PCR).

Identification of blood meal origin.

The DNA extracted from blood-fed female abdomens was used for a multiplex PCR assay with primers specific to human, pig, dog, goat, and cow hosts, in accordance with the protocol described by Kent and Norris.²⁸ Although this choice does not cover the whole range of potential hosts available in the area (e.g., rodents, birds, or reptiles), it does account for the most significant ones.

Each reaction consisted of 12.5 µL, including 0.0625 µL of the five host primers, 4.875 µL of distilled water, and 6.25 µL of AmpliTaq enzyme. The PCR conditions included an initial denaturation at 95°C for 5 minutes, followed by cycles of denaturation at 95°C for 60 seconds, annealing at 56°C for 60 seconds, extension at 72°C for 60 seconds, and a final extension step at 72°C for 7 minutes. Subsequently, the PCR products were separated on a 1.5% agarose gel for 35 minutes and visualized using ethidium bromide under ultraviolet light.

STATISTICAL ANALYSES

Data were recorded and processed using Excel, and statistical analyses were performed in R software v. 3.4.3. The proportions of mosquito species among HDs were compared using the χ² test. Associations between mosquito densities and environmental variables were analyzed using generalized linear mixed models (GLMMs). Generalized linear mixed models were fitted using the R package “glmmTMB”²⁹ with a negative binomial link function. Prior to fitting the GLMMs, we conducted a pairwise correlation of the variables of interest associated with house, climatic, and mosquito-related factors. Mean relative humidity, maximum temperature, and precipitation were correlated (<95%), so we opted for mean temperature, minimum relative humidity, and 10-day precipitation. We also observed a correlation between the number of persons in the house and the number of children; hence, we included the number of persons in the models. Predictor variables encompassed collection month, collection period (morning, afternoon), collection position (indoor, outdoor), rainfall, temperature, relative humidity, number of residents in the compound, construction type, presence and number of bed nets, presence and number of animals, vegetation, HD, and their interactions. These were included in the models, with the date of collection and house numbers considered as random effects.

From these full models, we selected the minimal model using a stepwise backward procedure based on the lowest Akaike information criterion (AIC) values by removing the factors with the highest P-value in the model. If the removal of a variable resulted in a change in the AIC value of more than 2 and the resulting model remained parsimonious (after the diagnosis of residuals and overdispersion in the R package DHARMa), the procedure was repeated until the removal of variables no longer improved the model. We set a significance level of P = 0.05 for all statistical analyses.

RESULTS

Characteristics of surveyed houses.

During the survey period, 656 houses were investigated in the study area, 196 in Baskuy and 230 each in Bogodogo and Nongremassom (Supplemental Table 1). The average number of residents per household was 8.48, 6.86, and 5.22 in Baskuy, Bogodogo, and Nongremassom, respectively (Supplemental Table 1). Households with at least one bed net accounted for 76.53%, 78.70%, and 81.74% (χ² = 1.78, df = 2, P = 0.41), with an average number of bed nets per resident of 0.36, 0.33, and 0.37, respectively, in Baskuy, Bogodogo, and Nongremassom (Supplemental Table 1). Approximately 50% of the compounds had domestic animals, with an average of four animals per compound reported in all HDs (Supplemental Table 1).

Adult mosquito species composition and abundance.

A total of 39,483 mosquitoes belonging to the Aedes, Anopheles, Culex, and Mansonia genera were collected. Among Aedes species, Ae. aegypti was the most abundant, with total counts of 1,360, 1,403, and 1,493 in Baskuy, Bogodogo, and Nongremassom, respectively (Table 1). Other collected aedine mosquitoes included Ae. ochraceus and Ae. hirsutus. The relative abundance of Ae. aegypti varied across districts, with percentages of 9.67%, 10.36%, and 12.56% in Baskuy, Bogodogo, and Nongremassom, respectively (χ² = 50.41, df = 2, P <0.001) (Table 1). Among the other mosquito species, Culex quinquefasciatus was the most abundant in all HDs, accounting for 85.75%, 80.26%, and 78.67% in Baskuy, Bogodogo, and Nongremassom, respectively (χ² = 2 46.2, df = 2, P <0.0001). After Cx. quinquefasciatus, Anopheles gambiae s.l. was more prevalent in Nongremassom (7%) and Bogodogo (5.9%) than in Baskuy (3%) (χ² = 229.55, df = 2, P <0.0001).

Table 1.

Diversity and abundance of mosquitoes collected according to genus and species and HDs in the study areas

Mosquito Species	No. (%) of Mosquitoes			P-Value
	Baskuy	Bogodogo	Nongremassom
Aedes aegypti	1,360 (9.67)	1,403 (10.36)	1,493 (12.58)	<0.001
Aedes ochraceus	1	0	0	–
Aedes hirsutus	1	0	0	–
Aedes sp.	50	37	35
Anopheles gambiae	422 (3)	798 (5.9)	831 (7%)	<0.001
Anopheles domicola	2	0	0	–
Anopheles funestus	0	0	1	–
Anopheles rufipes	0	0	1	–
Culex quinquefasciatus	12,063 (85.75)	10,869 (80.26)	9,340 (78.67)	<0.001
Culex decens	146	264	154	–
Culex nebulosus	3	155	3	–
Culex poicilipes	0	1	0	–
Culex sp.	19	14	12	–
Mansonia africana	0	1	0	–
Mansonia uniformus	1	0	1	–
Lutzia tigripes	0	1	1	–

HDs = health districts.

Resting location of adult Aedes aegypti mosquitoes.

The proportion of outdoor resting Ae. aegypti mosquitoes (313/630, 49.68%) was similar to that indoors in Baskuy, but significantly higher in Bogodogo (412/637, 64.68%) and Nongre-Massom (380/641, 59.28%) compared to indoors (Figure 2).

Figure 2.

Proportions of bloodfed (grey) and other stages (white) female Aedes aegypti collected outdoors (courtyard) and indoors (rooms) according to health districts.

However, the relative proportions of blood-fed females collected indoors were significantly higher than those collected outdoors in all HDs (Figure 2). In Baskuy, Bogodogo, and Nongremassom, 38.80%, 41.78%, and 45.51% of blood-fed females were collected indoors, compared with 26.20%, 33.01%, and 21.84% collected outdoors, respectively, relative to the total number of Aedes mosquitoes collected indoors and outdoors.

Identification of blood meals in adult females of Ae. aegypti.

A total of 385 blood meals from 640 females were successfully amplified by PCR for blood meal origin (Table 2). Our results show homogeneous and high anthropophily in all HDs based on amplified samples, with 96.1%, 91.0%, and 95.5% of blood meals on humans as single hosts in the HDs of Baskuy, Bogodogo and Nongremassom, respectively, and no statistically significant difference between HDs (χ² = 2.93, df = 2, P = 0.23). Other identified blood meal hosts were domestic mammals, such as pigs, cows, and goats, at very low frequencies. Considering multiple mixed-host blood meals, including human-animal, the human blood indices in all three HDs ranged from 97.24% to 99.1%. These proportions were statistically similar between HDs (χ² = 1.3, df = 2, P = 0.52).

Table 2.

Blood meal origin of blood-fed Aedes aegypti females collected in HDs

Health District	No. of Blood-Fed Ae. aegypti	No. of Blood-Fed Ae. aegypti Amplified	No. (%) of Mosquitoes with Indicated Blood Meal Origin
			Human	Human-Cow	Human-Dog	Human-Pig	Cow	Dog	Goat	Pig
Baskuy	198	128	123 (96.09)	3 (2.34)	–	–	–	1 (0.78)	1 (0.78)	–
Bogodogo	242	145	132 (91.03)	3 (2.07)	4 (2.76)	2 (138)	1 (0.69)	2 (1.38)	1 (0.69)	–
Nongremassom	200	112	107 (95.54)	3 (2.68)	1 (0.89)	–	–	–	–	1 (0.89)

HDs = health districts.

Factors associated with Ae. aegypti abundance.

Significant predictors of adult Ae. aegypti mosquito abundance per household included the health district, vegetation in the compound, minimum relative humidity, collection location (indoors versus outdoors), house type, number of persons in the house, number of Ae. aegypti immature stages, and interaction between HD and month (Table 3). The mosquito densities were two times higher in Bogodogo than in the reference HD (Baskuy) and almost two times higher in outdoor collections than indoors.

Table 3.

Outcome of the GLMM model of adult Aedes aegypti mosquito density per house^*

Predictors†	Incidence Rate Ratio	CI	P-Value
(Intercept)	0.04	0.00–0.39	0.005
HD [Baskuy]
Bogodogo	2.18	1.26–3.79	0.006
Nongremassom	1.42	0.85–2.38	0.177
Month [August]
October	0.75	0.39–1.45	0.394
September	1.62	0.89–2.95	0.118
Vegetation [No]
Yes	1.34	1.00–1.79	0.049
Minimal relative humidity	1.04	1.01–1.07	0.005
Location [Indoor]
Outdoor	1.85	1.53–2.23	<0.001
House type [Mixed]
Modern	1.42	1.09–1.84	0.008
Traditional	0.65	0.31–1.39	0.268
No. of persons (in house)	1.03	1.01–1.05	0.004
No. of Ae. aegypti immature stages	1.00	1.00–1.00	<0.001
HD [Baskuy] × Month [August]
Bogodogo × October	0.50	0.22–1.13	0.094
Nongremassom × October	0.68	0.31–1.49	0.339
Bogodogo × September	0.24	0.11–0.55	0.001
Nongremassom × September	0.44	0.21–0.94	0.034

GLMM = generalized linear mixed model; HD = health district.

^*Nonsignificant terms not included in the model were temperature, presence and number of bed net, presence and number of animals, collection period (morning, afternoon), and rainfall.

^† Reference categories are shown in brackets. Significant predictors are highlighted in bold, and nonsignificant predictors are listed immediately thereunder.

Aedes aegypti female adult abundance was positively associated with the presence of vegetation, the number of persons in the house, and the number of Ae. aegypti immature stages. The interaction between month and locality also significantly affected the adult mosquito abundance in the house, particularly for Bogodogo and Nongremasson in September, with reduced catches compared with that of the reference. Consistent with the analyses presented above (Figure 2), the collection location (indoor versus outdoor) was a key factor, with female Ae. aegypti mosquitoes more abundant outdoors. However, factors such as the presence of vegetation, type of construction, and number of residents were positively associated with the abundance of female adult Ae. aegypti mosquitoes in the household. A second GLMM on female Ae. aegypti showed similar predictors to those affecting Ae. aegypti densities, except for the 10-day cumulative rainfall, which replaced relative humidity in the previous model (Supplemental Table 2).

A third GLMM explored the factors predicting blood-fed Ae. aegypti abundance in the household (Supplemental Table 3). The results indicated that the locality, type of house, minimum relative humidity (correlated with rainfall), number of residents in the house, and the interaction between HD and month significantly influenced blood-fed Ae. aegypti densities. In Bogodogo, nearly three times more blood-fed Ae. aegypti mosquitoes were collected than in the reference (Baskuy), and modern houses yielded almost twice as many mosquitoes as the reference mixed house type. Minimum relative humidity and number of residents had a positive impact on the abundance of blood-fed females in the compound, whereas the interaction between HD and month had a negative effect, particularly for the Bogodogo × September interaction, indicating low catches in that month (Supplemental Table 3).

DISCUSSION

A previous study conducted during the 2016 and 2017 dengue outbreak in rural, periurban, and urban areas in and around Ouagadougou revealed that the outdoor density of adult Ae. aegypti mosquitoes was significantly higher than the indoor density.³ Therefore, the aim of this study was to enhance the data on the bionomics of Ae. aegypti in urban areas of Ouagadougou and ensure the reproducibility and consistency of characteristics of Ae. aegypti bionomics over time, which is essential for adapting vector control strategies. The findings of the current study also indicated predominantly exophilic behavior in this population of Ae. aegypti, with a relatively higher proportion of blood-fed mosquitoes collected indoors, most of which had fed on humans, consistent with previous research in Burkina Faso³ and a high level of anthropophily in West Africa.^5,8

Culicidae fauna in the study areas.

Although this study focused on Ae. aegypti, mosquitoes from other genera, including Anopheles and Culex, were also collected. Culex quinquefasciatus, followed by Ae. aegypti, was the most common mosquito species in all HDs, with Anopheles gambiae s.l. being less common. Aedes aegypti was the main Aedes species collected in all HDs and seems to be well established in Ouagadougou. Previous studies have already reported a wide variety and large number of larval containers in the city of Ouagadougou, responsible for the proliferation of this vector.^3,16

Aedes aegypti resting and blood-feeding behavior

A trend towards exophily was observed for Ae. aegypti in the HDs of Bogodogo and Nongremassom, as previously reported by Badolo et al.³ A recent study in Ghana reported that 76% of Ae. aegypti mosquitoes were collected outdoors, compared with 24% indoors.¹⁰

Although an exophilic tendency has been reported for Ae. aegypti in West Africa in most of these recent studies, in our study, the proportions of blood-fed females collected indoors were significantly higher than those outdoors in the study area, as also shown by Badolo et al.³ The reason for this may be that 1) adult females are likely to be endophagic or endophilic or 2) females return indoors at some point in the gonotrophic cycle after feeding outdoors. Further studies should elucidate the feeding behavior of Ae. aegypti, particularly the blood-feeding location and the movement and preferential resting locations post-blood feeding for the development of tailored control strategies against this vector. In contrast, many studies outside Africa have reported Ae. aegypti as typically endophilic, resting inside houses.^30–32

The dengue outbreak in Burkina Faso, combined with the exophilic behavior of the vector, suggests that the outdoor environment is the most likely place of transmission. During the outbreak period, the vector control responses implemented by the Ministry of Health included intradomiciliary spraying in the houses of dengue patients and their neighbors and spatial spraying in public places including hospitals, as well as larval source management. The strategy also includes sensitization of the population to the use of repellents and wearing clothing that covers the body as much as possible.

Our results also showed that the density of blood-fed females collected indoors varied between HDs, with Bogodogo having a higher density than the other two HDs, and the month of September was associated with lower density in Bogodogo. The reason for this situation is not very clear but may be linked to variation in ecology, to predominance of housing types, and very possibly to variation in precipitation between these three HDs.

Blood-feeding preferences of Ae. aegypti were consistent among the three HDs and indicated a very high frequency of feeding on humans. This finding aligns with previous studies conducted in West Africa^3,8,9 and contrasts with laboratory host choice experiments involving guinea pigs versus humans, where populations of Ouagadougou exhibited more flexibility in their host selection.³³ This difference could be due to the lower density of animals than humans in urban areas or variations in sampling methods. By also feeding on domestic animals, Ae. aegypti could contribute to the transmission of zoonotic pathogens, as evidenced by the circulation of dengue viruses in domestic dogs in Thailand,³⁴ or potentially act as secondary vectors of Rift Valley fever during explosive periods of high transmission.³⁵

Factors influencing Ae. aegypti adult abundance.

Several factors associated with mosquito abundance in the house can be grouped into three categories: those related to climatic factors (relative humidity, precipitation), the physical and demographic characteristics of the house (house type and number of residents), and the HDs and the location (indoors/outdoors). Our results showed that the abundance of Ae. aegypti mosquitoes was positively influenced by climatic factors such as minimal relative humidity or 10-day cumulative rainfall, which are correlated predictors, as shown by other studies.^3,36 The number of people sleeping in the house positively influenced the abundance of adult Ae. aegypti mosquitoes, particularly that of the female blood-fed portion. The strong anthropophilic preference of African Ae. aegypti mosquitoes documented in several studies^3,5,37 would explain the effect of the number of house residents on mosquito density and has also been demonstrated in Brazil.³⁶ The presence of vegetation in the household could provide resting places for mosquitoes, and its influence on the density of Ae. aegypti mosquitoes is likely associated with its favorable microclimate, as the model suggests an effect of relative humidity or rainfall.

Relevance to vector control of Ae. aegypti.

The high level of resistance to pyrethroid insecticides in Ae. aegypti from Ouagadougou^38–40 together with its outdoor resting behavior demonstrates that effective control of this vector is likely to be a major challenge. The detection of high proportions of blood-fed female adults resting indoors suggests that at least some stages of the gonotrophic cycle may occur indoors. If so, then indoor residual spraying, preferably a highly refined targeted version,^41,42 could effectively impact the vector population. In contrast, outdoor feeding and resting vectors are notoriously difficult to target effectively, and despite the hope with simple strategies such as attractive toxic sugar baits,⁴³ the most effective means to date has been the release of Wolbachia-infected mosquitoes⁴⁴ and probably fumigation using an appropriate insecticide.

During the last dengue outbreak, the above-mentioned control strategies (intradomiciliary spraying, spatial spraying, and larval source management) were the main responses. However, efforts must be intensified as soon as the rainy season begins, with particular emphasis on reducing the vector’s production sources. This could have a considerable impact on the density of adult populations and transmission of the dengue virus.

CONCLUSION

This study provides valuable information on the resting and blood-feeding behavior of the Ae. aegypti population in Ouagadougou, a city where dengue is endemic and that has experienced recent outbreaks. A significant proportion of adult Ae. aegypti mosquitoes was found outdoors, displaying a notable inclination towards exophily. The pronounced anthropophily of this vector increases the transmission risk of arbovirus diseases, particularly dengue, which circulates almost year-round in Ouagadougou. It is advisable to implement control measures that consider the outdoor resting behavior of Ae. aegypti.

Note: Supplemental material appears at www.ajtmh.org.