Niger is a landlocked country in West Africa, covering an area of 1,267,000 km². The territory extends between 11˚37’ and 23˚33’ north latitude and between 0˚06’ and 16˚ east longitude from the Greenwich meridian. Geographically, approximately 80% of the territory consists of the Sahara and Sahel, while a southern strip characterised by more abundant vegetation constitutes the Sudanese zone [1]. The latter receives more than 600 mm of annual rainfall, while Niamey, the capital, receives between 300 and 600 mm per year, with highly developed trade. Water is mainly available from the banks of the Niger River, which crosses the country for a distance of 550 km.

A descriptive and analytical longitudinal study was conducted in ten localities of Niger between January and December 2023 to investigate the spatial and temporal dynamics of Aedesmosquitoes and arbovirus circulation: Guidimouni, Tessaoua, Niamey, Liboré, Say, Agadez, Boboye, Madarounfa, Madaoua and Gaya (Figure 1). On the basis of criteria such as the ecology of the area, the density of human populations, the level of urbanisation, the density of vectors and seroprevalence data from previously detected cases of arboviruses, four sites (Niamey, Gaya, Tessaoua and Guidimouni) were selected for entomological monitoring of vectors (eggs, larvae and adults) on a fortnightly basis for one (1) year with a bi-month collection. Each locality is stratified into three (3) zones: an urban zone, a peri-urban zone and a rural zone. For adult mosquito surveillance, the same households were monitored throughout the study period for BG-Sentinel trapping and Prokopack collections. In contrast, larval surveys were conducted in 15 randomly selected households at each bimonthly sampling round.

Source: This map is based on our GPS coordinates recorded during the study’s data collection sessions.

Figure 1. Study sites.

Collectionoflarvae: this was carried out to determine the epidemiological risks of transmission of arboviruses and the typology of Aedesaegypti breeding sites. In each of the 4 sites: Niamey, Gaya, Tessaoua and Guidimouni, larvae surveys were conducted in 15 randomly selected households at each bimonthly sampling round among urban, peri-urban and rural areas. In each house, all water storage containers (Figure 2, Figure 3) and abandoned containers (Figure 4) were inspected, and the geo-referenced coordinates (GPS) were recorded. After collection, larvae were transported to the insectary of Entomology Laboratory of Health and Sanitary Research Center of Niamey and maintained under controlled environmental conditions at a temperature of 27˚C ± 1 and relative humidity of 78% ± 1. Aedes emerged were pooled of 10 or few mosquitoes were formed according to the type of capture and stored in RNALater (SIGMA®) at −80˚C [24].

Risk indices (Breteau Index BI, Habitat Index HI, Container Index CI) were estimated according to the WHO protocol, 2014 (CI > 3%, HI > 4% and BI > 5%) [25]. The number of inspected houses per locality did not reach the WHO recommended threshold of 100 houses, the estimation of the BI may be affected by sampling variability and potential over or ender estimation. Therefore, a Generalized Linear Mixed Model (GLMM) approach was additionally applied to strengthen the robustness of comparisons between localities.

Figure 2. Prospection of peri-domestic containers (water storage).

Figure 3. Container of Aedesbreeding site (abandoned containers).

Figure 4. Container of Aedesbreeding site (watering through).

Eggcollection: Thirty (30) nesting boxes (Figure 5) per locality were laid for one (1) year, with weekly collections of nesting papers for all 10 sites. The egg-laying papers were collected every 7 days and dried for a maximum of 24 hours in the shade. They were then packed in airtight bags to prevent the eggs from drying out and conserved in 27˚C ± 1 of temperature and 78% ± 1 humidity conditions at insectary. Every month, all the egg-laying boxes were observed with a magnifying glass, the eggs counted were placed in water (Figure 6) at the insectary and the density of emerging species was estimated. These eggs will be used to identify the Aedes species present in the locality and to detect arboviruses, insecticide resistance determination. After identification, mosquitoes were pooled according to sex, locality, capture method and collection period to preserve epidemiological stratification and stored in RNALater (SIGMA®) at −80˚C before analysis.

Figure 5. Ovitraps used for this study.

Figure 6. Hatching of Aedeseggs in the insectary.

Collectionofadultmosquitoes: This study took place at 4 sites: Niamey, Gaya, Tessaoua and Guidimouni. At each site, 6 human skin lure BG Sentinel traps were placed every 2 months inside and outside of 3 houses randomly selected for all the study period. Mosquitoes were collected every 12 hours for 2 successive days (Figure 7).

In the same localities, series of aspirations using the prokopack (Figure 8). were carried out over 2 successive days inside and outside of 3 other houses in order to determine the spatio-temporal dynamics of the species and their host-seeking activity and resting behaviour. mosquitoes were pooled according to sex, locality, capture method and collection period to preserve epidemiological stratification and stored in RNALater (SIGMA®) at −80˚C before analysis.

Figure 7. Mosquitoes collection by BG sentinel traps.

Figure 8. Mosquitoes collection by prokopack.

All mosquitoes collected or having emerged were morphologically identified by binocular glass using a dichotomous identification key [24]. After identification, mosquitoes were pooled according to sex, locality, capture method and collection period to preserve epidemiological stratification and stored in RNALater (SIGMA®) at −80˚C before analysis. Therefore, some pools contained fewer than 10 mosquitoes leading in a higher number of pools than expected. For RT-PCR analyses, 500 mosquitoes were collected by BG Sentinel Traps, 200 for prokopack collection, 112 mosquitoes for respectively each egg and larvae emergence were randomly selected. Hence 924 mosquitoes were analyzed by RT-PCR for circulation of DENV, YFV, CHIKV, ZIKV.

2.5.1. Viral RNA Extraction from Mosquitoes

The mosquito pools were removed from the RNALater, transferred to 2 mL Eppendorf tubes containing Leibovitz L15 medium, ground and automatically homogenized using the Tissue-Lyser II (Qiagen, Hilden, Germany 2013-25, LT). The homogenate was then centrifuged at 10,000 rpm at 4˚C for 20 min. Viral RNA was extracted from the supernatant using the QIA amp Viral RNA Mini Kit (Qiagen, Hilden, Germany). Then, RNA was stored at −80˚C until molecular screening.

2.5.2. Viral Detection by Real-Time RT-PCR

The detection of Dengue [26], Yellow Fever [27], Zika [28] and Chikungunya [29] viruses, was screened by real-time RT-PCR on all mosquito shreds using Lyophilized one-step RT-PCR Polymerase Mix (TIB MOLBIOL, Berlin, Germany). The reaction mixture consists of 5 μL of extracted RNA, 10 μL of enzyme, 1 μL (10 μM) of each virus forward primer, 1 μL (10 μM) of each virus reverse primer, 0.5 μL (10 μM) of each virus probe, and 2.5 μL of water, PCR grade. Real-time RT-PCR was performed using Applied Biosystem 2500. The thermal cycling profile of this assay consists of a 10-min reverse transcriptase (RT) step at 55˚C 95˚C for 1 min, and then 40 cycles of 95˚C for 15 s and 60˚C for 30 s. Human positive controls for DENV, ZIKV, YFV and CHIK were used as positive controls to validate the results. Any sample with a Cycle threshold (Ct) ≤ 38 [30] was considered positive due to the low viral load in mosquitoes. The primers and probe used in this study are shown in Table 1.

Table 1. Quantitative RT-PCR primers used for viruses detection.

Larval densities, which were used to determine risk factors, were determined as a function of the total number of larvae collected per type of site.

For the entomological risk indicators that were evaluated, the World Health Organization (WHO, 2014) protocol (CI > 3%, HI > 4% and BI > 5%) [25] was used according to the following formulas:

Breteauindex (BI): corresponds to the number of positive gites per 100 houses visited BI: (Number of positive gites/Number of houses inspected)*100 House index: percentage of houses with at least one positive gite HI: (Number of positive houses/Number of houses inspected) × 100 Container index: Percentage of positive containers (larval gites) among all the containers in water (potential gites) CI: (Number of positive containers/Number of containers in water) × 100 For the determination of the stegomyian indices. The number of positive containers was modelled using a poisson regression farmework with the logarithm of inspected houses. Model fit and overdispersion were evaluated and comparison between localites were performed at a 95% confidence interval with significance defined a p < 0.05. Statistical analysis were performed using Chi-square tests to compare mosquito proportions between localities, collection sites, and behavioural categories. Statistical significance was set at p < 0.05 with 95% confidence intervals. All the analyses. were carried out using Excel and R software.

This study is part of the arboviruses surveillance activities carried out by the Medical and Health Research Centre (CERMES), the Reference National Laboratory of the Ministry of Public Health for 7 diseases including arboviruses in Niger. However, access to households was authorised by the consent of the heads of household before each survey.

Figure 9. Spatio-temporal variation of Aedes presence by ovitraps in Niger.

Aedesspp eggs were detected in seven study sites indicating widespread distribution across Niger. Ovitraps positivity varied by locality and ecological setting, with higher egg densities observed in urban and peri-urban area (Niamey, Gaya, Madaoua and Guidimouni) compared to rural settings (Say, Tessaoua, Boboye). Temporal variation showed increased eggs abundance during the rainy season (June-October-December) with a peak in September. suggesting favourable breeding conditions linked to rainfall patterns. Adult emergence identified Aedes aegypti in 7 sites (Niamey, Gaya, Guidimouni, Tessaoua, Madaoua, Say and Liboré) and Aedesluteocephalus only in Madaoua (Figure 9).

Overall, 360 houses either 90 houses by site were visited and 1620 containers inspected during the study. Of these houses, 23 (6.38%) were found to be infested with mosquito larvae in 40 (2.46%) containers. A total of 2597 larvae were collected and 2403 larvae emerged with 965 for Niamey, 748 for Gaya, 75 for Guidimouni and 615 for Tessaoua.

3.2.1. Typology of Larval Breeding Sites

The types of potential sites encountered at all 4 collection sites were tyres, cans, jars, barrels, water troughs and abandoned containers. In Niamey the sites most infested with Aedes larvae were barrels and cans, while in Gaya it was abandoned containers and barrels that were most positive for larvae. In Guidimouni and Tessaoua, the containers most infested with larvae were cans and jars (Figure 10).

Figure 10. Typology of Aedesspp breeding sites in Niger.

3.2.2. Stegomyian Indices

The Breteau Indices (BI) (20% and 15.55%), the Container Indices (CI) (3.19% and 3.71%) and the Habitat Indices (HI) (20% and 8.89%) for Niamey and Gaya respectively exceed the WHO threshold (CI > 3%, HI > 4% and BI > 5%), while they are relatively low in Guidimouni and Tessaoua (Table 2).

Although the GLMM analysis, using a Poisson Regression model with the number of positive containers as the outcome variable and the number of inspected houses as an offset, demonstrated significant heterogeneity between localities (χ² = 16.6, df = 3, p < 0.001). Pairwise comparisons revealed significantly higher infestation risk in Niamey compared with Tessaoua and Guidimouni.

The GLMM approach confirmed that the elevated infestation observed in Niamey and Gaya was not solely due to sampling effort but reflected genuine spatial variation in Aedesaegypti breeding site abondance.

Table 2. Stegomyian Indices of Aedesaegypti breeding sites in Niger localities.

Note:Legend: House Index (HI), Container Index (CI), Breteau Index (BI), Number of visited houses (nh), Number of positive houses (nph), Number of containers (nc), Number of positive containers (npc).

On the graphic below (Figure 11), a predominance of mosquitoes of the genus Culexsp in all the locations monitored was determined. Among the Aedes mosquitoes species, larvae of Aedesaegypti and Aedesvitattus were collected in Gaya in September and Aedesaegypti in the 3 other localities, Niamey, and Guidimouni with a peak in September. In Niamey larvae were regularly found during all the months of collection except March and September with a peak in January. In Guidimouni, a low density of Aedes species was noted in September, while in Tessaoua, only Culex larvae were found during larval surveys. A few Anophelesspp were collected only in Gaya and Niamey.

Figure 11. Distribution of Culicidae after larval emergence.

The spatial variation in culicine species collected during this study was determined in Table 3. The number of mosquitoes of the genus Culexsp was 81.06%, 29.68%, 79.52% and 77.54% respectively in Gaya, Guidimouni, Niamey and Tessaoua. As for Anopheles species, the density of Anophelesgambiae was 9.24% in Gaya, 36.6% in Guidimouni, 4.42% in Niamey and 10.01% in Tessaoua.

Table 3. Spatial dynamics of Culicidae species in Niger.

According to the Table 3, for all the Culicidae collected in each of the four localities, Aedes aegypti was collected in Gaya, Guidimouni, Niamey and Tessaoua respectively in proportions of (150/1547) 9.69% and (162/896) 18.08%, (417/2598) 16.05% and (174/2267) 7.67%.

Specifically, regarding to the Aedesspecies, Niamey recorded the highest abundance 46.2% (417/903), followed by Tessaoua 19.3% (174/903).

According to the collection areas, Aedesaegypti was collected most in the urban area of Niamey (25.42%), followed by the peri-urban area of the city (7.05%).

Mosquito species distribution differed significantly across study localities (Chi-square test, χ² = 1557.23, df = 15, p < 0.001; 95% CI), indicating marked spatial heterogeneity in Culicidae composition.

A similarity in the temporal dynamics of Aedesaegypti for Gaya, Guidimouni and Tessaoua where in January there is the end of a peak while in March there is a considerable decrease in the population of these species. Between March and May, the population increases and then stabilises from May for Guidimouni and Tessaoua and in September for Gaya. In the city of Niamey, a permanent distribution was observed throughout the collection periods, with an increase from January to March and then a decrease between March and September. After September, the population increased (Figure 12).

Figure 12. Spati-temporal dynamic of Aedesaegypti in Niger.

Aedesaegypti in Niamey and Gaya are much more active between 6 p.m. and 6 a.m. than between 6 a.m. and 6 p.m. whereas in Guidimouni and Tessaoua the trends are indistinguishable. We can therefore deduce that Aedesaegypti in Niamey and Gaya have a higher host-seeking activity at twilight, whereas in Guidimouni and Tessaoua they were both during the day and at night (Figure 13).

Figure 13. Host-seeking activity moment of Aedesaegypti by BG Sentinel Traps in Niger.

In Gaya Aedesaegypti have a host-seeking activities outside than inside the houses respectively (66% vs 34%) and Niamey (77.46% vs 22.54%), whereas in Guidimouni (49.38% vs 50.62%) and Tessaoua (50% vs 50%) it is both outside and inside the house (Figure 14). Thus, we can suggest that Aedesaegypti have a higher outdoor host-seeking activity in Gaya and Niamey while in Guidimouni and Tessaoua they have both outdoor and indoor activity. The spatial distribution of Aedesaegypti host-seeking activity differed significantly between indoor and outdoor environments across localities (Chi-square test, χ² = 63.26, df = 3, p < 0.001; 95% CI). Outdoor host-seeking activity was higher in Gaya and Niamey.

Figure 14. Spatial variation of the host-seeking activity of Aedesaegypti in Niger.

Prokopack collections indicated a higher proportion of Aedesaegypti resting outside habitats regardless of Niamey (68.89%) and Gaya (62.79%) while in Guidimouni and Tessaoua they rest outside and inside habitats. However, this pattern may be influenced by environmental conditions (Figure 15).

Figure 15. Variation of Aedesaegypti resting behaviour.

The distribution of Aedesaegypti was almost uniform between the 4 collection sites except in Niamey where the abundance was very high in the urban and peri-urban area but also permanent throughout the study. This exclusive detection of Aedesaegypti in urban and peri-urban localities, alongside rare detection of Aedesvittatus (Gaya) after larval emergence and Aedesluteocephalus in Madaoua from respectively after eggs emergence underscores the dominance of anthropophilic subspecies Aedesaegypti in transformed environments through the creation of artificial breeding sites [14][31]. The presence of Aedesaegypti, Aedesvittatus and Aedesluteocephalus has been demonstrated in Niger [32][33]. The absence of Aedesalbopictus, the invasive Asian tiger mosquito, is noteworthy and consistent with its limited distribution across West Africa where have been incriminated in the vertical transmission of Zika and the maintenance of the Zika virus has been demonstrated in these sylvatic species in Senegal [34]. In fact, the Aedesaegypti vector is confined to urban and peri-urban environments where its survival is maintained by human populations.

The differential distribution of larval breeding habitats across Niger sites reflects established patterns of urbanization-driven vector ecology in West Africa. Container types such as barrels, cans and abandoned containers were the most productive larval sites, consistent with observations from neighboring Burkina Faso, where similar domestic and peri-domestic containers dominated Aedes breeding ecology during the 2016-2017 dengue outbreaks [14]. These results could be explained by the fact that the type of container influences the density of nymphs [8] also the difference in the level of urbanisation and sanitation in these 2 localities. Niamey is the Niger’s capital and the population enjoys a higher level of sanitation than Gaya. From a climatic point of view, Gaya is located in the Sudanian zone, where rainfall exceeds 600 mm, whereas Niamey receives between 300 and 600 mm/year [35] and trade is highly developed. It therefore goes without saying that the traditional risk indices described by the WHO are higher in these 2 localities, as urbanization, rainfall, sanitation levels and trade are all favourable conditions for the proliferation of Aedes mosquitoes [10][11][14][36][37].

The Stegomyian indices (Breteau Index, Container Index, and Habitat Index) exceeded WHO epidemic thresholds in Niamey (BI = 20%, CI = 3.19%, HI = 20%) and Gaya (BI = 15.55%, CI = 3.71%, HI = 8.89%), confirming substantial entomological risk in these urban and peri-urban zones. To confirm this observation, the consistency between the GLMM results and classical stegomyian indices strongly suggests that the elevated infestation observed in Niamey and Gaya reflects true ecological differences rather than sampling artefacts. Indeed, despite equivalent sampling intensity (90 inspected houses per locality) Niamey and Gaya recorded substancially higher numbers of positive containers compared with Guidimouni and Tessaoua. Similar elevated indices have been reported in other West African cities experiencing arboviral outbreaks [14]. The heterogeneity in breeding site typology across sites reflects localized adaptive responses of Ae.aegypti populations to microhabitat heterogeneity and urban infrastructure variations, a phenomenon documented across Ouagadougou, Ghana, and Senegal [38].

The permanent distribution of Ae.aegypti throughout the year in Niamey contrasts markedly with the seasonal patterns observed in Gaya, Guidimouni, and Tessaoua, where populations exhibited marked declines during dry months (March-May). This phenological divergence reflects both climatic variation and the urban heat island effect that characterizes capital cities in West Africa. Urban areas, with their modified microclimates, reduced water infiltration, and persistent anthropogenic water storage, provide stable breeding conditions regardless of seasonality, whereas peri-urban and rural zones depend on rainfall-fed containers and natural habitats.

Host-seeking activity patterns revealed striking behavioural variation across sites. Niamey and Gaya populations demonstrated marked nocturnal outside activities (77.46% and 66% outdoor host-seeking, respectively), with peak activity between 6 p.m. and 6 a.m., particularly at Conversely, Guidimouni and Tessaoua populations exhibited more mixed inside and outside behaviour, hours. with approximately equal proportions of host-seeking activity indoors and outdoors. This behavioural plasticity has been attributed to rapid urbanization and the evolution of human specialist populations in West African cities over the past 20 - 40 years, driven by increased availability of human hosts and artificial breeding sites twilight [39]. Such behavioral shifts have profound implications for vector control strategy design, as traditional indoor residual spraying programs may be less effective against predominantly outdoor-biting populations [40].

Resting behaviour patterns corroborated host-seeking data, with predominantly exophilic resting in Niamey and Gaya (68.89% and 62.79% outdoor resting, respectively) and mixed endophilic-exophilic patterns in Guidimouni and Tessaoua. These findings are consistent with a broader regional pattern across West Africa, where urban Ae.aegypti populations exhibit pronounced exophily while maintaining residual endophily, necessitating integrated vector control approaches [14]. Environmental conditions, including ambient temperature, humidity, and shelter availability, likely modulate these behavioral patterns, underscoring the need for context-specific surveillance and control strategies [41].

The discrepancy observed in Tessaoua between low domestic stegomyia indices and relatively high abundance of adult Aedesaegypti suggests the household larval surveys may not have capture the most productive breeding habitats. Similar observations have been reported elsewhere, traditional stegomyia indices showed limited correlation with adult mosquuito abundancd and transmission risk [14].

Despite serological in human sera (IgM/IgG) and molecular evidence (by RT-PCR) of DENV, YFV, ZIKV, and CHIKV circulation in Niger’s human population from 1963 to 2024 [22][36][42][43], RT-PCR screening of 924 Ae.aegypti individuals across all life stages yielded no detectable arboviruses. This apparent discrepancy between human seroprevalence and vector infection rates may reflect several mechanisms: (1) temporal asynchrony between mosquito surveillance and human infection peaks; (2) low infection prevalence in mosquito populations due to recent transmission decline; or (3) limited vector competence of Niger’s Ae.aegypti populations for certain virus strains. Similar negative RT-PCR results have been reported in Ghana despite confirmed DENV-2 circulation in human populations [8], suggesting that West African Ae.aegypti populations may exhibit variable or reduced competence for some arbovirus strains compared to populations from endemic regions.

Vector competence studies from the region are crucial for risk assessment. Recent investigations have documented that West African Ae.aegypti exhibit heterogeneous competence depending on virus strain, viral genotype, and population origin [44]. Senegal’s Ae.aegypti populations have demonstrated high disseminated infection rates for local DENV-2 and YFV isolates [44] while populations from Burkina Faso show variable competence contingent on both virus isolate and mosquito strain [45]. The refractory nature historically attributed to West African Ae.aegyptiformosus subspecies (forest-dwelling, zoophilic) has been challenged by recent evidence showing that some West African populations, particularly the anthropophilic Ae.aegyptiaegypti subspecies dominant in urban areas, can efficiently transmit multiple arboviruses [46]. Senegal hosts a population that is highly anthropophilic (78.6% human blood meals) and has demonstrated active dengue circulation [46]. highlighting the regional epidemiological variability.

Niger’s vector surveillance data should be integrated with serological monitoring of human populations and retrospective epidemiological investigations of reported arboviral cases to better assess the public health significance of local Aedesaegypti populations. The absence of arboviral RNA detection in mosquitoes collected during this study may reflect low virus circulation during the survey period, although low infection prevalence and sampling limitations cannot be excluded. Nevertheless, the widespread presence of Ae. aegypti across Niger, particularly in Niamey and Gaya where high vector densities, elevated Stegomyia indices, and year-round occurrence were observed, highlights the potential for arbovirus transmission should viruses be introduced through human movement or cross-border circulation. In addition, the relatively high abundance of adult Ae.aegypti and repeated ovitrap positivity recorded in Tessaoua, despite low domestic larval indices, suggest that important breeding habitats may occur outside surveyed households, indicating that local transmission risk could be underestimated when assessments rely solely on household-based larval surveys.

The heterogeneous distribution of Ae.aegypti across the study sites indicates that vector surveillance and control strategies in Niger should be adapted to local ecological conditions. In Niamey and Gaya, where vector abundance and entomological risk indicators were highest, integrated vector management approaches combining community-based source reduction, improved management of water-storage containers, waste tire disposal, and targeted vector control interventions may be particularly beneficial [40]. These measures are especially important given the predominance of domestic containers and discarded tires among the breeding habitats identified during the study.

Differences in the indoor and outdoor distribution of collected mosquitoes among study sites suggest spatial variability in vector activity and habitat use. Such variations may influence the effectiveness of vector control interventions and highlight the need for locally adapted surveillance strategies [14]. Because adult collections were based on BG-Sentinel traps and Prokopack aspirations, these findings should be interpreted as patterns of mosquito occurrence and capture location rather than direct evidence of biting preference or feeding behavior. Consequently, control strategies should consider both domestic and peri-domestic environments when targeting Ae.aegypti populations in Niger.

Insecticide resistance surveillance is critical. A recent survey in Niamey documented the presence of pyrethroid resistance-associated kdr mutations (F1534C, V1016I, V410L), population, [47] indicating that resistance management strategies should be integrated with vector control programs. Metabolic resistance mechanisms have been documented in Burkina Faso populations [48] and similar investigations in Niger will be essential to guide rational insecticide selection.

This study’s primary limitation was the circumscribed temporal window (2023) and sampling intensity at four intensive surveillance sites, which limits inference to broader seasonal and spatial patterns. Additionally, the study had limited statistical power to detect very low infection rates that are commonly observed in natural Aedes population. The sample size for RT-PCR analysis, while substantial (924 individual Ae.aegypti), may have been insufficient to detect low-prevalence virus circulation, particularly if infection rates in vectors are <1% - 2%. Longitudinal surveillance over multiple years, coupled with human serological and clinical surveillance, is essential to characterize arbovirus transmission dynamics.