Spatial Distribution and Biodiversity of Anopheles Mosquito Species Across Climatic Zone in Burkina Faso: Implications for Malaria Vector Control

1. Introduction

Over the past two decades, substantial progress has been achieved in malaria control, especially through the large-scale implementation of preventive and therapeutic interventions [1,2]. Vector control strategies such as insecticide-treated nets (ITNs), and indoor residual spraying (IRS), have played a central role in malaria control. These interventions are credited with over 70% of the substantial decline in malaria cases during this period [1]. Despite these achievements, malaria remains a major public health concern in sub-Saharan Africa [3], accounting for over 263 million reported cases and nearly 600,000 deaths in 2023 according to the WHO. This persistence is driven by the emergence of drug resistance in Plasmodium species, insecticide resistance in mosquito populations, and behavioural changes that undermine the effectiveness of existing control tools [4,5]. The disease is caused by five Plasmodium species, with P. falciparum being the most prevalent and virulent in Africa [6]. Transmission occurs through female Anopheles mosquitoes, a genus comprising nearly 500 species worldwide, of which about 100 species are recognised as malaria vectors [7].

Mosquitoes (Culicidae) are widely distributed across tropical and temperate regions, comprising more than 3,500 species grouped into three subfamilies. Among these, Anopheles species are of major medical importance as malaria vectors [8,9]. Africa hosts a wide diversity of Anopheles mosquitoes, with hundreds of species described across the continent [10]. However, only a subset is responsible for most malaria parasites transmission, and their distribution varies across ecological zones. In sub–Saharan Africa, the dominant vectors are An. gambiae s.s., An. coluzzii, An. arabiensis, and An. funestus [11]. Other species such as An. nili, An. rufipes, An. pharoensis, and An. coustani typically occur at lower densities and act as secondary or potential vectors depending on local ecological conditions. The importance of these species varies across regions, reflecting differences in climate, vegetation, and human activities [12].

In Burkina Faso, the main malaria vectors are An. gambiae s.s., An. coluzzii, An. arabiensis, and An. funestus [13,14]. Ecological and climatic factors influence their distribution. Anopheles gambiae s.s. is typically associated with temporary, rain-dependent breeding sites, whereas An. coluzzii predominates in irrigated and semi-permanent habitats such as rice fields; An. arabiensis is now present in urban areas, probably reflecting its adaptation to human-modified environments [15].

Human-driven environmental changes, such as urbanisation, deforestation, and agricultural expansion, influence mosquito ecology [16]. These alterations modify habitat availability, species composition, and population abundance, which can increase in vector species adapted to those altered ecosystems. Consequently, human vector contact may increase, thereby affecting the transmission dynamics of mosquito-borne diseases, including malaria [16]. To better understand how these environmental changes influence vector populations and transmission risk, biodiversity indices were used to provide reliable quantitative tools for describing and comparing mosquito populations across spatial and temporal gradients. Species richness (S) indicates the number of species present, the Shannon index (H′) integrates both richness and evenness, while the Simpson index (D) reflects dominance patterns within the populations. By combining information on species richness and relative abundance, these indices offer valuable insights into community structure and help elucidate the ecological complexity underlying malaria transmission systems. Applying such metrics to Anopheles populations enhances understanding of vector diversity and informs adaptive, evidence-based control strategies [17]. Several studies have examined the diversity, distribution, and ecology of mosquitoes in different regions of Burkina Faso; however, most have focused primarily on malaria vectors and were geographically limited [9].

To address these limitations, the present study provides an updated, nationwide assessment of Anopheles species diversity across the climatic zones of the country, combining morphological and molecular identification methods with biodiversity indices. Understanding mosquito biodiversity within these changing environments is essential for designing effective surveillance and control strategies, which rely on accurate identification and characterisation of vector species across ecological contexts. By exploring spatial and ecological variations in vector composition, this work aims to enhance understanding of Anopheles biodiversity and support the development of more effective, locally adapted malaria control strategies.

2. Materials and Methods

2.1. Study Sites and Ecological Zones

Entomological surveys were conducted in 67 health districts across the regions of Burkina Faso from September to December 2022. The country has a tropical climate with two distinct seasons: a rainy season lasting 3–6 months (May–October) and a dry season of about 6 months (November–April). These 13 regions are located in three main ecological zones: the Sahelian, Sudano-Sahelian, and Soudanian (Figure 1). Each is characterised by distinct climatic and environmental features that influence mosquito vector distribution. The Sahelian zone, located in the northern part of the country, represents the driest ecological region. It has a mean annual temperature of 29.1 °C and rainfall ranging from 300 - 600 mm per year, mostly falling between June and September. It is characterised by sparse vegetation, consisting mainly of grasslands, thorny shrubs, and drought-tolerant tree species. Agricultural activity is limited, with sorghum and millet as the dominant crops, typically cultivated near seasonal water bodies that may also serve as mosquito breeding habitats. The Sudano-Sahelian zone has a dry season from November to April, with average annual rainfall ranging between 600 and 900 mm and a mean annual temperature of 28.4 °C. Dry forests, savannas, and scattered trees dominate vegetation. Agricultural production is mainly based on cereal cultivation, particularly millet and maize, while vegetable farming is also carried out on a smaller scale. Furthermore, human activities such as vegetable cultivation and unregulated urban expansion contribute to the development of mosquito breeding habitats. The Soudanian zone, which receives the highest rainfall, 900–1200 mm annually, mainly from May to October, is rich in vegetation and constitutes the most humid ecological region in the country [18]. Gallery forests, dense savannas, and wooded landscapes dominate. Cotton and cereal crops are widely cultivated, and the combination of abundant rainfall, lush vegetation, and standing water creates favourable ecological conditions for mosquito proliferation [19,20]. Malaria control in these regions is based on the use of antimalarial drugs combined with preventive measures, including insecticide-based interventions and seasonal malaria chemoprevention for children under five years of age [21].

2.2. Mosquito and Metadata Collection

Indoor-resting mosquitoes were collected early in the morning (06:00–09:00 a.m.) using pyrethroid spray catches (PSC) from ten randomly selected houses per site across 67 health districts in Burkina Faso. [22]. Each collection site was georeferenced, and mosquito specimens were identified to the genus level using standard morphological keys (Gillies and Coetzee, 1987). All specimens were counted, identified, and preserved in 80% ethanol for subsequent analyses. PCR assays were performed on An. gambiae complex (s.l.) specimens from 33 districts, selected based on the three main climatic zones, to determine species composition, specifically distinguishing An. gambiae s.s., An. coluzzii, and An. arabiensis.

2.3. Molecular Identification of Anopheles gambiae s.l.

Molecular analyses were conducted to identify An. gambiae complex species. Genomic DNA was extracted from a single mosquito using a 2% CTAB protocol [23]. The extracted DNA served as the template for PCR amplification targeting the SINE200 region to identify species within the An. gambiae complex (s.l.) (Scott et al., 1993; Santolamazza et al., 2008). A single pair of primers, S200 X6.1 F (5′-TCGCCTTAGACCTTGCGTTA-3′) and S200 X6.1 R (5′-CGCTTCAAGAATTCGAGATAC-3′), was used with identification based on the size of the amplified fragment: approximately 479 bp for An. coluzzii, 249 bp for An. gambiae s.s., and 223 bp for An. arabiensis. PCR reactions were carried out in a 20 μl reaction, which contained 3 µL of template DNA, 0.4 µL of each primer, 4 µL of 5× Hot Firepol Blend Master Mix, and 12.2 µL of molecular-grade water. Thermocycler conditions were 95°C for 15 minutes followed by 35 cycles of denaturation at 95°C for 1 minute, annealing at 62°C for 1 minute, extension at 72°C for 1 minute, final extension step at 72°C for 5 minutes, and a 4°C hold. PCR products were visualized on 2% agarose gels stained with ethidium bromide.

2.4. Data Management and Metadata

We managed our data in accordance with our approved data management plan (DMP), an approach was designed to uphold the FAIR principles, making data Findable, Accessible, Interoperable and Reusable, thereby guaranteeing its integrity and supporting reproducible results [24]. Field data were collected using the KoBoToolbox platform [25] on ruggedized tablets, with customized electronic forms incorporating validation rules to minimize entry errors. A two-step verification process [26] was applied, and the two resulting datasets were programmatically cross validated using the arsenal package [27], with discrepancies resolved by reference to original filed records. Missing data were handled to predefined outlined in our DMP. For sporadic missing environmental covariates (<5%), multiple imputation by chained equation (MICE) was performed using the nice package [28] and results from five imputed datasets were pooled. The final validated dataset was archived in non-proprietary CSV format on secure, access-controlled servers with automated daily backups. In line with our commitment to data sharing, the de-identified dataset will be deposited in the Dryad Repository [29] upon manuscript acceptance, assigned a DOI and released under CCO waiver to ensure global accessibility and reusability.

2.5. Statistical Analyses

Statistical analyses were conducted in R version 4.5.1 to investigate mosquito species abundance, diversity patterns and their association with climatic zones and seasons. Species abundance and relative abundance were computed by aggregating counts of each Anopheles species by district and period. Relative abundance for each species was calculated as the proportion of its abundance to the total abundance within each district/period combination according to the following formula:

$${Relative Abundance}_i={\displaystyle \frac{N_i}{{\sum }_{j=1}^S{N}_j}}$$

where Ni represents the abundance of species i and S denotes the total number of species in the population, which allows comparisons of species composition across spatial and temporal scales while accounting for variations in sampling effort [30].

Biodiversity metrics (species richness, Shannon diversity index and Simpson diversity index) were calculated to characterise population structure. Species richness (S) was quantified as the number of species with non-zero abundance in each district/period combination. The Shanon index (H’) was computed as:

$$H\prime =-\sum _{i=1}^S{p}_{i}ln\left(p_i\right)$$

where pi represents the proportion of species i in the population. This index quantifies species diversity by combining richness and evenness with greater emphasis on rare species (Shannon, 1948). The Simpson index (D) was calculated as:

$$D=1-\sum _{i=1}^S{p}_i^2$$

which represents the probability that two randomly selected individuals belong to the same species (range 0 < D ≤ 1), was calculated. For the primary analysis, D was transformed into a component, D’ = 1 – D, representing the probability that two randomly selected individuals range [0, 1). This bounded metric (D) was selected for subsequent beta regression modeling. All diversity indices were computed using the “vegan version 2.7.1” package in R with appropriate handling of zero-inflated data through boundary adjustment (0.001 for zeros and 0.999 for ones) [32].

PCR-confirmed species data from An. gambiae s.l. specimens were used to validate morphological identifications and to assess species distribution patterns across climatic zones.

A contingency table of species counts by zone was built and chi-square tests of independence were performed to evaluate the association between climatic zone and species composition. The strength was quantified using Cramer’s V:

$$V=\sqrt{{\displaystyle \frac{{chi}^2}{n\times min\left(r-1,c-1\right)}}}$$

where chi² denotes chi-square statistic, n represents the total sample size, and r and c correspond to dimensions of the contingency table [33]. Species proportions with 95% confidence intervals were calculated using the Clopper-Pearson exact method for binomial proportions [34] and visualised to highlight compositional differences among zones.

To investigate the effects of climatic zone and season on biodiversity patterns, we used beta regression models with the Simpson index as the response variable. Given the bounded nature of diversity indices (0-1) and the presence of zero values, we implemented both standard beta regression and zero-inflated beta regression approaches using the “glmmTMB version 1.1.12” package [35]. Models were fitted with climatic zone and season as fixed effects and district as a random. Model selection was performed using the Akaike Information Criterion (AIC) to identify the most parsimonious model [24,36]. Separate models were fitted for each climatic zone to account for zone-specific dynamics, followed by a combined model including all zones to assess overall patterns.

In cases where frequentist models exhibited convergence issues or poor fit, we employed Bayesian beta regression models with the “brms version 2.23.0” package as an alternative [37]. These models were implemented with four chains, 2000 iterations (1000 warmup) and default weakly informative priors. Convergence was assessed using the R-hat statistic (target <1.01) and effective sample sizes (target > 400) [38]. Posterior predictive checks were conducted to validate model fit, and summaries were reported with 95% credible intervals.

3. Results

3.1. Species Composition of Anopheles Mosquitoes Across Burkina Faso

A total of 30,521 Anopheles specimens were collected between September and December 2022 across 67 districts in three climatic zones of Burkina Faso: Sahelian, Sudano-Sahelian, and Soudanian. Anopheles gambiae s.l. was the most dominant species, representing 94.38% (28,820/30,521) of the overall specimens collected (Figure 2). Other species included An. rufipes (4.06%; 1,239/30,521), An. funestus (1.36%; 415/30,521), An. pharoensis (0.11%; 33/30,521), An. nili (0.04%; 12/30,521), and An. coustani (0.01%; 2/30,521) (Table 1).

The Sahelian Zone was shown to be dominated by An. gambiae s.l., accounting for 91.85% (4,743/5,164) of the total collected mosquitoes. Anopheles rufipes was the secondary (8.03% (415/5,164)) dominant species after An. gambiae complex. The other Anopheles species were composed of An.funestus, An.pharoenis, An.coustani collected at frequencies less than 1% each. These results confirm the predominance of An.gambiae s.l. as the main malaria vector in this climatic area. The mosquito composition remained relatively consistent during the sampling period across the Sahelian region, with An. gambiae being the predominant species from September to December, with frequencies exceeding 85%. We observed clear geographical variability at the district level over the course of the trial. For example, Gorom-Gorom district reported steady counts from September (420) through December (224), whereas Kongoussi recorded no mosquitoes in September and December, but reported 229 An. gambiae s.l. in October and only 2 An. rufipes in November. In October, An. rufipes was found at 18.98% (141/743) in the central district of Dori and at 24.41% (208/852) in Sebba, both within the same region (Table S1). The presence of this species suggests that local environmental factors, such as specific larval habitat characteristics, favour An. rufipes in these areas, even within landscapes where An. gambiae s.l. remains the dominant species.

In the Sudano-Sahelian zone, An. gambiae was the predominant Anopheles species, accounting for 95.19% (17,602/18,492) of the collected mosquitoes. The other species included An. rufipes at 4.1% (751/18,492), along with An. coustani and An. funestus, which were collected at low frequencies. Regarding the spatiotemporal distribution, the dominance of An. gambiae was consistent across most districts and months, particularly during the peak abundance observed in October, which represented 95.89% (10,586/11,040) of all collections. However, significant temporal shifts were noted in certain districts as the season progressed into November and December. For example, in Kombissiri, the proportion of An. gambiae s. l. decreased from 84% (420/500) in October to 37.37% (37/99) in November, coinciding with an increase in the proportion of An. rufipes from 15% (75/500) to 47.47% (47/99). Similarity, in Solenzo a notable shift was observed between October and December, with An. rufipes accounting for 44.66% (46/103) of the catches in October prior to An. gambiae s.l. Re-establishment dominance at 96.87 % (433/447) in December (Table S1). These spatiotemporal dynamics suggest a complex ecological; succession of anopheline species which may have implications for seasonal malaria transmission dynamics and vector control strategies

The anopheline species in the Soudanian zone was overwhelmingly dominated by the An. gambiae complex, which represented 94,32% (6,457/6,865) of the total specimens collected. The following secondary species was An. funestus with 4.32% (301/865), while other species such as An. rufipes (1.06%), An. pharoensis (0.22%) and An.nili (0.01%) were significantly rarer. This indicated that An. gambiae complex is the primary vector in this ecological zone with An. funestus emerging as a significant secondary vector of interest. About the space-time pattern, the dominance of the An. gambiae complex was highly dominant across most districts and throughout the sampling period from September to December, consistently accounting for over 90% of the monthly total catches (e.g. up to 94.85% in October). However, An. funestus reached significant levels in certain districts. For instance, in Gaoua district, An. funestus accounted for 38.64% (153/396) of the catches in October and became the most prevalent Anopheles species in December at 66.67% (26/39). Similar proportions of this species were observed in Dano in November, at 46% (23/50), and in Léo during December, at 57.14% (4/7) (Table S1). These findings confirm the overall dominance of the An. gambiae complex, but in some districts and periods, An. funestus may play a relatively larger role in malaria parasite transmission and monitoring.

3.2. Biodiversity Indices Across Climatic Zones and Collection Periods

Biodiversity patterns of Anopheles mosquitoes were assessed across 67 districts in three climatic zones of Burkina Faso (September–December 2022) using species richness, Shannon (H′), and Simpson (D) diversity indices. Species richness quantified the number of Anopheles species per zone, while the Shannon and Simpson indices integrated both species abundance and evenness to describe population diversity and dominance patterns, revealing marked spatiotemporal variation driven by seasonal and environmental factors (Figure 3).

The arid sahelian Zone exhibited the lowest biodiversity with richness ranging from 1-3 species per district-period (Figure3). This low diversity was most observed in Barsalogho, Kongoussi and Thiou where only one species was collected, resulting in Shannon and Simpson diversity indices of zero across all sampling periods. The highest diversity was recorded in Sebba during October (Richness = 3, Shannon = 0.85 and Simpson = 0.46), reflecting temporary habitat suitability during peak of rainfall (Table S2). Dori and Titao showed moderate diversity in October to November (Shannon = 0.78 – 0.53, Simpson = 0.41-0.51), but diversity reduced to zero by December as water dried up. This temporal decline reveals the significant environmental constraints in the region. Only An. gambiae s.l. (An. coluzzii) appears to survive in the few remaining permanent water sources, while other species cannot withstand the dry conditions.

The transitional Sudano Sahelian Zone displayed intermediate biodiversity with greater spatiotemporal heterogeneity. Richness varied widely (0-5 species) with Solenzo recording the highest exceptional diversity in October (richness = 5, Shannon = 1.14 and Simpson = 0.56) in this zone was comparable to Soudanian diversity (Table S2). In Kombissiri, species diversity remained relatively high across sampling periods. A notable peak in November (richness =3, Shannon = 1.76, and Simpson = 0.77) confirms that the area’s variety of larval breeding sites provides a stable foundation for a multi-species mosquito population. The zone’s biodiversity reflects the interaction between seasonal water availability and competitive dynamics among An. gambiae s.l.

The humid Soudanian Zone exhibited the highest and most stable biodiversity, with richness spanning 0 – 4 species but fewer zero diversity periods (Table S2). During the peak diversity period of October-November, several districts, notably Dano, Do, Gaoua and Kampti, consistently showed high species richness. This was accompanied by elevated Shannon diversity indices, surpassing 1.07 in Dano, 1.42 in Do and Kampti 0.50 during November, indicating a more balanced species distribution as reflected in their respective Simpson indices (0.53 and 0.51). Banfora and Zabré showed the widest temporal variation, with October peaks (richness = 3-4) collapsing to monospecific or absent collection by December, reflecting seasonal breeding site dynamics in floodplain areas. Gaoua’s sustained diversity into December (richness = 3, Shannon = 0.97, Simpson = 0.56) and suggested permanent water bodies supporting year-round multi-species communities, potentially including An. funestus alongsides An.gambiaes s.l. (Table S2). The higher Simpson values in the zone indicate more equitable species proportions, consistent with greater habitat heterogeneity and reduced environmental stress compared to northern zones.

3.3. Modeling Simpson Diversity Across Climatic Zones

Simpson’s diversity index was calculated for the Anopheles population across 67 districts in Burkina Faso, spanning three climatic zones: Sahelian (n = 21 samples), Sudano-Sahelian (n= 96), and Soudanian (n=57), for a total of 174 sampling observations representing site-level Anopheles assemblages from dry (November to December) and wet (September to October) seasons. Due to the high prevalence of structural zeros (37.4% overall, ranging 36.5-42.9% by zone) reflecting single species dominance, we applied beta regression on epsilon transformed values (ε = 5.8×10⁻⁴ to 1.1×10⁻³) using the glmmTMB package. For each zone, we compared null models against season effects via Akaike Information Criterion (AIC), with null models consistently preferred (Δ AIC = 1.2-2.0), indicating negligible seasonal influence on diversity. Globally, the null model outperformed the full zone × season interaction (AIC = -639 vs. -631, ΔAIC = 7.7), providing moderate evidence against climatic effects. The mean Simpson’s index was uniformly low across zones (0.094-0.132), with a global estimate of 0.110 (95% CI: 0.087-0.134), suggesting persistent dominance of few Anopheles species. Model diagnostics using the DHARMa package confirmed adequate fit at the zone level (dispersion p > 0.75; uniformity p > 0.01), though global uniformity showed minor deviation (p < 0.001), likely attributable to structural zero-inflation rather than model misspecification. This pattern underscores ecological homogeneity in Anopheles assemblages across Burkina Faso’s climatic gradients, with single species dominance (often An.gambiae s.l.) persisting irrespective of aridity of seasonality. Such low diversity may constrain malaria transmission dynamics and vector control efficacy, warranting species specific intervention over zone tailored strategies.

3.4. Composition of Anopheles gambiae s.l. Species Across Climatic Zones

Polymerase chain reaction (PCR) analysis was conducted on 2,026 An. gambiae s.l specimens collected between September and December across three climatic zones in Burkina Faso (Table1). The molecular identification revealed clear climatic structure of sibling species within the complex: An. coluzzi, An. gambiae s.s., and An. arabenssis (Figure 4). The analyses revealed distinct species distributions across the climatic gradient. Anopheles coluzzii was the predominant species in both the Sahelian (74.9%, 143/191), and Sudano-Sahelian (71.0%, 632/890) zones (Table S4). In the Soudanian zone, An. gambiae s.s. accounted for the majority of specimens (53.7%, 508/945), while An. coluzzii was less frequent (36.1%, 341/945). Only a small proportion of mosquitoes collected were An. arabiensis, which varied from 8.9% (17/191) in the Sahelian zone to 14.3% (127/890) in the Sudano-Sahelian zone (Table 1). A chi-square test showed that species distribution patterns were strongly dependent on climatic zone (χ² = 353.86, df = 4, p < 0.00001), with a moderate effect size (Cramer’s V = 0.3). The analysis of residuals provided clear detail, showing that An. gambiae s.s. was disproportionately abundant in the Soudanian Zone (residual = 11.06) while being significantly underrepresented further north (residual = - 4.05 in Sahelian, - 9.52 in Sudano-Sahelian). PCR-based diversity indices showed all zones had richness = 3 species, but diversity increased southward. Analysis of diversity indices revealed a pronounced major latitudinal cline. Population evenness was highest in the Soudanian zone (Shannon = 0.93, Simpson = 0.57) due to the influential presence of An. gambiae s.s., decreased in the transitional Sudano-Sahelian Zone (Shannon = 0.80, Simpson = 0.45), and was most suppressed in the arid Sahelian zone (Shannon = 0.72, Simpson = 0.40). This pattern results from the increasing monopolization of the mosquito population by An. coluzzii as conditions become drier.