Aedes aegypti Abundance, Distribution, and habitat Characteristics in Dar es Salaam, Tanzania Measured Through Two-Year Longitudinal Survey of Larval and Adult Sampling Methods

1. Introduction

Mosquito vectors of arthropod-borne viruses, particularly Aedes aegypti, have gained increasing public health attention due to their role in recurrent and severe outbreaks of arboviral diseases worldwide [1,2]. These species transmit pathogens responsible for dengue fever, chikungunya, Zika, and yellow fever, which are among the most important arboviral infections globally. Aedes aegypti is considered the most efficient vector, largely due to its ability to thrive in or near human domestic environments and its strongly anthropophilic behaviour [3,4]. This species is believed to have originated in Africa and later spread globally, mainly during the transatlantic slave trade era [5]. Two distinct subspecies of Ae. aegypti have been described: Aedes aegypti aegypti (Aaa), a light-colored domestic form, and Aedes aegypti formosus (Aaf), a darker, ancestral forest-dwelling form [6]. While Aaf is primarily associated with a sylvatic transmission cycle of viruses, Aaa predominates in urban environments, where it plays a central role in urban virus transmission [7]. However, despite these ecological differences, increasing evidence indicates that the two subspecies are becoming sympatric, resulting in interbreeding [6,8,9,10,11]. Land-use changes, particularly the conversion of forested areas into human settlements, are thought to drive this shift by increasing contact between sylvatic and domestic populations, potentially promoting sympatry and genetic exchange between the two forms.

Beyond its evolutionary history and ecological adaptation, Ae. aegypti is now widely distributed across all continents except Antarctica [12,13]. Its global spread has largely been facilitated by international trade, particularly through global shipping networks involving used tires and other water-holding containers [12,14] which can harbor eggs and larvae and thereby enhance passive dispersal. The ability of Ae. aegypti eggs to resist desiccation for up to six months facilitates long-distance dispersal, as shipping durations are often shorter than the egg survival time [15]. In addition, rapid urbanization, driven by rural-to-urban migration and poor settlement planning, often characterized by inadequate water supply and sanitation infrastructure, has further facilitated vector expansion [12]. Climate change has also been identified as an important driver of mosquito species distribution [14,16,17]. The species is currently present in over 167 countries worldwide [18], including many that previously did not harbor the vector. Projections suggest that global temperatures may increase by 1.8 to 4.0 °C by the end of the 21st century [17]. The increase in global temperature may expand the geographic suitability of Ae. aegypti in some temperate regions that were previously climatically unsuitable.

In addition to these broad distribution drivers, local ecological factors particularly breeding habitat availability play a critical role in determining Ae. aegypti presence and abundance. The species typically breeds in artificial water-holding containers, including discarded containers, used tires, and domestic water storage vessels [19,20]. Depending on the location of the breeding habitats, the species can breed either indoors or outdoors. Its ability to utilise breeding sites both indoors and outdoors enhances its capacity to colonise a wide range of environments [21]. Importantly, the breeding ecology of Ae. aegypti varies geographically. Populations outside Africa, particularly Latin America and Southeast Asia, are generally more closely associated with human-dominated environments and frequently utilise artificial water-holding containers in and around houses [22,23], whereas African populations exhibit greater ecological diversity, with some populations utilizing natural breeding habitats including tree-holes in addition to domestic containers [24].

While breeding habitat availability influences spatial patterns of abundance, temporal factors such as seasonality also play an important role in shaping Ae. aegypti population dynamics. Vector abundance is generally seasonal, with higher densities observed during the wet season [25,26]. During this period, the risk of pathogen transmission increases as a result of increased mosquito vectorial capacity [27] as result of higher population density, biting rate due to human crowding and mosquito survival. However, this pattern may not consistently apply to Ae. aegypti, particularly in urban settings, where population abundance can remain relatively stable throughout the year [28]. This stability is largely driven by human activities such as water storage and improper disposal of water-holding containers such as tyres, which ensure continuous availability of breeding habitats [28]. In many African cities and other developing countries, water scarcity is common, necessitating the storage of water in large containers, which serve as ideal breeding habitats [19]. These human behavioral practices create persistent breeding opportunities, enabling vector populations to persist during both wet and dry seasons and consequently facilitating continuous virus transmission.

Currently, vector control remains a key strategy for preventing arboviral diseases such as dengue, particularly in the absence of widely available antiviral therapies and universal vaccines. Common control strategies typically involve integrated approaches, including insecticide application, source reduction, environmental management, and vector surveillance [29,30,31]. In recent years, use of population-level innovations such as release-based approaches, including Wolbachia-infected mosquitoes, sterile insect technique (SIT) and gene drive systems [32,33], has expanded the vector control arsenal. However, in endemic regions with constrained resources, large-scale implementation of these strategies remains challenging. Therefore, targeted interventions at high-risk locations such as airports and seaports as well as hotspots have been proposed as a cost-effective complementary approach. These sites may serve as important points for the introduction of arboviruses and invasive vector populations, while also providing opportunities for monitoring established populations and early detection of new introductions [34]. Furthermore, because Ae. aegypti exploits a wide range of breeding habitats, with considerable variation in productivity among container types [20], prioritizing the most productive breeding habitats may further enhance the cost-effectiveness of control interventions.

In Tanzania, arboviral outbreaks such as dengue fever are increasingly reported [19,35,36,37,38], with Ae. aegypti identified as the primary vector [19,39]. Although several studies have documented the occurrence of Ae. aegypti species and the burden they cause, comprehensive information on seasonal variation, habitat productivity and fine-scale spatial distribution of Ae. aegypti remains limited. Such ecological data are critical for designing effective control strategies, as they help to identify the most productive habitats and optimal timing and locations for interventions. Therefore, this study aimed to assess density, spatiotemporal distribution, and breeding habitat preferences of Ae. aegypti, in order to inform evidence-based vector control programs.

2. Materials and Methods

2.1. Study Area

The study was conducted in Dar es Salaam City [39], the largest commercial city in Tanzania. Located at 6.48°S and 39.17°E along the Indian Ocean coast, the city has an estimated population of nearly 5.5 million [40]. It comprises five administrative districts (Ilala, Kigamboni, Kinondoni, Temeke, and Ubungo), with a total of 70 wards. The city is an epicentre of dengue outbreaks, with transmission occurring during the short rains (November-December) and the long rains (March-May). Based on previous dengue outbreak reports [19,41], Ilala, Kinondoni, and Temeke districts were selected for this study. The city experiences high temperatures year-round, peaking between October and February, and receives an average annual rainfall of approximately 1100mm.

2.2. Study Design and Sites

A repeated sampling design (longitudinal survey) was conducted over two-years (2022-2024) to collect Aedes mosquitoes (SOM Figure S1). The sampling was performed only outdoors due to unwillingness and unavailability of household owners that lead to indoors sampling infeasibility. Longitudinal sampling was chosen over cross-sectional designs as it is more effective in predicting the transmission risk of dengue virus (DENV) [42] and other arboviral diseases. In each district, a multiple stage sampling approach was used to select four wards: two representing high-income areas and the other two representing low-income areas [43,44] Table 1.

High-income wards were characterized by well-constructed houses, improved sanitation infrastructure, paved roads, and reliable piped water whereas the low-income wards were characterized by poorer housing, inadequate sewage systems, and unpaved roads. Within each selected ward, 20 households were randomly selected from lists obtained from the ward council which were assigned numbers. If the household owner of the selected house, refused to take part in the study, the adjacent house was recruited. Households were spaced at least 100m apart in each hamlet, and sampling covered a radius of up to 5km from a central point which was either ward market or ward council office. To increase the likelihood of capturing Aedes mosquitoes, additional sampling was conducted in all storm drains, public spaces, industrial areas, and motor vehicle repair garages available in each ward within 5km of sampling. These locations are known as potential Ae. aegypti breeding sites.

2.3. Household Recruitment

Each selected household was visited, and a head of the household was provided with an informed consent form requesting participation in the study. Participants were informed about the purpose of the study, their role and their right to withdraw at any time. Following consent, the head of household or other adult representative was administered a structured questionnaire adapted from [45] to collect data on household characteristics including wall and roof materials, number of occupants, and socioeconomic status. Technicians conducted a rapid inspection to identify and record the number and types of potential mosquito breeding sites currently on the premises [19]. All sampling locations (households, storm drains, public spaces, industrial areas, and vehicle garages) were geo-referenced to generate a spatial map of sampling sites. Each sampling location was visited once per month throughout the study period.

2.4. Mosquito Collection

Adult mosquito collections were conducted using four entomological sampling methods: BioGent AG Sentinel (BG) traps (Biogents AG, Regensburg, Germany) targeting host-seeking Aedes mosquitoes, Ovitrap (modified Biogents Gravid Aedes Trap (BG-GAT; Biogents, Germany)) for gravid females, and Prokopack aspirations for resting adults. In each sampling round, a combination of BG-Sentinel and BG-GAT was deployed outdoors in one household for 24 hours and inspected the following day. After the inspection of aforementioned traps, resting mosquitoes around the household premise were also collected using a Prokopack aspirator. All collected mosquitoes were transported to a temporary field laboratory and morphologically identified to species level using the identification key by Wilkerson et al. 2021 [18]. All Ae. aegypti were sorted by sex and physiological status (unfed, blood-fed and gravid), then preserved individually in Eppendorf tubes containing RNAlater prepared locally at Swiss Tropical and Public Health (TPH) Institute as described by Tenywa et al. 2025 [39]. Samples were then arranged in 96-well boxes, placed in a cooler box, and transported to the Ifakara Health Institute laboratory for storage at −80 °C. Following adult collection, larval surveys were conducted by inspecting all water-holding containers for the presence of larvae and pupae using dippers for larvae and pupae. Upon examination, the number of immature mosquitoes per container was recorded.

2.5. Data Analysis

Data analysis was performed using STATA software (StataCorp LLC, College Station TX, USA) and R package (version 4.5.3; R Core Team, 2026).

2.6. Vector Density

Descriptive analysis was performed to summarise habitat characteristics, adult and immature mosquito densities at district level. A mixed-effect negative binomial regression model was used to estimate incidence rate ratio (IRR) and 95% confidence intervals (95%CIs) for mosquito density [46] as part of a sensitivity analysis, and the Breteau Index (BI: number of positive containers with larvae/pupae per 100 inspected houses) across districts. A binary logistic regression was used to estimate the House Index (HI: percentage of houses positive for Ae. aegypti larvae or pupae) and Container Index (CI: percentage of containers infested with Ae. aegypti larvae or pupae). Statistical significance was assessed at the 5% significance level and 95% confidence intervals were reported. Furthermore, scatter plots were used to assess the correlation between the number of larvae collected and the number of adult mosquitoes captured by each trap type. Pearson’s correlation coefficient was calculated, and a linear regression line was fitted to illustrate the relationship between larval and adult mosquito abundance.

2.7. Spatial Distribution of Vectors

All spatial analyses and visualisations were conducted in R (version 4.5.3; R Core Team, 2026) [47] using the readxl, dplyr, lubridate, sf, ggplot2, and ggspatial packages. Mosquito capture records were aggregated to the household level by summing trap counts per unique GPS coordinate pair, and subsequently to the ward level for choropleth mapping. All spatial objects were projected to WGS84 (EPSG:4326) prior to analysis, and ward-level Ae. aegypti capture totals were joined to the 2023 Dar es Salaam ward boundary shapefile by normalized ward name keys. Choropleth maps were generated for the combined study period and stratified by dry and rain seasons, with a shared continuous color scale across seasonal panels to enable direct visual comparison.

2.8. Temporal Distribution of Vectors

Descriptive analyses were conducted to examine temporal trends in mosquito abundance. The relationship between vector density and seasonal variability (e.g., temperature and humidity) was assessed using a generalised linear mixed model with log link.

3. Results

3.1. Mosquito Density

A total of 220,947 mosquitoes were collected over a two-year period. Of these, 3.8% (n = 8,496) were Ae. aegypti, 0.1% (n=198) were Anopheles gambiae complex and 96.1% (n=212,253) were Culex quinquefasciatus.

Temeke district had a significantly higher Ae. aegypti density (IRR=2.14; [95%CI: 1.59-2.88], p<0.001), accounting for 49.9%(n=4,235) of all Ae. aegypti collected. This was followed by Ilala (25.8%, n=2,195), with Kinondoni having the least density (24.3%, n=2,066), (IRR=0.23; [95%CI: 0.16-0.33], p<0.001) (Table 2). Overall, Ae. aegypti density was nearly twice as high in the wet season compared to the dry season (IRR= 1.72; [95%CI: 1.48-1.98], p<0.001) (Table 2).

Aedes aegypti density varied across wards within each district (Figure 1 & SOM Table S1) with wards categorised as high income constituting higher density than low-income category in all districts. Within Temeke district, Chang’ombe ward recorded the highest Ae. aegypti density (59.0%, n=2,497). Within Ilala district, Kivukoni had the highest density (46.4%, n=1,018), while within Kinondoni district, Kijitonyama ward recorded the highest density (72.1%, n=1,488) (SOM Table 1). A similar pattern was observed for C. quinquefasciatus across districts, except in Kinondoni, where Mbweni ward had a higher C. quinquefasciatus density despite a lower Ae. aegypti density.

Seasonal analysis showed that Ae. aegypti density was significantly higher during the wet season in Ilala (IRR=2.48; [95%CI: 1.94-3.18], p<0.001) and Temeke (IRR=2.08; [95%CI: 1.50-2.89], p<0.001), but not in Kinondoni (IRR=1.29; [95%CI:1.00-1.67], p=0.064) (Table 3).

3.2. Larval Indices

Of the three districts, Kinondoni showed the highest larval indices across all parameters: Container Index (CI) (OR=6.92; [95%CI:5.30-9.03], p<0.001), House Index (HI) (OR=12.65; [95%CI: 10.05-15.94], p<0.001) and Breteau Index (BI) (IRR=16.53; [95%CI:12.72-21.46], p<0.001) (Table 4). Overall, significantly lower CI, HI and BI were observed during the dry season compared to the wet season (Table 4).

3.3. Aedes aegypti Habitat Positivity

A total of 42,596 potential breeding habitats representing five habitat categories were identified and inspected for the presence of Ae. aegypti larvae (Table 5). Overall, 2.5%(n=1070) of the inspected habitats were positive for Ae. aegypti larvae. Positivity varied among habitat types, with used tyres having the highest proportion of positive habitats (5.7%), followed by water-storage containers (3.8%). Although water-storage containers had a lower positivity rate than tyres, they accounted for the largest number of Ae. aegypti-positive habitats and contributed the highiest proportion of the total larvae collected (Table 5 ; SOM Figure S2). Cans and coconut shells were the most abundant habitats types but had the lowest positivity rate (1.4%). Habitat size analysis showed that medium-sized habitats were the most common across the study sites; however, small-sized habitats measuring <50 cm in depth exhibited the highest positivity rate for Ae. aegypti larvae (18.7%) (Table 5). Clean-water habitats constituted the majority of inspected habitats but turbid habitats had a higher positivity rate (17.6%) than very turbid (9.7%) or clear-water habitats (8.8%). Overall these descriptive findings suggest that habitat type, size and water turbidity may influence the likelihood of Ae. aegypti infestation.

3.4. Spatial Distribution

Ward-level aggregation of Ae. aegypti captures across the 12 surveyed wards revealed marked spatial heterogeneity in vector abundance over the study period (2022–2024) (Figure 1). By urban zone, the highest cumulative captures were recorded in the southern and central wards of Dar es Salaam, with Chang’ombe (n = 2,497), Temeke (n = 1,552), and Kijitonyama (n = 1,488) collectively accounting for the greatest vector burden. Wards in the inner city zone, including Kivukoni (n = 1,018), Upanga Magharibi (n = 816), and Makumbusho (n = 251), recorded moderate capture counts. The lowest captures were observed in the peripheral wards, comprising the northern coastal wards of Bunju (n = 130) and Mbweni (n = 197), the southern peri-urban ward of Chanika (n = 204), and the southern coastal wards of Kurasini (n = 58) and Charambe (n = 128) (SOM Table 1).

Stratification by season revealed that the spatial pattern of Ae. aegypti abundance was broadly consistent between the dry and rainy seasons across all surveyed wards (Figure 1 (B&C)). Chang’ombe, Temeke, and Kijitonyama maintained the highest capture intensities in both seasons, while peripheral wards including Chanika, Bunju, and Mbweni recorded low captures across both seasons.

3.5. Temporal Distribution

The results showed that Ae. aegypti were present throughout the year. Increased densities were observed during the wet months (March-May) in both the first and second years of collection (Figure 2). It was shown that, Temeke district exhibited higher densities during the peak months of Ae. aegypti compared to the other two districts (Figure 2).

3.6. Correlation Between Larval and Adult Ae. aegypti

Assesment of the relationship between larval and adult Ae. aegypti abundance by trap types showed no significant correlation between larval density and the number of mosquitoes captured by different trap type (Figure 3). However, mosquito capture varied among trap type, with BG-traps capturing the highest number of adult mosquitoes, followed by prockopack aspirators, across all districts (Table 6).

3.7. Environmental Data

Temperature and relative humidity were recorded throughout the study period using a data logger installed at one of the study sites. The overall mean temperature and relative humidity were 28.6 °C and 67.7% respectively, with values ranging from 15.3 °C–34.7 °C and 35.5%–96.6% respectively (Figure 4). On average both temperature and relative humidity were within the optimal ranges generally considered suitable for Ae. aegypti survival and development. Because the three study sites were geographically close to one another, a single data logger was used assuming that spatial variation in temperature and relative humidity among the sites was minimal.

4. Discussion

4.1. Species Composition of Aedes mosquitoes

East and West Africa are located within tropical and subtropical regions and share many ecological and climatic characteristics that favour the establishment and persistence of mosquito vectors. Despite these similarities, the distribution of Aedes species differs markedly between the regions. In West Africa, both Aedes aegypti and Aedes albopictus co-exist [8,12,13,14,48,49,50], whereas in East Africa studies have reported the presence of Ae. aegypti with limited or no evidence established for Ae. albopictus populations [14,19,28,51]. In the present study, we conducted a two-year longitudinal survey and only Ae. aegypti among Aedes species was identified throughout Dar es Salaam during the study duration, corroborating previous reports from other parts of Tanzania [52,53,54]. Similar findings have been reported in neighbouring country Kenya, where ecological and dispersal studies have consistently documented Ae. aegypti as the predominant species [6,21,24,28,55,56,57]. Furthermore, genetic diversity studies have further revealed the presence of both Aedes aegypti aegypti (Aaa) and Aedes aegypti formosus (Aaf) subspecies [10,28]. While Aaf is predominantly found in western Kenya, both subspecies occur in eastern part of the country [10,28]. Given the extensive geographic connectivity between Tanzania and Kenya mainly through trade, migration, and sociocultural interactions, it is plausible that both subspecies may also be present in Tanzania. This hypothesis, however, requires confirmation through genomic investigations. Such information is epidemiologically important because Aaa is highly synanthropic and plays a major role in urban arboviral transmission [58], whereas Aaf is generally associated with sylvatic environment and wildlife transmission cycles [58,59].

4.2. Spatial Distribution and Abundance of Ae. aegypti

Our findings demonstrated substantial heterogeneity in Ae. aegypti abundance across Dar es Salaam. The abundance and distribution of Ae. aegypti are largely influenced by habitat suitability and human behavioural practices [60,61]. The species preferentially breeds in water accumulated in artificial containers, including discarded containers, water-storage vessels, used tires and flower pots [19,20,28,62,63]. Among these habitats, water-storage containers and used tyres have frequently been reported as the most productive breeding habitats [28,64]. These reports are broadly consistent with our findings, which identified used tyres and water-storage containers as habitats with relatevely high Ae. aegypti infestation rates, at 5.6% and 3.8%, respectively. The higher infestation rate observed in used tyres suggests that they may be particularly important breeding habitats where they are present and retain water. Therefore, the Tanzania vector control programmes should consider targeting used tyres in Ae. aegypti control interventions, particularly in communities where these habitats are common. Control of Ae. aegypti breeding in used tyres could include approprate disposal, recycling or repurposing them as mosquito control devices, such as lethal ovitraps including Ovillanta ovitrap, which are designated to attract gravid Ae. aegypti [65,66]. Such approaches may provide an opportunity to transform an important breeding habitat into a vector-control tool. Furthermore, this study demonstrated lower infection rate in water-storage containers than used tyres, despite having a lower infestation rate, water-storage containers were considerably more abundant and accounted for a larger number of positive habitats as well as high proportion of larvae. Thus, their overall contribution to Ae. aegypti abundance may be substential because of their high availability in the study communities. This finding highlights the importance of considering both habitat-specific infestation rates and the abundance of available habitats when prioritising breeding sites for Ae. aegypti control.

In this study, Temeke district accounted for nearly half of the collected mosquito populations. The high mosquito abundance observed in the district may be attributed to a combination factors including demograph, infrastructure and behaviral practice. According to the 2022 Tanzanian census, Temeke is one of the most densely populated districts in Dar es Salaam and has a limited access to reliable piped water and sanitation infrastructure [40]. Consequently, residents commonly store water in drums, buckets and other containers for domestic use, potentilly increasing the availability of suitable oviposition habitats for Ae. aegypti. The district also exhibits numerous vehicle repair garages spatially distributed across the area and a relatively high number of warehouses [40], which may contribute to the accumulation of discarded tyres that can serve as productive Ae. aegypti breeding habitats [67]. Furthermore, mapping of Aedes larval habitats using drone imagery and supervised machine learning has documented the placement of tyres on household roofs in the district, reportedly for storm protection [67], further increasing the availability of suitable larval habitats [62,67].

Interestingly, a recent study on dengue virus prevalence in mosquito reported the highest infection rates in Ae. aegypti mosquitoes from Temeke district [39]. Although adult mosquito density alone does not necessarily predict dengue transmission intensity [42,68], vector abundance remains an important component of transmission dynamics because the basic reproduction number of mosquito-borne pathogens is strongly influenced by vector density [69,70]. Consequently, the elevated mosquito densities observed in Temeke may contribute to an increased risk of arboviral transmission in this district. Furthermore, at the ward level, high-income wards generally reported higher mosquito densities than low-income wards within the same district. Although these areas often have better water infrastructure, intermittent water supply encourages residents to store water in large containers including septic tanks that may remain filled for prolonged periods, thereby facilitating mosquito breeding. Similar observations have been reported in Puerto Rico, where Ae. aegypti abundance remained high despite relatively good water infrastructure [25], water storage containers were highlighted as the main source of the mosquito infestation. Currently, source reduction remains the primary WHO-recommended strategy for Ae. aegypti control [71,72] while interventions including larviciding, adulticiding and emerging approaches such as Wolbachia-based infected mosquitoes, sterile insect technique, and genetic modification technologies are recommended as complementary tools [72]. The success of these interventions depends on accurate information regarding breeding habitats, resting sites and population dynamics. Therefore, the spatial distribution and density estimates generated in this study provide important evidence for guiding targeted vector surveillance and control programmes in Dar es Salaam.

4.3. Seasonal Dynamics of Ae. aegypti

Mosquito abundance generally tends to vary seasonally, with peaks occurring during rainy periods when breeding habitats become more abundant due to rainfall [19,26,48,73]. However, this pattern is not universal for Ae. aegypti. Several studies have demonstrated relatively stable year-round populations in urban settings despite seasonal variation in rainfall [25,74,75]. This persistence is often attributed to permanent or semi-permanent breeding habitats, including water-storage containers and cryptic habitats such as septic tanks [76,77]. The reported observations are in conformity with findings in this study particularly in Kinondoni district, where Ae. aegypti population remained relatively stable throughout the year, sugesting that Kinondoni may be exhibiting permanent, human-maintained aquatic habitats that buffer the mosquito population between seasons, thus highlighting potential different timing for controlling the mosquito species in this district in respect to the two districts. Given the chronic challenges in domestic water supply experienced in Dar es Salaam, it is plausible that water storage containers constitute the principal breeding habitats sustaining year-round mosquito populations. This interpretation is echoed by our habitat productivity data, which identified water-storage containers as the most abundant and productive breeding sites (Table 5). Because the prevention of Aedes-borne diseases especially dengue, mainly relying on vector control, targeted management of water-storage containers should be prioritised in these settings, particularly through covering water-storage containers with tightly fitting lids or mesh screens, or larviciding the containers as well as mosquito proof containers [71,72]. Numerous studies have demonstrated that these interventions can substantially reduce Ae. aegypti populations and consequently lower arboviral transmission risk [78,79,80,81].

4.4. Larval Indices and Their Utility in Predicting Adult Mosquito Abundance

House Index (HI), Container Index (CI) and Breteau Index (BI) are the conventional indices used as indicators of Aedes mosquito infestation in an area [82,83]. These indices quantify household infestation, breeding-site distribution, and container positivity [82,83], and have historically been used as a convetional instrument for estimating arboviral outbreak risk [68], where when HI and BI is above 1% and 5 thresholds respectively, indicates potential risk for dengue outbreaks [82,83,84]. In our study, all three indices were lower than those previously reported in Dar es Salaam [19,85]. Among all districts, Kinondoni district exhibited the highest values for each index (Table 4), consistent with previous findings [85]. Notably, however, Kinondoni did not exhibit the highest adult mosquito density, instead, Temeke had the largest adult mosquito population. This discrepancy highlights the well documented limitations of HI, CI and BI as predictors of adult vector abundance [68,86], suggesting that operational control of the species in these sites should combine adult surveillance with habitat-productivity measuremnts. This is explained by factors such as larval mortality and lifespan, which lead to inconsistency between the observed immature abundance measured and the corresponding adults [42,87]. Thus, pupal-based indicators such as Pupae Index (PI), Pupae per Person Index (PPI) and Pupae per House Index (PHI) have been proposed as more informative predictors of adult vector productivity [86,88].

All three entomological indices (HI, CI and BI) are strongly influenced by seasonality, with previous studies reporting higher values during the rainy season [53,89]. The present study demonstrated a similar seasonal pattern of the indices. This increase is likely attributed to greater availability of water-filled containers due to rain that serve as breeding habitats for Ae. aegypti, leading to higher mosquito densities. Consequently, the risk of arboviral transmission may increase during rainy season as vector abundance is an important determinant of the transmission potential. Consistent with this argument, most dengue outbreaks reported in Tanzania have occurred during the rainy months, particularly between March and May [19,37], when the country experiences prolonged and heavy rainfall. These findings highlight the importance of strengthening vector control interventions, with particular emphasis on the removal or management of temporary water-holding containers especially before the onset of the rains [19,67].

4.5. Breeding Habitat Types

Inspection of potential breeding habitats in this study revealed that cans and coconut shells were the most abundant habitat types. However, tyres and water-storage containers exhibited the highest positivity rates for Ae. aegypti larvae, indicating that Aedes productivity does not depend on habitat abundance but rather on habitat type. Similar findings have been reported previously in Dar es Salaam [19] and elsewhere [25,28,64]. Therefore, targeting water-storage containers and tyres for Ae. aegypti control program in Dar es Salaam settings is an essential aspect. Aedes aegypti generally prefers to oviposit in artificial containers containing clean to moderately organic water [90]. Studies indicate that the species commonly breeds in water ranging from near-neutral to mildly alkaline pH [91,92], whereas highly acidic or alkaline conditions can reduce egg hatchability, larval survival and overall habitat suitability [91]. In this study we observed that habitats containing turbidy water and ≤50 cm in depth were particularly favourable for Ae. aegypti productivity. This indicates that water type plays a significant role in the species’ productivity. Therefore, breeding habitat with turbidy water should be the main targets when conducting Ae. eagypti programs in the city. However, further studies should characterise water pH and alkalinity to better understand the best parameter range required for the species in the study site.

4.6. Climate Drivers of Ae. aegypti Abundance and Future Implications

Climate change is increasingly recognised as an important driver of the geographic expansion and transmission potential of Aedes mosquitoes. It impacts aspects such as rainfall patterns, humidity and temperature, causing potential environmental catastrophes such as floods or droughts and global warming [14] which can alter mosquito vector ecology and disease risks [14,16,17]. While rainfall contributes to the formation of breeding habitats, temperature and humidity directly influence mosquito development, survival, reproduction and vector competence [93,94,95]. These factors can also accelerate viral replication within mosquitoes by shortening the extrinsic incubation period (EIP) [96,97]. For example, modelling studies suggest that the dengue virus EIP may decrease from approximately seven days to five days at a temperature of around 32 °C [96]. In the present study, the average temperature and relative humidity were 28.6 ^°C and 67.7% respectively, conditions within optimal range for Ae. aegypti survival and reproduction. Given the projections that global temperature may continue to increase substantially during the twenty-first century [16,17,98], these findings raise concerns regarding the continued expansion of Ae. aegypti populations and the potential emergence of arboviral transmission in areas that were previously unsuitable for the species.

4.7. Study Limitations

The sampling of both adult and immature mosquitoes was conducted exclusively outdoors because the household owners refused access to their indoor environment. This limits our ability to determine whether the observed mosquito density accurately reflects the overall Ae. aegypti population in the study area, particularly given the species’ ability to utilise both indoor and outdoor habitats.

Entomological indices such as Pupae Index (PI), Pupae per Person Index (PPI) and Pupae per House Index (PHI) have been proposed as measures more directly related to Ae. aegypti productivity and abundance. In this study, we used House Index (HI), Container Index (CI) and Breteau Index (BI), which primarily measure the extent of larval habitat infestation rather than adult mosquito abundance. This difference may partialy explain the apparent discrepancy between adult Ae. aegypti abundance and larval indices across districts. For example, Temeke had the highest adult mosquito abundance but relatively lower HI, CI, and BI values, where Kinondoni had higher larval indices but lower adult mosquito abundance. This suggests that HI, CI and BI do not reliably reflect adult mosquito abundance, as adult abundance is also influenced by factors such as habitat productivity, larval survival, adult emergence, dispersal and mortality.

4.8. Recommendations

Overall, this study demosntrated substantial spatial heterogeity in Ae. aegypti abundance across Dar es salaam, with mosquito density varying considerably between districts and among wards within district. Adult mosquito abundance was also higher in areas classified as high income than in low-income areas. These findings suggest that Ae. aegypti surveillence and control programmes should adopt a risk-based approach that prioritise areas with consistently high mosquito abundance while maintaining minimal efforts in low-density areas. Such interventions should also account for seasonal variation in mosquito abundance, because adult Ae. aegypti abundance increased during the rainy season, therefore intensified surveillance and control activities could be initiated before and during the onset of the rains to reduce mosquito population before the peak is reached.

The study also demonstrated that larval indices did not consistently reflect adult Ae. agypti abundance across districts. For example, Temeke had relatively high adult mosquito abundance despite lower larval indices, whereas Kinondoni showed the opposite pattern. Therefore, reliance on larval indices may not adequately capture adult mosquito abundance in Dar es Salaam. Thus, vector surveillance programmes should consequently prioritise adult moquito surveillence especially where the primary objective is to estimate adult mosquito abundance or identify areas at increased risk of dengue transmission. Nevertheless, larval indices can remain useful for assessing breeding site infestations and evaluting source-reduction activities.

Moreover, the study further identified used tyres and water-storage containers as important Ae. aegypti breeding habitats. Control programmes in Dar es Salaam should therefore prioritise these habitat types through targeted source-reduction and larval source-management activities. For example used tyres should be removed, properly stored, covered, recycled or managed to prevent water accumalation while on the other hand water-storage containers should be tightly covered or fitted with appropriate screens and regualrly emptied, cleaned or treated where necessary. For successifulness of these interventions, they should be supported by community engagement and behavioural-change activities.

5. Conclusion

This study demonstrated that Ae. aegypti is widely distributed across Dar es Salaam, with Temeke district exhibiting the highest mosquito abundance. The observed spatial distribution and abundance of Ae. aegypti indicate a continued risk of arboviral disease transmission in the city, particularly in Temeke district, where vector densities were the highest. Given that vector density is a key component of mosquito vectorial capacity, routine vector surveillance and control interventions should be maintained throughout the city, with particular emphasis on high-density areas. The study further showed that Ae. aegypti persist throughout the year, highlighting the importance of permanent and semi-permanent breeding habitats, especially water storage containers in sustaining vector populations. Interventions targeting these habitats, including covering water storage containers with lids or mesh, and larval source management through larviciding, may contribute to reducing year-round vector persistence. In addition, the study revealed inconsistencies between conventional entomological indices (HI, CI and BI) and observed mosquito density particularly in Temeke and Kinondoni districts.