The selection of study sites was based on previous entomological surveys and the degree of urbanization.

The commune of Kintele (4˚9'0''S, 15˚20'32''E) is situated 25 km north of Brazzaville, with an average altitude of 201 m. The estimated population is 11,105 inhabitants. Economic activities are primarily supported by small and medium-sized enterprises and state-owned markets. Housing structures consist of bricks buildings with sheet metal roofs. The relief is made up of valleys, plains, hills and plateaus. The climate is tropical and humid, characterized by distinct wet (October to May) and dry (June to September) seasons. Annual average rainfall and temperature are 1370 mm and 25.5˚C, respectively. Kintele is intersected by two rivers, the Djiri and the Blue Chatelet. The predominant vegetation is savannah trees, sometimes accompanied by shrubs [9]. A study conducted in 1987 on malaria transmission indicated that the Anophelesgambiaes.l. complex contained the primary malaria vector species in Kintele, namely An.gambiaes.l. and An. funestus [10].

The village of Djoumouna (4˚22'34''S, 15˚9'36''E) is located 25 km southeast of Brazzaville in a rural area at an average altitude of 217 m [11]. The estimated population is 635 inhabitants. Main activities in this region encompass agriculture, fish farming and subsistence livestock. Dwellings are clustered and constructed from cement or clay bricks, with sheet metal and cob roofs. Djoumouna is located within a region of degraded secondary forest. The climate is humid tropical, marked by a rainy season (October to May) and a dry season (June to September). Data provided by the National Civil Aviation Agency between September 2016 and August 2017 [12] indicates average monthly rainfall and temperature of 134 mm and 26.1˚C, respectively, with an average maximum relative humidity of 90%. Djoumouna is bordered by four rivers: Djoumouna, Kinkoue, Lomba, and Loumbangala, which contribute water to a network of fish ponds. These ponds are periodically emptied for fish recovery or when irrigation canals are obstructed, creating stagnant water pools conducive to malaria vector proliferation. Furthermore, natural sites sustain Anopheles population year-round. The vegetation is composed of diverse species from to different plant families [13]. Extensive studies were conducted in the village between the 1970s to the 1990s [14][15], which showed that species within to the An.gambiaes.l. complex were the principal malaria vector species of malaria. Regarding insecticide resistance, An.gambiaes.l. species were sensitive to deltamethrin [16].

Adult culicids were collected in Kintele and Djoumouna (See Figure 1Figure 1 and Figure 2Figure 2) between September 2016 and August 2017, and in August 2019 in Djoumouna alone. Collections were conducted monthly, over 3 consecutive days in each locality. The bedrooms were categorized based on the absence or presence of LLINs and their condition (good or torn). These LLINs were sourced from the distribution carried out by the NMCP between 2011 and 2012, as well as from LLINs sold in markets and distributed through campaigns in the DRC.

The collection method used was the manual collection of residual fauna in bedrooms [17]. The collected Anopheles mosquitoes were used entomological assessment.

Morphological identification of the specimens was conducted at the genus and species levels using the dichotomous keys developed by Edwards [18] and Gillies and De Meillon [19]. After collection, specimens were placed in individual tubes containing silica gel and cotton, subsequently stored at −20˚C for further analysis.

Figure 1. Map of Kintele with the location of adult collection sites (source: CERGEC 2020).

Figure 2. Map of Djoumouna with location of adult collection sites (source: CERGEC 2020).

2.3.1. DNA Extraction

DNA was extracted from mosquitoes preserved with silica gel using the protocol described by Morlais et al. [20]. The procedure involved grinding the legs and wings of the mosquitoes in 200 μl of 2% CTAB (Cetyl Trimethylammonium Bromide) in an Eppendorf tube. Subsequently, the tubes were incubated at a minimum of 65˚C for 5 minutes using a dry bath, followed by a 3-minutes cooling period on ice. To each tube, 200 μl of chloroform was added. After manually shaking the tubes, they were centrifuged at 12,000 rpm for 5 minutes. The supernatant from each tube was carefully transferred to a new tube, to which 200 μl of isopropanol was then added. The new tubes were subsequently centrifuged at 12,000 rpm for 15 minutes, and after that, the supernatant was discarded, leaving the DNA pellet. This pellet was washed with 200 μl of cold 70˚C ethanol, followed by centrifugation at 12,000 rpm for 5 minutes. The tubes were then emptied of ethanol and placed in a speed-vac to allow the DNA pellet to dry for 5 minutes. The resulting DNA was dissolved in 20 μl of bi-distilled water and stored at −20˚C until PCR testing.

2.3.2. Identification of the An.gambiae Complex Species by PCR

The molecular identification of species belonging to the An.gambiaes.l. complex was carried out following the method developed by Wilkins et al. [21]. PCR was performed using Thermus aquaticus polymerase from Sigma, USA. The reaction volume was 12.5 μl, consisting of the following components: 6.4 μl of water; 1.25 μl of 10× buffer; 1.25 μl of 2 mM dNTP; 0.5 μl of each primer (UN, AR, ML, M1, and S1); 0.1 μl of Taq polymerase, and 1 μl of DNA template.

The PCR reaction mixture was initially incubated at 95˚C for 5 minutes, followed by 30 amplification cycles: denaturation at 95˚C for 30 seconds, hybridization at 58˚C for 30 seconds, and elongation at 72˚C for 30 seconds. A final elongation step was performed at 72˚C for 5 minutes.

The size of the amplified DNA fragments was assessed through electrophoresis on a 2% agarose gel. The agarose gel was prepared by dissolving 2 g of agarose in 100 ml of 0.5× TBE buffer, supplemented with 5 μl of ethidium bromide (BET). The gel was then subjected to electrophoresis at 170 volts for 20 minutes. Visualization was done under UV light. The sizes of the resulting bands were determined using a 1 kb DNA ladder as reference; 221 bp for An.gambiae, 333 bp for An. coluzzii and 387 bp for An. arabiensis.

In this study, data were collected in a Microsoft Office Excel 365 file, subsequent analyses were conducted using R software (version 8.4.1). Figures were generated with GraphPad-Prism software (version 3.6.3). For categorical variables involving Anopheles species, bedroom types and localities, proportions or percentages were calculated. To provide a measure of certainty, the binomial method was applied to establish the 95% confidence intervals for these parameters. Comparisons of proportions were carried out using the statistical tests such as Pearson chi-square (KHI-2) and Fisher’s exact. When dealing with anopheline densities equal to or greater than 5, nonparametric including the Kruskal-Wallis and Mann-Whitney tests were used. For means comparisons, the one-way ANOVA method was used, along with the and the Kruskal-Wallis test. To ascertain whether one locality exhibited a significantly higher percentage of An. coluzzii compared to another, the Odds Ratio was used with a 95% confidence interval.

A total of 1818 female mosquitoes were collected, comprising four genera: 1131 Anopheles, 681 Culex (with Culex quinquefasciatus), 5 Mansonia (Mansonia uniformis) and 1 Aedes with (Aedes albopictus). Among the Anopheles mosquitoes, 1130 (99.9%) were females belonging to the Anophelesgambiaes.l. complex, and 1 female (0.1%) belonged to the Anopheles funestus group.

Table 1. Monthly distribution of Anophelesgambiaes.l. complex females in Kintele and Djoumouna (September 2016 to August 2017, and August 2019 in Djoumouna alone).

Furthermore, within the An. gambiae complex 136 (12.04 %) females were collected in Kintele and, and a larger number of 994 (87.96 %) in Djoumouna.

The statistical analysis indicated a significant difference between the collection localities and the diversity of collected species (χ² = 1301.2; p-value ˂ 0.001).

The monthly distribution of An.gambiaes.l. females, categorized by bedroom type, is detailed in Table 1.

Table 1 shows the number of females collected based on bedroom types in both locations. The highest average density was collected in bedrooms with torn LLINs (3.68 females), while the lowest density was found in bedrooms with LLINs in good condition (0.40 females). In November and December 2016, no dwellings with bedrooms lacking LLINs were visited.

Variations in population densities of the An.gambiae complex are shown in Figure 3Figure 3 and Figure 4Figure 4.

Figure 3. Monthly variation of Anophelesgambiaes.l. densities according to the absence or presence of LLINs in good condition or torn in Kintele between September 2016 and August 2017.

In bedrooms without LLIN, anopheline densities increased slightly from September to March, reaching a peak in April. Subsequently, they remained relatively constant between June and August.

For bedrooms with torn LLINs, anopheline densities showed a consistent pattern from September to August.

Similarly, in bedrooms with LLINs in good condition, anopheline densities were relatively constant from September to August.

Statistical tests revealed no significant differences in anopheline densities based on collection month (Kruskal-Wallis test; p-value = 0.348) or bedroom type (Kruskal-Wallis test; p-value = 0.1821). Additionally, the Kruskal-Wallis test indicated that anopheline densities in bedrooms without LLINs and those with torn LLINs were statistically identical (Kruskal-Wallis test; p-value = 0.063).

In bedrooms without LLINs, anopheline densities slightly increased from September to October, followed by a decrease from November to December (Figure 4Figure 4). Densities then increased from December to February, peaking in February. A significant decrease in density was observed in June before a gradual increase through July to August 2019.

Figure 4. Monthly variation of Anophelesgambiaes.l. densities based on the absence or presence of LLINs in good condition or torn between September 2016 and August 2017 and in August 2019 in Djoumouna.

For bedrooms with torn LLINs, four peaks in anopheline densities were observed in October, February, April, and August. Densities then slightly decreased in August 2019.

In bedrooms with LLINs in good condition, densities remained relatively stable from September to June, showing minimal variation throughout the study period. Thus, no distinct peak was observed. Densities then declined from June to August, nearly reaching zero.

Statistical tests indicated that there was no significant difference between in anopheline densities based on collection months (Kruskal-Wallis test; p-value = 0.1213). However, the Kruskal-Wallis test revealed a significant difference in anopheline densities based on bedroom type (Kruskal-Wallis test; p-value = 0.0003).

A total of 343 females belonging to the An.gambiae complex was identified by PCR in the two study localities. Molecular analysis of the sampled females reaveled that the An.gambiaes.l. complex consisted of An.gambiae (72.3%), An. coluzzii (27.1%), An. arabiensis(0.3%) and hydrids (between An.gambiaeand An.Coluzzii) (0.3%) in both locations. The Kruskal-Wallis statistical test showed that there was a significant difference in the species collected between the two localities (Kruskal-Wallis; p-value = 0.0426). Furthermore, a significant difference was observed in the species composition of the Anophelesgambiaes.l. complex in the two localities.

The seasonal distribution of Anophelesgambiaes.l. complex in Kintele and Djoumouna is outlined in Table 2.

Table 2. PCR identification of Anophelesgambiaes.l. complex species collected during this study.

Table 3. Identified species of the Anophelesgambiaes.l. complex.

Figure 5. Monthly distribution of Anophelesgambiaes.l. based on collection seasons in Kintele and Djoumouna between September 2016, August 2017 and in August 2019 in Djoumouna alone.

Table 3 and Figure 5Figure 5 present the species within the An.gambiaes.l. complex identified by season. All three species were collected during the rainy season.

The KHI-2 statistical test revealed a significant difference in the densities of An.gambiae and An. coluzzii between Kintele and Djoumouna during the rainy season. The KHI-2 statistical test results for each species were as follows: (KHI-2 = 26.10; p-value ˂ 0.001) for An.gambiae and (KHI-2 = 21.32; p-value ˂ 0.001) for An. coluzzii. Similarly, in the dry season, the test demonstrated a significant difference in the densities of An.gambiae andAn. coluzziibetween Kintele and Djoumouna (KHI-2 = 32.87; p-value ˂ 0.001).

In Kintele, during the rainy season, the monthly distribution of An.gambiae species was identified using the Kruskal-Wallis test (p-value = 0.0058). An.gambiae was the only species identified during the dry season. Both An.gambiae andAn. coluzzii exhibited significantly higher abundance during the rainy season compared to the dry season (Kruskal-Wallis test; p-value = 0.0477).

In Djoumouna An.gambiae exhibited significantly greater prominence than other species during both the rainy season (Kruskal-Wallis test; p-value = 0.0006) and the dry season (Kruskal-Wallis test; p-value = 0.0417).

Figure 6. Diagnostic gel for species identification of the An.gambiaes.l. complex.

Caption:

1) Wells 1 and 16: 1 kb base-pair ladder.

2) Wells 2, 5 and 14 : Individuals An. coluzzii (C).

3) Wells 3, 4, 6, 7, 8 to 12 and 15: Individuals An.gambiae (G).

4) Well 13: An.gambiae/ An. coluzzii hybrid individual (H).

Figure 6Figure 6 shows an agarose gel showing separated DNA bands. Within this gel, two species were identified, and the most identified species was An.gambiae, comprising 10 out of the 14 identified females. In the case of An. coluzzii, 3 females out of 14 specimens were identified. Additionally, one hybrid female was identified.

The assessment of the entomological indices and transmission evaluation gives the following results:

Aggressive density represents the average number of bites an individual receives in the specified area within a unit of time.

Parity rate indicates the proportion of parous females (those having already laid eggs) within a sample of anopheline population.

The sporozoite index is the proportion of females observed with sporozoites in their salivary glands among the dissected females.

The entomological inoculation rate expresses the number of infective bites an individual experiences within a specific time frame. It is a measure of the intensity of malaria transmission within a given area.

These results show that the highest aggressive density (0.22 p/h/n) was recorded in bedrooms without LLINs, while the lowest (0.07 p/h/n) was observed in bedrooms with LLIN in good condition.

1) Regarding the number of infected anopheles, 2 females were collected from bedrooms with LLINs in good condition, in contrast to 1 female from bedrooms without LLIN. Notably, no infective females were collected from bedrooms with torn LLINs.

2) The sporozoite index (SI) exhibited a higher greater value (4.17%) in bedrooms with LLINs in good condition, while it was null in bedrooms with torn LLINs.

3) Concerning the rate of parity, bedrooms with torn LLINs showed a higher rate (38.09%), while bedrooms with LLINs in good condition displayed a lower rate (28.7%).

4) The entomological inoculation rate indicated one infective mosquito bite per person every 5.5 months (166.6 days) in bedrooms without LLINs, and one infective bite per person every 11.1 months (333 days) in bedrooms with LLIN in good condition.

Statistical tests showed significant difference in aggressive rates across thedifferent bedroom types (KHI-2 = 24.94; p-value ˂ 0.0001). Notably, the aggressive rates in bedrooms without LLINs and with torn LLINs were significantly different (KHI-2 = 8.55; p-value = 0.0034). While there was no significant difference in the number of infective females and sporozoite indices between bedroom types (Kruskal-Wallis test; p-value = 0.9723), the same was true when comparing bedrooms without LLINs and with torn LLINs (Fisher’s exact test; p-value = 0.8541). Additionally, there was no noteworthy difference in parity rates when comparing the various bedroom types of (KHI-2 = 2.53; p-value = 0.2823). Likewise, the party rates in bedrooms without LLINs and with torn LLINs were identical (KHI-2 = 0.21; p-value = 0.6501). Furthermore, disparities in entomological inoculation rates were observed among different bedroom types (KHI-2 = 77.34; p-value ˂ 0.0001). Lastly, significant differences in entomological inoculation rates were noted between bedrooms without LLINs and with torn LLINs were significantly different (KHI-2 = 101.14; p-value ˂ 0.0001).

Table 4 presents the entomological indices of malaria transmission over the entire study period in Djoumouna.

These findings reveal that the most substantial aggressive density (1.73 p/h/n) was recorded in bedrooms lacking LLINs. In contrast, the lowest density (0.13 p/h/n) was recorded in bedrooms with LLINs in good condition.

1) Regarding the number of infected anopheles, the highest number (15 females) was obtained from bedrooms with torn LLINs. Conversely, the lowest number (1 female) was found in rooms with LLINs in good condition.

2) Regarding the sporozoite index (SI), it was higher (2.14%) in bedrooms without LLINs and in bedrooms with LLINs in good condition. The lowest index (1.86%) was recorded in bedrooms with torn LLINs.

3) Regarding the rate of parity, the highest rate (82.23%) was observed in bedrooms with torn LLINs, while the lowest rate (4.39%) was found in bedrooms with LLINs in good condition.

4) Regarding the entomological inoculation rate, it was estimated to one infective mosquito bite per person every 27 days in bedrooms without LLINs. In bedrooms with LLINs in good condition, there was one infecting bite per person every 1.2 years (370.3 days). For bedrooms with torn LLINs, the rate was one infective bite per person every 2.2 months (66.6 days).

Table 4. Entomological indices by type of bedroom in Djoumouna.

Statistical tests show significant differences in aggressive rates among the various bedroom types (KHI-2 = 431.78; p-value ˂ 0.0001). The aggressive rates in bedrooms without LLINs and with torn LLINs were significantly different (KHI-2 = 17.25; p-value ˂ 0.0001). However, there was no significant difference in the number of infective females and sporozoite indices among the different types of bedroom types (Kruskal-Wallis test; p-value = 0.9712). Likewise, no significant difference was observed between bedrooms without LLINs and those with torn LLINs (Fisher’s exact test; p-value = 0.7107). Similarly, there was no significant difference in parity rates across the different bedroom types (KHI-2 = 1.33; p-value = 0.5152). Furthermore, the parity rates in bedrooms without LLINs and with torns LLIN were similar (KHI-2 = 0.74; p-value = 3889). Lastly, a significant difference in entomological inoculation rates between the different bedroom types (KHI-2 = 140.37; p-value ˂ 0.0001) was noted.