Acta Tropica

Volume 114, Issue 2, May 2010, Pages 116-122

Malaria vectors in the Republic of Benin: Distribution of species and molecular forms of the Anopheles gambiae complex

• Luc Djogbénou ^ab⁎,
• Nicole Pasteur ^c,
• Sahabi Bio-Bangana ^a,
• Thierry Baldet ^ab,
• Seth R. Irish ^d,
• Martin Akogbeto ^a,
• Mylène Weill ^c,
• Fabrice Chandre ^b

1. ^⁎ Corresponding author. Present address: Institut Régional de Santé Publique, Ouidah, 918 Cotonou, Benin. Tel.: +229 21341674; fax: +229 21341672.
2. ^a Centre de Recherche Entomologique de Cotonou, 06 BP 2604 Cotonou, Benin
3. ^b Laboratoire de Lutte Contre Les Insectes Nuisibles/Institut de Recherche Pour le Développement, 911 Av. Agropolis, 34394 Montpellier, France
4. ^c Université Montpellier 2 - CNRS, Institut des Sciences de l’Evolution, Equipe Génétique de l’Adaptation, C.C. 065, Place Eugène Bataillon, 34095 Montpellier, France
5. ^d London School of Hygiene and Tropical Medicine, Keppel Street, WC1E 7HT London, UK

Abstract

Members of the Anopheles gambiae complex are among the best malaria vectors in the world, but their vectorial capacities vary between species and populations. A large-scale sampling of An. gambiae sensu lato was carried out in 2006 and 2007 in various bioclimatic areas of Benin (West Africa). The objective of this study was to collate data on the relative frequencies of species and forms within the An. gambiae complex and to produce a map of their spatial distribution. Sampling took place at 30 sites and 2122 females were analyzed. Two species were identified through molecular methods. The overall collection showed a preponderance of An. gambiae s.s., but unexpectedly, An. arabiensis was reported in the coastal-Guinean bioclimatic area characterized by a mean annual rainfall of >1500 mm where only An. gambiae s.s. was reported previously. Our study of Benin indicates that An. arabiensis would be adapted not only to the urban areas but also to the rural humid regions. Among 1717 An. gambiae s.s., 26.5% were of the M form and 73.3% were S form. Few hybrid specimens between the M and S forms were observed (0.2%). Only the spatial distribution of the M form appears to be mainly a function of bioclimatic area.

Factors that influence the distribution of these malaria vectors are discussed. This study underlines the need of further investigations of biological, ecological, and behavioral traits of these species and forms to better appreciate their vectorial capacities. Acquisition of entomological field data appears essential to better estimate the stratification of malaria risk and help improve malaria vector control interventions.

1. Introduction

In West Africa, An. gambiaesensu stricto and Anopheles arabiensis constitute the main vectors of malaria. Both species are members of the Anopheles gambiae complex that comprises at least seven sibling species (Gillies and Coetzee, 1987; Coetzee et al., 2000). An. gambiaes.s. is subdivided into reproductively isolated sub-populations. Five chromosomal forms were initially identified (Forest, Savanna, Bamako, Mopti, and Bissau) based on the patterns of chromosome 2 inversions (Coluzzi et al., 1985; Touré et al., 1998). The distributions of mosquito species are related to bioclimatic factors, and in West Africa, it appears that the different chromosomal forms of An. gambiae s.s. may occur sympatrically but are segregated environmentally (Coluzzi et al., 1979; Coluzzi et al., 2002). Subsequent studies revealed that An. gambiae s.s. contains two molecular forms, M and S, recognizable by the differences of their rDNA sequences, either in the intergenic spacer or in the internal transcribed spacer (della Torre et al., 2001; Gentile et al., 2002). Both molecular forms co-exist in West Africa whereas the S form is the only one reported in eastern Africa (della Torre et al., 2005).

Currently, identification of field-collected specimens of the An. gambiae complex requires multiple steps. Individual mosquitoes are first identified to the species level by a PCR reaction based on species specific differences in the rDNA region, which includes part of the 28S coding region and part of the IGS region (Scott et al., 1993). In a second step, An. gambiae s.s. specimens are identified at the molecular form level by either one method based on PASA (Favia et al., 1997), or another method based on RFLP (Fanello et al., 2002).

The M and S forms are presently considered as units of an on-going incipient speciation process (della Torre et al., 2001), and many studies have reported a partial or complete reproductive isolation. Evidence of reproductive isolation is based on three types of observations. First, even the M and S forms mate freely in the laboratory, yielding fertile progeny (Diabaté et al., 2007), M/S hybrids are absent or very rare in nature, notably in areas in sympatry (della Torre et al., 2001; Taylor et al., 2001; Wondji et al., 2002).

2. L. Djogbénou et al. / Acta Tropica Volume 114, Issue 2, May 2010, Pages 116-122

Second, genetic studies using microsatellite loci (Wang et al., 2001; Wondji et al., 2002; Lehmann et al., 2003; Stump et al., 2005) and/or insertion polymorphisms of short interspersed elements (Barnes et al., 2005) revealed an important genetic differentiation between sympatric M and S populations in Cameroon, Burkina Faso, Mali and Kenya. Overall, studies using molecular markers seem to indicate that the differentiation between the M and S forms is restricted to a few regions of the genome, one close to the centromere of the 2nd chromosome, and the other in the centromeric region of the X chromosome that contains the rDNA used for differentiating the two molecular forms (Gentile et al., 2001; Lehmann et al., 2003; Turner et al., 2005). Third, genes conferring resistance to insecticides display large frequency differences in M and S sympatric populations. The kdr allele (conferring cross-resistance to DDT and pyrethroids) is very frequent in the S form but not or less in the sympatric M mosquitoes from Côte d’Ivoire, Nigeria and Burkina Faso (Chandre et al., 1999; Awolola et al., 2003; Diabaté et al., 2003). However, this allele is present in both forms in Southern Benin, probably as a result of introgressive hybridization (Weill et al., 2000). Recently, the ace-1^R allele (responsible for cross-resistance to carbamate and organophosphate compounds) was found to have the same DNA sequence in M and S molecular forms from Côte d’Ivoire and Burkina Faso. Because the ace-1^S susceptible allele displays a high DNA sequence polymorphism, the presence of the same ace-1^R sequence in the two molecular forms indicates a recent introgression (Djogbénou et al., 2008). The large genetic heterogeneity observed within natural populations of An. gambiaes.s. is expected to be accompanied by biological, ecological and behavioral heterogeneities leading to different vector capacities. Previous study carried out on this subject in Burkina Faso showed that the M form was predominant in (and therefore better adapted to) permanent breeding sites such as rice fields, whereas the S form was predominant in temporal habitats notably rain-filled puddles which are productive during the wet season (Diabaté et al., 2005).

In order to improve the understanding of these adaptations, it appears necessary to acquire precise data on the spatial distribution of the members of An. gambiae complex as well as of the two molecular forms of An. gambiae s.s.

We report here the distribution of species and molecular forms of the An. gambiae complex observed in Republic of Benin, a West African country not yet investigated which covers three strongly contrasted bioclimatic regions. In the south, near the Atlantic coast, a Guinean-bioclimatic zone with two rainy seasons (April–July and September–November) and an average annual rainfall of >1500 mm with degraded tropical forest. In the north, a Sudanian semi-arid bioclimatic zone with only one rainy season from June to October (mean annual rainfall below 900 mm) and characterized by a dry savanna. In the center, an intermediate bioclimatic zone (tropical Sudano-Guinean climate) with an average rainfall of 1000 mm per year, characterized as humid savanna. As one travels from north to south, mean annual rainfall increases, the number of dry season months decreases, and the vegetation becomes taller and more compact. We compared data from this study to those acquired in the past in the same bioclimatic regions to discuss the long term variation in the relative frequency of each species and molecular form.

2. Materials and methods

2.1. Study areas

The study was conducted in 30 areas, distributed along a south-north transect across Benin, and representing permanent, semi-permanent and temporal breeding sites from urban and rural localities. Seven, eleven and twelve localities were sampled in the Guinean, Sudano-Guinean, and Sudanian bioclimatic zones, respectively (Table 1 and Fig. 1).

3. L. Djogbénou et al. / Acta Tropica Volume 114, Issue 2, May 2010, Pages 116-122

Fig. 1: Map of republic of Benin showing the position of the 30 study sites within each bioclimatic area.

2.2. Sample collections

Mosquitoes were collected as larvae or pupae during the rainy seasons, between March and July 2006, between April and May 2007, and in October 2007. The sampling was guided by the availability and the accessibility of larvae in breeding sites of An. gambiaes.l. Larvae and pupae were stored in separate bottles for each breeding site of each locality. They were transported to the CREC laboratory. Pre-imaginal stages were then reared to adults as follows: larvae from different breeding sites of a same locality were mixed in rearing bowls (each bowl contained larvae from several breeding sites) in order to ensure that the samples used for species and molecular forms identification will be representative of a locality and not of specific breeding sites.

4. L. Djogbénou et al. / Acta Tropica Volume 114, Issue 2, May 2010, Pages 116-122

Larvae of the An. gambiae were sorted out morphologically and fed with fish food (TetraMikromin, Tetra Werke, Germany). They were reared to adults under standard conditions, and deep frozen for further analyses.

2.3. Molecular analyses

DNA was extracted from individual adults (Collins et al., 1987). DNA from each specimen was first analyzed using the An. gambiae species specific PCR (Scott et al., 1993). Then the DNA of all mosquitoes identified as An. gambiae s.s. was subjected to the PCR-PASA assay described by Favia et al. (1997) for identifying the molecular M and S forms. The use of similar standardized techniques for sampling and processing mosquitoes across surveys ensured data consistency.

3. Results

3.1. Species distribution

Among the 2122 mosquitoes of the An. gambiae complex collected over 30 sites of the Benin Republic (Table 2), two species, An. gambiae s.s. and An. arabiensis, were identified (no other member of the complex was detected). They represented 81.4% and 18.6% of the identified mosquitoes, respectively.

An. gambiae s.s. was present in all samples, and was the only species present in 17 out of the 30 sampled sites. In other sites, its frequency varied between 8.1% and 98.8%. In the humid coastal-Guinean area, it represented 99.5% of mosquitoes of the An. gambiae complex. It was also more common than An. arabiensis in 8 of the 11 sites of the Sudano-Guinean zone and in 11 of the 12 sites of the Sudanian areas.

An. arabiensis was observed in less than half of the sampled sites (13/30 = 43%) where it was sympatric with An. gambiaes.s. It was rare in the coastal-Guinean area (frequency of 2.8% in only one of the seven studied sites: the urban site of Allada), but present in 7 of the 11 Sudano-Guinean sites and in 5 of the 12 Sudanian sites. This species was observed at high frequency (>50%) in several urban areas (Bohicon, Savalou, Glazoue, and Parakou) of the Sudano-Guinean and Sudanian areas (Table 2).

3.2. Identification of the molecular forms of An. gambiaes.s.

Of the 1728 An. gambiaes.s. collected, the majority (99.4%, n = 1717) was successfully identified to molecular level using PCR (Table 2). Over all samples, the S molecular form was predominant (73.3%). The two molecular forms of An. gambiaes.s. was sympatric in 15 out of the 30 studied sites. Four M/S heterozygotes (0.2% of the total analyzed) were found in four different sites where the two molecular forms were sympatric (Tori Bossito, Allada, and Niaouli in the coastal guinean zone, and N’dali in the sudanian zone).

The M form was present in all sites of southern Benin, including the coastal-Guinean zone and southern part of the Sudano-Guinean zone up to Bohicon. The M form was the only form present on the Atlantic coast (Cotonou and Pahou), and up to Bohicon, its frequency was above 20% in 7 of 8 sites (including 4 sites where it was about 60% or more). Then from Abomey to Kandi, it was absent or present at a frequency usually below 10%. The frequency of the M form shows a strong decline with increasing latitudes (Fig. 2). The exception is the Malanville site, where the M form was the only one present, probably because of the abundance of rice fields irrigated by the Niger River.

5. L. Djogbénou et al. / Acta Tropica Volume 114, Issue 2, May 2010, Pages 116-122

Fig. 2: Comparison of the frequencies of the M molecular form of Anopheles gambiae in the different sites between bioclimatic zones. The bar chart indicates the frequency of this molecular form in each site.

On the x-axis, the name of different localities is aligned from north to south.

The S form was the only molecular form of An. gambiaes.s. in 12 sites (3 in the Sudano-Guinean area and 9 in the Sudanian area), and it was the predominant form in 11 other sites (3 coastal-Guinean, 5 Sudanian-Guinean, and 3 Sudanian).

4. Discussion

In this study, we used molecular tools to accurately identify species and forms of the An. gambiae complex and, thus, to analyze their precise geographic distribution in the Republic of Benin. Such studies are important for the implementation of targeted vector control interventions, especially in this country where several large campaigns based either on massive distribution of long lasting nets or indoor residual spraying are currently supported by organizations like the Global Fund or President Malaria Initiative. This field survey is also the first one providing data on An. gambiae complex over large areas which have never been documented in Benin.

An. gambiae is a complex, with seven sibling species that are closely related and morphologically indistinguishable from each other by routine taxonomic methods. They are however different with respect to ecological and behavioral characteristics and to vectorial competence (White, 1974). In West Africa, An. gambiae s.s. and An. arabiensis are the two main species of the complex that transmit malaria, with the former being the most efficient vector due to its high anthropophily (Pates et al., 2001; Besansky et al., 2004).

The species of the An. gambiae complex observed in Benin were An. gambiaes.s and An. arabiensis. An. melas, member of the same complex, which has been reported in the coastal lagoon area of South Benin (Akogbeto and Di Deco, 1995) and to some extent along the whole coast of West Africa, is not considered here. Due to its ecological preferences, this halophilic species is limited to saline coastal habitats (e.g. mangrove swamps) which have not been sampled in our study. In addition to its geographical restriction, An. melas usually plays a secondary role in malaria transmission due to its strong zoophily and exophily. Despite it behavior, the role of An. melas during dry season need to be clarify.

An. gambiae s.s. was predominant (>80%) and had a wide distribution across the country. This agrees with a recent study by Corbel et al. (2007) carried out in 4 localities from Benin and conforms to the well known distribution of An. gambiae s.s. in West Africa (see review of Coetzee et al., 2000).

However, we observed large and sometimes unexpected variations in the relative proportion of An. arabiensis and An. gambiaes.s. when considering different localities. In West Africa, An. arabiensis is generally concentrated in the arid savannah areas due to its high tolerance of dry environment (Gillies and Coetzee, 1987). Previous studies conducted in Nigeria using cytogenetics (Coluzzi et al., 1979) and microsatellites (Onyabe and Conn, 2001) have revealed that An. arabiensis existed also in the Guinean region, south of the country. The authors hypothesized that An. arabiensis extended its range southward, from drier savannah into humid forest habitats. In Benin, the last study which can serve as a stepping-stone is that of Akogbeto and Di Deco (1995) which found An. arabiensis only in the dry savannah areas in the north. In the present study, we sampled An. arabiensis at relatively high frequencies in the centre of Benin, characterized by a mean annual rainfall of 1000 mm, (Glazoue, Savalou, Dassa-Zoume, Paouignan, Bohicon and Parakou). Furthermore, two individual mosquitoes belonging to An. arabiensis were found in a city (Allada) located in the Guinean climate area, a more humid region. Although differences in sampling technique may explain the disparity in distribution between the two studies and that only few specimens have been found in the southern humid area, our recent work suggests a change in the distribution of A. arabiensis in Benin. Thus, we give credence to the hypothesis made by Lindsay and Martens (1998) that An. arabiensis may be in process of expanding its range southward from arid savannah into more humid area affected by drought and/or human activities (deforestation and urbanization). This situation corroborates findings from the neighboring country, Nigeria, which has a same climatic pattern as that of Benin republic (Coluzzi et al., 1979; Onyabe and Conn, 2001; Kristan et al., 2003).

Several studies reported that abiotic and biotic characteristics related to climate, topographic, human disturbance of the natural environment (deforestation, urbanization, and irrigation) can favor the spread and colonization of new areas by efficient malaria vectors, increasing the risk of transmission (Antonio-Nkondjio et al., 2005; Guerra et al., 2006). Overall, do malaria vectors change their ranges in response to environmental changes?

6–7. L. Djogbénou et al. / Acta Tropica Volume 114, Issue 2, May 2010, Pages 116-122

Do they colonize new areas that have become suitable? Does the ecological niche shift over time? The answers to these questions can be used to examine factors affecting the dynamics of malaria vectors species distribution. Nevertheless further studies are needed to confirm our preliminary results: the spreading of An. arabiensis in south Benin, not only in urban settlement but also in rural environment.

An. arabiensis and An. gambiae s.s. have a different vectorial competency (White, 1974) but An. arabiensis is also known to have a higher behavioral plasticity with respect to feeding and resting (Gillies and De Meillon, 1968; Chauvet and Rajaonarivelo, 1973; Gillies and Coetzee, 1987). The spreading of An. arabiensis may affect the distribution and abundance of these species and can therefore affect the transmission of malaria. Furthermore, more recent surveys indicated that this species is less resistant to insecticides (Czeher et al., 2008; Ndjemai et al., 2009) and should be easier to control. However, such An. arabiensis spreading, particularly in West African cities, may induce an increase of its anthropophily and thus of its vectorial capacity that will severely raised the risk of urban malaria in the next future. The vectorial competency of An. arabiensis newly adapted to urban zone should now be compared to those of An. arabiensis rural mosquitoes and to An. gambiae s.s. Our findings emphasize the need to understand how An. arabiensis populations evolve in the field and how its high adaptive potential could influence the design and implementation of an integrated vector control program.

Our study shows that the S molecular form of An. gambiae s.s. is predominant with a large distribution across the country. This agrees with findings from a study in Nigeria (Onyabe et al., 2003) and similar to reports from other west African countries such as Cameroon (Wondji et al., 2002) and Burkina Faso (Dabiré et al., 2009). Finally, the level of hybridization between M and S forms is variable in West Africa. While, a very low frequency of hybrids (less than 2%) was observed in Burkina Faso, Côte d’Ivoire and Nigeria (della Torre et al., 2005), a relatively high frequency of hybrids (up to 24%) was recorded in Gambia and Guinea Bissau, both countries located in the extreme western part of the continent (Caputo et al., 2008; Oliveira et al., 2008). In Benin, we observed only 4 M/S hybrids among the 1717 An. gambiae s.s. tested (0.2%), and 3 of them were found in the coastal-Guinean bioclimatic region. This observation reinforces the hypothesis of incomplete but substantial barriers to gene flow between the two molecular forms in this country. In particular it strengthens the previous observation that insecticide resistance genes, such as the ace-1^R allele (conferring cross-resistance to organophosphates and carbamate insecticides) found in the two forms was the result of a recent and ongoing introgression (Djogbénou et al., 2008).

The evolution of insecticide and drug resistance and the complexity of vectorial system have seriously challenged standard public health approaches to control malaria and to date, alternative measures are desperately needed. One approach to controlling transmission that is being actively pursued by several laboratories is to develop mosquitoes that are genetically modified to resist infection and then drive the refractory genes into wild vector populations. Another approach that may be effective is based on sterile male release. Both approaches that target particular vector populations require a thorough understanding of the spatial and temporal distribution of species and forms of An. gambiae s.l.

Acknowledgements

This work was supported by a grant from Institut de Recherche pour le Développement (BSTD/DSF/IRD). It is the contribution ISEM-XX-2010 of the Institut des Sciences de L’Evolution de Montpellier (UMR CNRS-UM2 5554).

Authors are grateful to S. Koudénoukpo for his technical assistance during laboratory assays and mosquito collections.

References

• About ScienceDirect
• What is ScienceDirect
• Content details
• Set up
• How to use
• Subscriptions
• Developers

• Contact and support
• Contact and support

• About Elsevier
• About Elsevier
• About SciVerse
• About SciVal
• Terms and Conditions
• Privacy policy
• Information for advertisers