Larval ecology and infestation indices of two major arbovirus vectors, Aedes aegypti and Aedes albopictus (Diptera: Culicidae), in Brazzaville, the capital city of the Republic of the Congo

Background Invasive mosquito species, such as Aedes albopictus in Congo can affect the distribution of native species, changing the vector composition and pattern of disease transmission. Here, we comparatively establish the geographical distribution and larval habitat preference of Ae. aegypti and Ae. albopictus and the risk of arbovirus disease outbreaks using Stegomyia indices in the city of Brazzaville, the capital of the Republic of the Congo. Methods Human dwelling surveys of water-holding containers for immature stages of Aedes was carried out in December 2017 in Brazzaville through a random cluster sampling method. A total of 268 human dwellings distributed in 9 boroughs and 27 neighbourhoods were surveyed across the city. Results Overall, 455 potential larval habitats were surveyed. Both Ae. aegypti and Ae. albopictus were collected across the city with an overall high prevalence of Ae. aegypti (53.1%) compared to Ae. albopictus (46.9%). Geographical distribution analysis showed that Ae. aegypti was more abundant (mean = 6.6 ± 1.4) in neighbourhoods located in downtown, while the abundance of Ae. albopictus was low (mean = 3.5 ± 0.6) in suburbs. Peridomestic containers, especially discarded tanks, were the most strongly colonized productive larval habitat for both mosquito species with the prevalence of 56.4% and 53.1% for Ae. aegypti and Ae. albopictus, respectively. Globally, the house index (HI), Breteau index (BI) and container index (CI) were high for Ae. aegypti (26.6%, 38.4% and 22.6%) and Ae. albopictus (33.3%, 49.6% and 26.6%) compared to the transmission risk threshold (5%, 5% and 20%) established by the WHO/PAHO. Overall, pupae-based indices (the pupae index and the pupae per person index) were not significantly different between Ae. aegypti (273.4% and 23.2%) and Ae. albopictus (228.8% and 19.5%). Conclusions The findings of this study suggest a high risk for transmission of arbovirus diseases in Brazzaville and call for an urgent need to implement vector control strategies against these vectors in the Republic of the Congo.


Background
Mosquito-borne diseases such as dengue, chikungunya, yellow fever and Zika are increasing concerns in most tropical and sub-tropical regions. Dengue fever is the most important of these diseases worldwide. One last estimate indicates that, around 390 million dengue infections occur annually leading to 96 million (67-136 million) clinical cases [1]. In 2016, Zika virus was declared as a health emergency of international concern by WHO [2] because of the association of this virus with microcephaly [3], Guillain-Barré syndrome [4] and myelitis [5]. The burden of these arbovirus infections in Africa remains unknown except for yellow fever that a modelling study based on Africa data sources estimated in 2013, 84,000-170,000 severe cases, of which 29,000-60,000 were fatal [6]. Formerly, dengue, Zika, and chikungunya were considered as scarce in West-Central Africa despite of the presence of major vector Aedes aegypti Linnaeus. However, in the past decade, several outbreaks caused by these viruses have been reported in West-Central African countries [7][8][9][10][11][12][13][14] suggesting a possible change in the dynamics of these diseases. These viruses are transmitted to vertebrates including humans mainly by the bite of an infected mosquito belonging to the genus Aedes notably Ae. aegypti and Aedes albopictus (Skuse). Aedes aegypti is native from Africa forests [15], and is nowadays found in sylvan and human-dominated environments. Two subspecies of Ae. aegypti were formally identified by Mattingly [16]. The dark form, Ae. aegypti formosus, confined to the African forest and the light form, Ae. aegypti aegypti, found in human-dominated habitats primarily outside Africa [16]. The Sylvan population of Ae. aegypti breed in natural containers such as rock pools, tree holes and leaf axils [16][17][18] and rarely feed on humans [15], whereas domestic populations feed mainly on humans, mate and rest indoors, breed in man-made containers in and around human dwellings [19]. Generally, Ae. aegypti collected in Central Africa whatever the environment matches Ae. aegypti formosus [20,21] suggesting that in Central Africa, two types of Ae. aegypti formosus, domestic and sylvan, co-occur. On the other hand, Ae. albopictus, originated from south east Asian forest, has invaded all the continents in past 30-40 decades [22]. This species was first reported in Central Africa in the early 2000s [23], and nowadays, is present in almost all countries of the region including the Republic of the Congo [24,25]. Aedes albopictus is the dominant species of Aedes in most urban areas located under 6°N in Central Africa where it tends to replace the native species Ae. aegypti [25][26][27][28]. The coexistence of Ae. aegypti and Ae. albopictus sharing often the same larval habitats has been documented in Central Africa [26,27,29,30]. Nevertheless, it was demonstrated that Ae. aegypti prefers man-made containers located in areas with high building density while Ae. albopictus rather prefers larval habitats surrounded by vegetation [27,30,31]. Interestingly, the emergence of dengue, Zika and chikungunya outbreaks in urban areas in Central Africa coincided with the invasion of the region by Ae. albopictus [32]. Indeed, Ae. albopictus was detected as the main vector during a concurrent dengue/chikungunya outbreak in Gabon in 2007 [9,33]. Both Ae. aegypti and Ae. albopictus were found infected with chikungunya virus during the massive outbreak that occurred in Brazzaville in 2011 with 11,000 cases [13,24]. Recently in 2019, Ae. albopictus was suspected as the main vector during the chikungunya outbreak affecting several locations in the Republic of the Congo [34]. It was also demonstrated that both Ae. aegypti and Ae. albopictus collected in Brazzaville are able to transmit yellow fever virus [35], Zika virus [36], and dengue virus [37]. As there is no efficient vaccine (apart from yellow fever) and specific treatment against these diseases, vector control remains the cornerstone to prevent outbreaks. However, implementing vector control strategies requires extensive background information on the larval ecology of Aedes species. However, in the Republic of the Congo, no such data are available apart from preliminary studies showing the co-occurrence of Ae. aegypti and Ae. albopictus across the country with a predominance of Ae. albopictus in all locations except in Brazzaville [25] and the dominance of Ae. albopictus in two periurban neighbourhoods from Brazzaville irrespective of the season [38]. Here, we present an extensive analysis of the levels of infestation, detailed comparative data of larval ecology, and geographical distribution of Ae. aegypti and Ae. albopictus in Brazzaville, the capital city of the Republic of the Congo to improve the control of these vectors and help prevent other arbovirus outbreaks in this major city.

Study area
Surveys were carried out in Brazzaville (4°16′04′′S, 15°16′31′′E), the capital city of the Republic of the Congo (Fig. 1). The city is located along the Congo River and spans an area of 263.9 km 2 with a population estimation of 1.4 million inhabitants. Brazzaville is laid out concentrically: the downtown is modern with urban buildings containing most administrative and commercial structures, while the suburb is unplanned and sparsely populated. Brazzaville is subdivided into 9 boroughs: Makélékélé; Bacongo; Poto-Poto; Moungali; Ouenzé; Talangaï; M'filou; Madibou; and Djiri. Each borough comprises several neighbourhoods. The climate is humid tropical with four seasons: a short dry season from January to February; a short rainy season from March to May; a long dry season from June to September; and a long rainy season from October to December. The mean annual precipitation ranges between 1500-1800 mm and the mean annual temperature is 25 °C [39].

Sampling and entomological surveys
Entomological investigations were carried out in November/December 2017, corresponding to the long rainy season period in which Aedes mosquitoes are at maximum densities. Surveys were undertaken in clusters of dwellings sampled randomly. Each cluster consisted of 10 dwellings per neighbourhood randomly selected (by drawing lots). In each borough, 3 neighbourhoods were randomly selected by drawing lots, and boroughs were categorized, based on geographical position, the downtown (city centre with high building density) or suburb (periphery area of the city characterised by high vegetation density compare to downtown). The downtown is located in Poto-Poto, Moungali, and Ouenzé boroughs, while the suburb is located in Makélékélé, Bacongo, M'filou, Madibou, Talangaï and Djiri. Two rural neighbourhoods were surveyed in Madibou (Fig. 1, Table 1). During the surveys, each selected dwelling was geo-referenced with a global positioning system (GPS, Garmin Etrex 10), surveyed to record all natural and/or artificial containers with water (potential containers), and those containing at least one larvae or pupae (positive containers). Positive larval habitats were also geo-referenced. For each potential/positive container, the type of container, the volume of container (low, < 5 l; medium, 5-50 l; and high, > 50 l) the volume of water inside the container (low, < 5 l; medium, 5-50 l; and high, > 50 l), the source of the water inside the container (tap or rain) and the quality of water [clear; turbid (cloudy water), and polluted (foul smelling water with organic matter and/ or with the presence of a layer of oil)] inside the container, the presence of plant debris inside the container, the presence of the vegetation (grass, tree or shrub) in the immediate vicinity of the container, the shade (full, partial or none), the colour (transparent, dark or light colour) of the container, the material (metal, plastic, rubber or miscellaneous), the distance between the container and the nearest building (0-5 m, 5-10 m), the distance of container to the nearest plant (grass, tree or shrub) (0-5 m, 5-10 m) and to the ground (0-1 m, 1-3 m) were noted as well the number of inhabitants per dwelling. On basis of the nature, the source, and the use of the water, potential containers were classified into three categories: domestic; peridomestic; and natural. Domestic containers (e.g. storage tanks) were defined as human-filled receptacles, while peridomestic (e.g. discarded tanks and used tyres), and natural receptacles (e.g. rock and tree holes, leaf axils) were those filled by rain [27]. Larvae and pupae found per container were collected, referenced and transported to the insectary and isolated from predators such as Culex (Lutzia) tigripes. Pupae per container were counted, isolated to larvae, and maintained until emergence to adults. Larvae were also reared to adults. Emerged adults per larval habitat were morphologically identified using a suitable taxonomic key [40,41]. The proportion of larvae and pupae, or pupae of each Aedes species was estimated based on the number of specimens identified for each species divided by the total number of larvae and pupae, or pupae of Aedes spp. identified.

Entomological indices
The level of infestation was estimated using traditional Stegomyia indices including the Breteau index (BI, the number of positive containers per 100 surveyed houses), house index (HI, the percentage of houses infested), and container index (CI, percentage of positive containers). Estimated thresholds of HI, BI and CI references were established by the WHO for dengue and yellow fever transmission. Whenever HI > 35%, BI > 50, and CI > 20%, the city is considered as high risk of urban transmission of yellow fever virus, whereas HI < 4%, BI < 5 and CI < 3%, indicated that the city is considered as low risk of the disease transmission [42]. Similarly, low HI < 0.1%, medium HI 0.1-5% and high HI > 5% were established for dengue transmission [43]. Additional indices based on absence/presence and number of pupae were also used including pupae index (PI, number of pupae per 100 surveyed houses) and pupae per person index (PPI, number of pupae per 100 inhabitants). The productivity of pupae in each container type was also assessed as defined (the number of pupae in each container type divided by the total number of pupae in all container types) [44].

Data analysis
All statistical analyses were performed using R version 3.1.5 [45] and R studio version 1.1.463, at α = 0.05 level of significance. Variables defined as categorical variables were expressed by percentages and confidence intervals, and numeric variables were expressed by means and standard deviations. The Shapiro-Wilk test for normality was used to assess the distribution of the data. Because data were not normally distributed, non-parametric statistical tests were used to compare variables; Chi-square test was used for percentages and Wilcoxon rank sum test for means. A binary logistic regression model was used to assess the relationship between larval habitat characteristics and presence of immature stages (larvae and pupae) or pupae only of each Aedes species (Additional file 1: Table S1). Odds ratios (OR) and their 95% confidence intervals (95% CI) were estimated. While a negative binomial regression model was used to assess the relationship between larval habitat characteristics (Additional file 1: Table S1) and number of immature stages and pupae only of Aedes spp. (based on estimates and their standard deviations). Both models were computed first with all factors related to the larval habitat and the environment, and then, each model was refined using a stepwise procedure based on Akaikeʼs information criterion (AIC) [46]. The GPS coordinates of each surveyed house, and each positive container of both species, were projected onto a map using the open-source software QGIS (version 3.4.1 Madeira) [47].

Pre-imaginal infestation
In total, 268 dwellings were surveyed in 27 neighbourhoods across Brazzaville, with 3076 inhabitants (Table 2). Among the surveyed dwellings, 111 (41.4%) were found positive, harbouring at least one positive container of Aedes (larva and/or pupa) ( Table 2). A total of 3787 specimens of immature stages of Aedes were collected,

Types and prevalence of water-holding containers
A total of 455 potential breeding containers for Aedes spp. were surveyed, of which 176 (38.7%) were positive for immature stages of Aedes spp. (Table 3). Containers found during the surveys were grouped into three categories and five types ( Table 3): domestic containers (water storage tanks, and flower-pots); peridomestic containers (discarded tanks, used tyres, and miscellaneous); and natural containers (axils of plants). Analysis revealed that Ae. albopictus (mean ± SD: 4.9 ± 1.2) was significantly more abundant than Ae. aegypti (mean ± SD: 3.6 ± 1.9) in used tyres (Wilcoxon rank sum test: W = 2199, P = 0.007) (Fig. 2). However, no significant difference was found in the prevalence of both species in the other container types (Wilcoxon rank sum test: W = 18094, P > 0.05) (Fig. 2). Discarded tanks were the most prevalent (Table 3), and the most productive water-holding container type for both Ae. aegypti (56.4% of pupae) and Ae. albopictus (53.1% of pupae) (Fig. 2). Nevertheless, no significant difference was observed in the pupae abundance of both species according to the container type (Wilcoxon rank sum test: W = 17414, P > 0.05) except for used tyres where Ae. albopictus pupae abundance (mean ± SD: 2.0 ± 0.6) was significantly higher (Wilcoxon rank sum test: W = 2050, P = 0.013) than those of Ae. aegypti (mean ± SD: 1.3 ± 0.7) (Fig. 2).

Risk factors for the presence and the abundance of Aedes species
Binary logistic regression analysis showed that several factors influence the presence of immature stages of Ae. aegypti and Ae. albopictus as well as the presence of pupae of both species (  (Table 5). Furthermore, negative binomial regression was used to explore the factors which can influence the number of pupae of Ae. aegypti and Ae. albopictus inside the containers. Analysis revealed that the number of pupae of Ae. aegypti was positively influenced by the source of water (tap) (estimate = 4.06, P = 0.0003), the presence of plant residues inside the container (estimate = 9.33, P < 0.0001), the presence of surrounding vegetation (estimate = 15.56, P < 0.0001), the volume of water inside the container (estimate = 4.79, P < 0.0001), the container volume (estimate = 23.68, P < 0.0001), the distance to the nearest plant (0-5 m and 5-10 m) (estimate = 19.56, P < 0.0001 and estimate = 18.94, P < 0.0001, respectively), the distance to the nearest building (5-10 meters) (estimate = 5.69, P < 0.0001), the distance to ground (1-3 m) (estimate = 4.94, P = 0.03), and the neighbourhood location (rural) (estimate = 3.55, P = 0.003). The number of pupae of Ae. albopictus was also positively influenced by the turbid aspect of water (estimate = 1.98, P < 0.0001), the presence of surrounding vegetation (estimate = 1.47, P = 0.0004), and the neighbourhood location (suburb) (estimate = 1.38, P = 0.001) ( Table 6). In contrast, the container colour (light and transparent) (estimate = − 11.27, P < 0.0001 and estimate = − 38.44, P < 0.0001, respectively), the absence of shade (estimate = − 9.75, P < 0.0001), but also the volume of water inside the container (estimate = − 84.72, P < 0.0001), the container volume (estimate = − 13.92, P < 0.0001) and the neighbourhood location (suburb) (estimate = − 25.81, P < 0.0001) were inversely associated to the number of Ae. aegypti pupae. While abundance of Ae. albopictus pupae was negatively associated to the absence of shade (estimate = − 1.12, P = 0.005) and the distance to nearest building (estimate = − 1.69, P = 0.001) ( Table 6).

Discussion
This detailed study presents a comparative analysis of the ecological adaptation of the two major arbovirus vectors Ae. aegypti and Ae. albopictus, in Brazzaville, the major      [24] suggesting its recent introduction, while Ae. aegypti was documented in the country since 1970 [48]. Overall, the prevalence of Ae. aegypti in the city of Brazzaville was higher than that of Ae. albopictus. This is in accordance with the previous data collected across Brazzaville in May 2017 corresponding to the early rainy season indicating that Ae. aegypti was the dominant Aedes species in the city [25]. However, the comparative analysis of the prevalence of both Aedes species across the city of Brazzaville indicates that Ae. aegypti is most prevalent in downtown while Ae. albopictus is most prevalent in suburban and rural areas. Indeed, the dominance of Ae. albopictus in neighbourhoods located in periurban areas in Brazzaville has been previously reported [38]. These observations are in accordance to previous studies in Central Africa which demonstrated that in sympatric areas, Ae. aegypti is most prevalent in neighbourhoods located in downtown with a higher building density, while Ae. albopictus is found more frequently in periurban areas surrounded by vegetation [27,29,30]. Importantly, seasonality can affect the pattern of abundance of both species as demonstrated previously [27,49], perhaps due to the difference in the tolerance of desiccation of the eggs of both species [50]. However, previous data collected in Central Africa suggest that this variation depends to the difference of time between the rainy season and the dry season among locations [30]. Larvae of both Ae. aegypti and Ae. albopictus were found colonizing all container types surveyed across the city. However, peridomestic containers, especially discarded tanks were the most prevalent and the most productive containers for both species. The preference of both species to colonize discarded tanks is in agreement with previous studies in Cameroun [29] and in Central African Republic [27]. These observations are contrary to those generally observed in other regions of the world, particularly in Asia, where Ae. aegypti larvae breed commonly in domestic containers such as water storage tanks [44,51,52]. Unplanned urbanization and lack of waste management can explain the proliferation of peridomestic containers such as discarded tanks and used tyres in Brazzaville. This can explain why Ae. aegypti was found associated with turbid water in this study. The prevalence of water storage containers (27%) observed in Brazzaville suggests a lack of running water in human-dwellings that would promote the storage of water. This prevalence is high compared to that reported in other cities from Central Africa such as Yaoundé and Douala in Cameroon [29].
No significant difference was observed in pupae abundance between the two species according to the container types, except in used tyres for which the abundance of Ae. albopictus pupae was higher than the abundance of Ae. aegypti pupae suggesting that Ae. albopictus infests more used tyres compared to Ae. aegypti. Indeed, the invasion of Ae. albopictus from Asia to other continents was suggested to be driven by the commercialization of used tyres [22,53]. Several container-related factors were impacting the presence and or the abundance of Aedes immature stages notably the turbidity of water which indicates that the presence of organic matter in the water which can supply food resources. In addition, turbidity of water can also serve to hide aquatic stages of Aedes from predators as suggested previously [27,54]. The presence of the vegetation around the container provides shade which reduces the water temperature in the Aedes larval habitat [27,55]. It was demonstrated that the variation of water volume inside the container, can modulate attractiveness to oviposition, space availability and food resources accessibility [54,56,57]. The coexistence of the invasive species Ae. albopictus and the native species Ae. aegypti in the same ecological niche, implies interspecific competition for resources, which leads to segregation of habitats according to macro-environmental variations such as urban environmental gradients, as shown previously [55,58] or the decrease of the abundance of the native species [28,59,60].
The level of infestation of both vector species, was assessed by calculating Stegomiya indices. Traditional Stegomiya indices, HI, BI and CI are commonly used to measure the success of vector control strategy to understand the vector ecology. However, these indices are considered by some authors as a poor predictors of epidemiological risk because they are generally not correlated with disease incidence or outbreak [61]. Based on that, Focks & Chadee [62] suggested that pupae based indices such as PI and PPI, which are more epidemiologically relevant indices, could be better predictors for arboviruses transmission; because of their correlation between total pupal densities and adult densities.
Traditional Stegomyia indices (BI, CI and HI) and pupal indices (PI, PPI) in our study were relatively high for both species across the city of Brazzaville. Although no significant differences were reported between the index values of the two species at the city scale, all index values significantly varied according to the boroughs and the environment (downtown vs suburb). It would be interesting to highlight that this pattern can change according to the season, as mentioned in a previous study in two neighbourhoods of Brazzaville which reported a decrease of Stegomyia indices in the dry season compared to the rainy season [38]. The overall BI (38.4%) and HI (26.5%) estimated in Brazzaville for Ae. aegypti were higher compare to those estimated in previous studies in other cities from Central Africa such as Yaoundé in Cameroon (BI: 10.1%, HI: 9.0%) [29] and Bangui in the Central African Republic irrespective to the season (early wet season: the BI and HI were 14.4% and 9.03%, respectively; late wet season: the BI and HI were 16.5% and 21.8%, respectively) [27]. Similar observations were reported for Ae. albopictus [27]. In addition, index values for both species estimated, were high compared to the reference epidemic thresholds of transmission risk established for yellow fever [42] and for dengue [43]. The high infestation indices of both species suggest a risk for large outbreaks of arbovirus infections such as dengue, yellow fever, Zika, and chikungunya in Brazzaville. Indeed, two large chikungunya outbreaks were reported in the Republic of the Congo during the past decade [13,34]. During these outbreaks, chikungunya virus was detected in both Ae. albopictus and Ae. aegypti in 2011 [24], while Ae. albopictus was suspected as the main vector during the 2019 outbreak [34]. In addition, it was demonstrated that both Ae. aegypti and Ae. albopictus from Brazzaville, are able to ensure yellow fever virus [35], dengue virus [37], Zika virus [36] transmission. To clearly establish the epidemiological importance of each Aedes species in the Republic of the Congo, additional experiments including feeding behaviour patterns, covering additional locations, and spanning several seasons are needed.

Conclusions
Our study revealed high infestation rates of Ae. aegypti and Ae. albopictus across Brazzaville, the major city of the Republic of the Congo, implying a strong potential for human arbovirus infection. Findings generated on the typology, geographical distribution, and productivity of larval development sites of both Aedes species could be useful to implement vector control programmes, including management of larval sources by establishing a targeted discarding of most of the productive larval habitats.